Salesforce Admins Podcast

Today on the Salesforce Admins Podcast, Admin Evangelist Josh Birk sits down with Kat Holmes, Chief Design Officer and EVP at Salesforce.

Join us as we chat about diversity, accessibility, and her book, Mismatch: How Inclusion Shapes Design.

You should subscribe for the full episode, but here are a few takeaways from our conversation with Kat Holmes.

What is a mismatch?

I brought Josh on the podcast to host this special deep dive episode of the Salesforce Admins Podcast, and we couldn’t think of a better guest than Kat Holmes. At Salesforce, she’s in charge of User Experience. But she’s also the author of the amazing book, Mismatch: How Inclusion Shapes Design.

The title of the book comes from the World Health Organization. In 2011, they redefined disability as “a mismatched interaction between the features of a person’s body and the features of the environment in which they live.” As Kat explains, thinking of design as a way to solve mismatches leads to innovative solutions you wouldn’t otherwise find.

The problem with designing for the “average user”

For decades, designers have tried to make things for the “average user.” Kat takes us through the fascinating history of the bell curve, which goes back to a 19th-century Belgian astronomer who set out to apply the principles of statistics and probability to sociology. The problem, as she points out, is all of the different types of users that this approach leaves out.

Kat’s favorite example is the keyboard. It’s an interface that’s incredibly efficient and enables pretty much everything we do with computers. But it was actually invented to help a blind Italian countess write letters without the need to dictate everything. And there are tons of other examples, like bendy straws and curb cuts. These designs solved one person’s specific mismatch problem but ended up benefiting all sorts of other people, too.

Designing with inclusion and the potential of AI

When you’re building something, Kat recommends recognizing the abilities on your team and thinking about who might be excluded. As she puts it, “What abilities are missing that are important to the design we’re making?” Then, find a way to include someone with those different abilities in your process.

We also get into AI and what the future holds. As it becomes easier and easier for admins to build things, it’s more important than ever to factor in things like accessibility and inclusion into the equation. And there’s a lot of potential to adapt to the interface to the user to give each person a different experience.

There’s so much more in this deep dive episode, so be sure to take a listen for . Be sure to subscribe so you don’t miss out.

 

Podcast swag

Learn more

 

 

Admin Trailblazers Group

Social

Full Transcript

Mike Gerholdt:
This week on the Salesforce Admins podcast, well, it's our Deep Dive episode. I said we're launching something new for April, and with a deep dive comes a guest host. Hey Josh, how are you?

Josh:
Hi, Mike. I'm doing pretty good. How are you?

Mike Gerholdt:
I'm excited because I listened in on this episode and I can't wait to see if this is the pilot episode of where the Deep Dive series is going. Buckle up, folks because it's going to be awesome.

Josh:
Right? I honestly think maybe we should just, I don't know if we're going to do better than this. This was a... And I hate saying things like when people are like, "Oh, who was your favorite guest?" I'm like, "I don't like picking the favorite of my children." Kat's going to get into the top five right away. I never thought I would talk about diversity when it comes to everything from the iPhone to bendy straws. Just almost [inaudible 00:00:58].

Mike Gerholdt:
Yeah, it's fascinating. Let's get into the episode with Kat.

Josh:
Today on the Salesforce Admin podcast, we are going to talk to Kat Holmes about things, diversity, inclusivity, and AI. Kat, welcome to the show.

Kat Holmes:
Thanks for having me.

Josh:
So let's talk about your early years. In one of your talks, you speak about growing up in Oakland, and that led you thinking and eventually promoting inclusion. Can you expand on that a little bit? What about Oakland fermented this for you?

Kat Holmes:
Yeah, in the way back machine. So growing up in a city that os incredibly diverse, all the way through my schooling, all of my community engagements, we really learned a lot about many different ways that people live. But the thing that was really interesting for me, all the way through college, so I went to college in the Bay Area as well. I never learned about the fundamentals of accessibility as part of my training as an engineer. I also studied pre-med. We just didn't learn about ways that people experience disability in the world. So it's kind of ironic, you'd come up in this environment where you have all these kind of movements that had happened, right? Free Speech Movement, there was the Black Panthers, and at the same time, we never learned about the Disability Rights Movement, which also started in Berkeley in the 19... I'm going to say '50s and '60s by students, Ed Roberts, students that really, they created some of the first accessible sidewalks in the United States-

Josh:
Oh, wow.

Kat Holmes:
... right here in Berkeley. And I just never knew it even though I was going to school right there on campus.

Josh:
Gotcha. Now, you've talked about how when you were 16 you encountered racism, and I believe even neo-Nazism for the first time. And that left you, and I believe I'm quoting you here, "Activated and angry." And I have to say, as somebody who has used the written word to try to exact revenge on his enemies, I can appreciate it. But when you say active and angry, what actions were you taking? What led you closer to activism?

Kat Holmes:
The encounter I had, this was when I was a junior in high school. It was right off of the school campus, and I was physically and verbally assaulted by a group of neo-Nazis. I'm going for lunch. And it was a pretty shocking... I had also just moved from Oakland to a suburb, and this is where this encounter happened. So it was really shocking to my system. But the thing that really got to me is, that I told the administrators of the school, the principal, and their response was that there was nothing they could do about it because it was off of school grounds, so therefore it was perfectly legal, and that's the part that angered me the most.
Because that sense of responsibility, here's the adults in the environment that are, I thought, there to provide my safety. And what I was really hearing is I only do that within a certain boundary. And the way I got activated was writing. New student in the school, and took the time to write a intense feeling-filled, sixteen-year-old article that was published in the newspaper about my experience. And so when I think about, for me, and it means many different things for different folks, but for me it was about saying what was true and saying what my experience was and what was true about that. And so finding ways to activate people through our experiences, really, to share those experiences. And that's what I really have taken through my entire life.

Josh:
Did it feel like you were taking the power back?

Kat Holmes:
I felt like I could make myself visible is the way to say it.

Josh:
Got it.

Kat Holmes:
In the moment where I felt very much like people were trying to keep me invisible.

Josh:
Got it. Moving on a few years, what exactly did you study at UC Berkeley?

Kat Holmes:
I studied orthopedic biomechanics and material science engineering. So my goal was to design prosthetic limbs for people and tried to find a way to eke that out of a combination of majors.

Josh:
I got to ask, and I am going to throw in an anecdote here, because my father-in-law actually is blind and has no hands. So prosthetic limbs is something we... I think we have a few in the house here, actually. Why prosthetic limbs? Where were you going with that?

Kat Holmes:
I had been really interested in materials and mechanics for a lot of my young adult life. One of the things that struck me was prosthetics. We often try to replicate a human [inaudible 00:06:13] to try to make some material look like skin or some material shaped like bone or nail. And I thought there were so many other kinds of materials that were more expressive or unique that actually when you pair them up with somebody, you ask them what their preference is, they may choose a really amazing leather over a polymer. So quite honestly, it was just curiosity, following curiosity, connecting with people that I knew in my life who used prosthetics, but also just there had to be a better way to do this.

Josh:
Gotcha. Gotcha. Now at Microsoft, you were, I believe, if I'm correct, you were involved in designing their first-ever smartphone. Which I have to say, I think might've been my first-ever smartphone, might've been exactly that smartphone, and I remember it pretty clearly because it had this wonderful keyboard that was this very nice, tactile keyboard. And I know that a lot of people out there probably think this sounds weird because we live in... This is before the age of the iPhone where touch screen basically started ruling the world. What was that like at that time? Because smartphones were really just basically being invented. And so what kind of challenges were you facing when it came to designing a product for something that didn't exist before?

Kat Holmes:
Just to clarify, I did not work on Windows Mobile, and Windows Mobile was a really relatively successful platform for Microsoft. I came in right about the time that the iPhone came out, so 2007. And it was this existential moment for Microsoft because like you said, there's this physical world, BlackBerrys, and Nokia phones, and some of those great tactile keyboards that you're talking about. And then the emergence of the iPhone was the pinch and zoom on a map.
Being able to still take a phone call, even though you're taking photos, amazing. And the first phone that I worked on for Microsoft actually ended up being a spectacular failure, it was called Kin. I don't know if anybody knows it, but we had a blast building this phone, and it was about tactility. It was really a phone for teenagers, and it's because Facebook was one of the first apps on the iPhone. It was just emerging as well. And so we thought, wouldn't it be cool if you could create a full on social media app just for teenagers all built into the phone?
So learning a lot about that time, what I'll say is the top lesson for me is we poured money, our hearts and souls. We developed beautiful hardware with a company [inaudible 00:08:57] Sharp. But we missed what the success of the iPhone was going to be. And that was the developer ecosystem, the App Store. So you can build the best phone in the world, but the game had changed and we hadn't realized it. The game was all about activating a tremendous ecosystem of applications and developers that could build on this platform. And so we were still thinking of it as a device-centered world when really it was a platform game.

Josh:
Yeah. Well, and to your credit, I think Apple itself, because for the first year, I want to say of the iPhone, they're just like, "Oh, no, if you want to do anything custom to this, you have to do it through a website. We're not going to let you past our Ivory Palace into the App Store." And then somebody course corrected and here we are now in the middle of history.

Kat Holmes:
Well, that's where I did then transition into Windows Phone. And so I did help build that product and that platform. And that was a really fun experience, a really interesting experience. I think we pushed the boundaries and the design of user interfaces for mobile, and that did change the game for a lot of companies and how they thought about mobile design.

Josh:
Nice. Can you give me a couple of specifics there? What were some corners that you turned that you feel we might be still seeing today?

Kat Holmes:
If you remember the iPhone in 2007 when it came out, I think we used the term lickable for the advertisements, it looked like pieces of candy. They were shiny, they looked like they had [inaudible 00:10:36] in the phone. And it's those kind of, we use the term in design of affordances. The shape of the button says, "Push here," because it's so clearly indicating that it wants to be touched.
One of the first things that we did with Windows Phone was flat UI is what we called it. And we took all of those affordances out, but it's because we wanted the content itself to come through, people's photos. An application's top metrics, maybe it's biometrics from your health app. We want that content to come through on the icon or, we now think of them as widgets, but at the time it was very revolutionary to say, "What if the icon was the photo? What if the icon was the biometric data?" And so on a home screen for a user, they'd look at this unique, only looks like their phone, doesn't look like anybody else's, flat window into all of their content. And that was pretty revolutionary at the time.

Josh:
To actually surface that detail right up to the phone so that you can just glance at it and be like, "Oh, it's Tuesday."

Kat Holmes:
It's right there. And we still see that. I think the iPhone and its widgets in particular, but many developers have tried to bring, what's the most important thing a user wants to know both so they can glance and go, but also to maybe entice them to come into the app.

Josh:
Right. One of my favorite T-shirts is from Apple's WWDC where they announced the App Store, and they must have still, the icons are literally the location and date and time of the WWDC announcement.

Kat Holmes:
Oh, that's cool.

Josh:
Yeah, they lifted that for sure.

Kat Holmes:
I want a cool T-shirt like that. I have so many cool T-shirts from my 25, 30 years in tech. That's maybe the best part of working in tech is you get cool T-shirts.

Josh:
You get cool T-shirts. I have found that every now and then I have to double check myself and make sure I don't have more than three Salesforce logos at a time. And then I just feel like that guy at that concert. So, yeah. Speaking of Salesforce, how would you describe your current job?

Kat Holmes:
I am the Chief Design Officer and Executive VP for our user experience team. So I lead product experience, which means anything that at the end of the day ends up in front of an end user, whether it's through our amazing admin community, architects, developers, we're thinking about the platform that you use to build that, but also the end experience that people are going to interact with.

Josh:
Got it. Now in your book Mismatch: How Inclusion Shapes Design, you talk about design leaning to the average person. How are you defining a mismatch here, and what are some examples of design that intentionally are not being inclusive because they're designing for the average person?

Kat Holmes:
Yeah, the first thing I'll say is that in all my training as an engineer, in addition to not learning about accessibility, I also was taught the myth of there being an average person. So I'll get to that in a moment. When I think of... So the term mismatch, I borrowed from the World Health Organization's definition of disability, and they dramatically redefined it in 2010, they defined it as a mismatched interaction between the features of a person's body and the features of the environment in which they live. And I loved that as an engineer, as a designer, because it meant that it was my responsibility, in the choices that I make for the product, to make sure that I was considering different types of abilities that somebody might have when they come to use that product. The responsibility sits with me, as a product maker.

Josh:
Got it.

Kat Holmes:
And so some examples of mismatches might be stairs at the front of a library. It's a public library, but somebody who uses a wheelchair, who has limited mobility, would not be able to access that front entry. So another great example is the keyboard. This is a mismatch for anybody who has limited use of their hands, or doesn't have hands, completely unworkable for interacting with a computer. And what I love about these examples of mismatches, it means that we can identify who might be experiencing the greatest mismatch when they come to interact with our program or application. We need to make sure that it works for voice as well as keyboards, or it needs to be for different types of audio in addition to tactility. But, what I love about this also is that it's not about trying to create one solution for all people. You often hear the term universal design. That really means creating one environment that works for everybody. What I love about the keyboard is it was actually invented by a blind countess from Italy, and an inventor named Pellegrino Turri. And the two of them worked together to create a device that she could use to type letters on her own, rather than dictating to somebody else who'd write it for her. So they invented this device originally for someone who is blind, but it went on to benefit so many more people. We've used this device multiple times today, all of us. And in that, they've created an inclusive design. It started first with somebody who's highly excluded from some sort of activity. And that solution that they created benefited many more people. And so when I think about coming back to your point on the average person the misinformation that I was certainly taught in engineering is that there's a bell curve of human abilities, or any kind of human dimension.
And if you think about that bell curve, that the middle of that bell curve is the average human. This is a concept that was created by Adolphe Quetelet, he was a Belgian astronomer in the mid-1800s. And he was actually super jealous, is the way I read it. Super jealous of Isaac Newton, right? Isaac Newton had created these laws governing, deciphering what was happening in the heavens like why does the moon move this way? And Quetelet, who was also an astronomer, he had a pretty curious bombing of his observatory and could not practice astronomy during the Belgian Revolution.

Josh:
Oh my gosh.

Kat Holmes:
So he turned all of his ambitions to be as famous as Newton towards human society, and he started measuring human bodily dimensions. He created the body mass index that we still use today. It actually used to be called the Quetelet Index, to determine is a person healthy based on weight and height, which is a pretty crude measurement. He also developed the foundation of IQ tests, and he also developed really dangerous frameworks that underlie eugenics.
And the challenge with what Quetelet did is he gathered data for as many people as he could, but in the mid-1800s, really hard to believe that he had a true global sample of human [inaudible 00:18:36]. He had a nice Belgian, maybe a couple of countries over sample. So he took all of his data and he was astonished to find his data fit to a curve, a normal curve, which is in mathematics we know of normal curve is there's a point where the tangent reaches a perpendicular. So he was astonished that it fit this normal curve. And he took the middle of that line and he said, "Well, that curve right in the middle must be the perfect person." [inaudible 00:19:07] perfect person.

Josh:
Oh, God.

Kat Holmes:
And that became the foundation for saying any deviation from the center of that curve was some kind of abnormality or error. So taking mathematics and applying it to humans can be very powerful in some ways and can be very dangerous in others. But it's why we refer to people as normal, is actually from a mathematical background. And what I was taught as an engineer is if you design something for the average, you're going to hit 80% of the population. And then there's edge cases. I like to talk about edge cases.
There's 20%. That's an edge case. All you have to do is really look around at humanity, or do some research of your own, to know that that is just not true. That's not actually how the world is. But it's so deeply entrenched. It happens maybe at large sets of data, like large public health issues, and you find anomalies, and that's good indicators. But when it comes down to one person's experience sitting in front of whatever technology you're configuring or building or designing, it actually just isn't true. So that's where inclusive design becomes a much more interesting paradigm.

Josh:
It's fascinating to me that when we say the word average, and we apply that to a person, that we are probably describing a 20- to 30-something-year-old white male in Belgium.

Kat Holmes:
Yes.

Josh:
It's slightly terrifying, too to be kind of honest. And speaking this. And I honestly, I just want to bring this up because when I was reading about it, it shocked me that this even exists. You talk about Robert Moses, who apparently had, I'm actually struggling to say this to be honest, that he utilized a racist lens in some of his urban planning, which, I'm like, that's supervillain-level stuff right there. What's an example of this? I think a lot of what we're talking about is sort of designed through intention and it's good intention. We don't think about the average person being a 30-year-old white male in Belgium, so people don't intend to exclude people. But here we have an example of somebody who did. What's the story there?

Kat Holmes:
It's a really fascinating study, and you always have to remember the context of the time and place. But Robert Moses was the... the term they gave him was the master builder of New York City. He was a city planner, but he had wide-ranging control and power over the design of New York City. And the practices that he employed, and some of these are documented in a book called The Power Broker, is thinking about the types of transit that people had access to or didn't have access to. And so he'd say, "Hey, the tunnels leading out of Manhattan, heading out to the beaches," Long Beach, let's say. So the height of an average public bus, let's say is X, and the height of an average car is Y. So he would design the tunnels coming out of the city to be low enough that a public bus couldn't pass underneath it.
In effect, it created limited access to those public spaces outside of the city. But the inherent, nefarious part is, people who predominantly relied on public transportation, or exclusively relied on public transportation, tended to be Black or African-American families or families of low income. And so it's that it can happen intentionally, and it can happen unintentionally when you think about, oh, I have a car, so I'm going to just make this tunnel to fit my car. And not really think about somebody who maybe doesn't and somebody who maybe uses other modes of transportation that you're in fact creating this physical barrier in participating in public spaces outside of the city.
So that's a great example of sometimes it is nefarious, and sometimes it is accidental or unintentional. I think as people who are problem solvers, we come to this discipline or our jobs because we like solving interesting problems, or we think about how we can solve these and make the world a better place. And it's that kind of intentionality that fascinates me because when we bring attention to it, you can't unlearn it. [inaudible 00:23:48] oh, I didn't realize I created something that made it uncomfortable for somebody else. Just [inaudible 00:23:53]. How can I be a better problem solver?

Josh:
And to flip that script completely to the other end, give me a little bit of backstory. Once again, it was fascinating to learn, why do we have bendy straws?

Kat Holmes:
The story behind the bendy straw is super fascinating. The first design actually came from a man who was watching his four-year-old niece try to drink a milkshake at a counter. And this is the old soda fountain days, and they had straight paper straws in those days, and she kept tipping the glass and spilling the milkshake while she was trying to drink out of this straight straw. So he went home and he put a nail inside of one of these paper straws and he wrapped a wire around the outside and created a flexible joint in the straw and then ended up patenting it. And that's how we have bendy straws.

Josh:
That's awesome. That is awesome. Okay, so let's talk specifics about if I am a designer, how can I identify and address these kind of potential exclusions while I'm working?

Kat Holmes:
The best way to identify this is really first looking at our own abilities, like what abilities... Often the products that we make, there's teams that are working together. So looking across that group and saying what abilities are represented? And it might be, oh, okay, we all have 20/20 vision, we all are right-handed. We all speak a particular language. These are the abilities that we represent. Now, what abilities are missing that would be really important to the design that we're creating?
And that might be, okay, somebody who has low vision, or somebody who speaks a different language. And it doesn't mean you have to solve every scenario, every potential language, every potential ability. But what are you making? And who's going to need to use it? Are you designing something that's going to be in healthcare? Do you potentially need to think about somebody who is not well? Somebody who maybe has a different cognitive state, maybe they're in an emergency situation? If that's the case, then how can we think about including people in your team who have either experienced that or are experiencing that difference in ability and bring them in as experts to advise and learn from, or even co-design that product with you. So that's really the starting point is recognizing exclusion and then asking yourself who's missing? Really seeking out their expertise.

Josh:
And what's the importance in collaborating directly with people who either have experience or are possibly experts in different forms of disabilities?

Kat Holmes:
There's a couple of lenses I think are really important. One is, we often do research in design and we think of it more as user or usability research, or we're putting something in front of a person and asking, how do you think this works? Or does this work for you? We're treating people a little bit more like a subject, a research subject, which is different than starting before we've designed anything, and going to someone who has a different set of abilities than we do, and asking them, how would you solve this problem? Or have you already solved this problem in some way, in your home or in your work? And learning from the workarounds that people already have, or the considerations before you even create any solution is incredibly insightful to the process. And so it gives us a way of A, thinking differently about expertise. I'm not the expert as the designer. The expert is the person who's experienced exclusion, but still somehow is making a living using the product that I created.

Josh:
Got it.

Kat Holmes:
And then I think the other part's just, quite candidly ego, just to check my ego as a designer, that there's collaboration has a way of opening up the creative process. And I think that keeping our egos in check is a really important factor, and bringing other people to the process and letting them be the experts to lead the way is a really great way to do that.

Josh:
So to paraphrase, don't design a solution and then take it to somebody and be like, "How bad is this for you?" But bring them into the process so that by the time you get to the point where we're trying the solution, you've already brought their feedback in.

Kat Holmes:
Well said. Yeah.

Josh:
Thanks. Now let's move that kind of conversation to AI because that's how the world's revolving these days. So when we talk about AI in collaboration, how do you think people should think about AI itself?

Kat Holmes:
That's a ginormous question. There's two lenses I'll put on for this conversation. I think AI as a tool that can help us think about and the things that we're not recognizing ourselves. What are other considerations I'm not considering? How do I think more broadly than my own experience? I think AI is a great tool to help us expand the starting points. I do this often just with our own tools, with Einstein or some of the other tools in the world that are AI-related. But it's just, "Hey, I'm thinking about getting started on this. Where are different considerations that I might have?" So it could be a way of expanding beyond our own biases.
I think the other lens is thinking of AI as a user of what we're designing. So there's a whole bunch of behaviors, AI or different types of machine learning, different types of generative and predictive, even machine learning, are going to bring to our applications or businesses that we're building. So if we think of AI as a user that itself is trying to solve some set of problems. It's going to encounter certain kinds of errors, it's going to need to make certain kind of adjustments on the fly. The more we can understand what kind of goals and what kind of barriers AI is going to encounter when they work with the data that we are providing, or working with the applications we're providing, the more we're going to be able to design this positive cycle of access and also safe parameters around what AI can access, what it can and can't do. And so it might be a nuance, but thinking about AI as a tool, versus thinking about AI as a user, I think gives us really two interesting places to design from.

Josh:
Gotcha. Because I think one of the things, it's very hard, and this is one of the reasons in my own AI talks, I always tell people, just go try it because it's really hard to describe why it's a new style of interface, simply because it's conversational and it's interactive. What sort of design challenges come up with something that's having more of a conversation with you than just pressing a submit button?

Kat Holmes:
The interesting thing about AI is that we're kind of in love with this conversational moment of AI, ChatGPT welcomed us to a really broad and accessible kind of AI through conversation. But most of machine learning and AI applications that I've worked with, and I've worked with different types of interactions since about 2010, a lot of them aren't conversational.

Josh:
Got it.

Kat Holmes:
And even in our devices, our smartphones, we may have different types of machine learning or AI that is vision-based, object recognition, or audio-based or tactile. So there's many different kinds of interaction models that come along with processing information through AI. And the unique design challenges, I think one of the biggest ones comes back to the mismatches we were talking about earlier.
AI could give us a tool to be much more adaptive, to meet people where they are, whether that is, we were talking a lot about physical abilities earlier, whether the person can see or hear, but what about cognitive differences? And that's a whole frontier that I think is fascinating. There's so many different ways that people learn or process information or want information presented to them. Can AI help us adapt a design or an interface or an application to meet people where they are? If they're a novice versus an expert, wouldn't it be interesting to think about the differences in experience that AI could create to meet people where they are? So that's one design challenge.
And then another prominent one that there's many leaders in this field is thinking about the biases in AI itself. And there's a lot more, I think, visibility and awareness of this now than there was, say, five years ago, certainly 10 years ago. But the training sets of data, or when I go into Midjourney and I say, "Create an image of a doctor treating patients." [inaudible 00:33:58]. What's the doctor look like and what does the patient look like? And has this algorithm been trained predominantly on sets of data that favor certain races or for certain experiences, genders. So that kind of bias is a very small example, but a lot of companies have learned early lessons in this. I think Tay at Microsoft being trained overnight, within hours by the Twitter community, formerly known as the Twitter community. And it just went sideways within hours. And so that risk of what we're teaching and how that shapes the design at the end of the day is a huge challenge as well.

Josh:
I kind of feel like the world should actually kind of thank Tay for being such a horrible, awful example of how things can go wrong.

Kat Holmes:
That's true. It happened in a relatively safe sandbox.

Josh:
Right. No doubt here, it's basically speaking Hitler. We all can agree, let's not do that.

Kat Holmes:
[inaudible 00:35:11]. Thank you, Tay.

Josh:
Thank you, Tay. And I really appreciate it because I've talked to women of color who they're kind of in a generation where they grew up with the concept of what an engineer looks like, and it's that crew cut guy with glasses and a shirt and a pocket protector in an IBM [inaudible 00:35:31]. And they didn't think they would be an engineer because they never saw anybody who looked like them be an engineer. And I feel like we just have that history that AI has. I don't know how AI is even going to try to catch up to it.

Kat Holmes:
The opportunity is there. The opportunity to create a different reflection of reality is there. And it really comes down to the choices that we make in the design of our AI. And who is designing that AI at the end of the day. Can we really broaden... One of the things I love is I think the skillset to become an AI designer will dramatically change because the things that I learned in engineering school, I learned FORTRAN, so that's not super helpful anymore. But if we don't need to learn some of these technologies that are going to turn over anyways, what is the important thing to learn about the design of AI and then what skills are needed? And that could open up the field dramatically to a wider range of people.

Josh:
Yeah. And it's one of the things I'm really excited about with Salesforce because the idea that an admin could use their preexisting skills as a flow builder to then also be an AI builder is very exciting to me. Do you have any tips for some... I think our community's really in the shallow end of this. They're slowly getting into the waters of it. When it comes to thinking of solutions for their users, do you have any suggestions or tips for lining up what we can do with AI with a user's skills or job or role?

Kat Holmes:
Being in the shallow end is I think where everybody is. There's maybe a very small population that really, really is deep in these waters. Most of the population hasn't even put their toe in yet. So if you're in the shallow end, welcome.

Josh:
You're in good company.

Kat Holmes:
... [inaudible 00:37:29]. And please keep learning and keep walking a little bit further in because this is the first wave of us who, coming into those shallow waters, are going to say, "This is how we apply it to life." This is where it makes a difference. And I think our admin community understands the work that people are trying to get done on a daily basis. They understand the challenges people encounter. And when we designed Prompt Builder, for example, we were really thinking about the community that understands what an end user is trying to do. We're thinking about the admin community who can say, "These are the most important mundane tasks that need to be repeated and automated or supported by AI."
And so I think the most important advice is lean into that understanding who's using your products or who's using Salesforce at the end of the day. And help us understand what more will serve the people, and the use cases that they have, in better ways. And going back to inclusive design, think about folks beyond, think about the edge cases or think about the folks who maybe are experiencing challenges without using Salesforce today, and how can we really make this a turning point using AI tools to make sure that we're doing a better job going forward.

Josh:
Yeah. Okay. I'm going to throw a hypothetical to you and we're going to pretend you have infinite time and money. Where do you think... One of the things I think is very interesting is that the hardware curve, I feel is still advancing. We're just now getting things like AR goggles that are associated with AI. Where are some edge cases that you think could AI really help with inclusivity? For instance, I was having a conversation with a friend and I was like, "Well, I have a nephew who is autistic, and he might benefit from glasses that could actually identify social cues that maybe his brain isn't wired for." Where do you think we might be going with this?

Kat Holmes:
There's this interesting debate, I think, between computing power, infinite times and resources to make trillions on infinite computing power. Versus reaching as many people as possible with something that's beneficial.

Josh:
Got it.

Kat Holmes:
I would lean towards reaching as many people as possible with something beneficial. We may be in a place with what we have today to transform a lot of lives if we can really connect the potential of the technology to what people are trying to achieve. So with infinite time and money, I think there's tremendous diversity in human... This is such an obvious statement, but it's one that we haven't really taken to heart as technologists. There's infinite diversity in human lives. And understanding unique medical needs, diagnosing those, giving people the power to diagnose them for themselves, or to at least understand some of what's happening in their lives.
I think about medical, I think about cognitive learning styles, education around the world, just thinking about how I learned versus I have an 11-year-old, it's my youngest kid who's learning on YouTube, so fast, guitar virtuoso overnight. And I'm like, "Oh, how'd you do that?" Well, they've been watching YouTube videos and [inaudible 00:41:26]. So the learning, the medical applications, and then I think, one of the things I'm really interested in is how language models are going to become local to devices. How are we going to get really personal, device-driven AI that can be a close companion, or just the applications of being able to embed that in different environments? And that's where I think about climate science. And could we combine sensor technology with local AI device technology and think about climate science differently on a global pattern.
And so we put all our money into computing power for one great AI. Or do we think about the diversification of many different kinds? And I'd say the past 20 years has taught us that this tremendous power in diversification of applications, like we said in the beginning through the iPhone, that whole ecosystem, many, many small things can sometimes solve a problem equally or better than one ginormous thing. And that's how, I'd apply my money towards the small and the mighty.

Josh:
I love it. Kat, thank you so much for the great time and conversation. This was a lot of fun.

Kat Holmes:
Thank you. It was really good to dive into these topics. I appreciate it.

Josh:
Thank you very much.
I want to thank Kat for the great conversation and information. And as always. Thank you all for listening. Mike, how you think we did?

Mike Gerholdt:
I think it was amazing. I also got into some of the discussions that you were talking about, especially around architecture. I think a lot of times we, as admins, think of, "Oh, well, how does this apply to tech?" Well, how does it apply everywhere? We're design thinkers everywhere. And some of this is really opening up. I mean, you've exposed to me the whole making ChatGPT do illustrations, and now I'm asking it stuff. Like, that's fascinating. That's not what I was thinking in my head, but that's clearly what other people, or a machine, was thinking.

Josh:
Yeah. And I'm really glad that we got Kat to really describe how admins are going to really be in a driver's seat. They have a really important role based on what they're already doing. Based on the solutions that they're already building and their relationship with current users.

Mike Gerholdt:
Yep, absolutely. And of course, any of the resources that Kat or Josh mentioned we'll include in the show notes, which can be found on admin.salesforce.com, including a transcript of the entire show. And be sure to join our Trailblazer community because we'll post there to discuss about it.
So with that, we'll see you in the cloud.

 

Direct download: Unlocking_Diversity_in_Tech__a_Deep_Dive_with_Kat_Holmes__Josh_Birk.mp3
Category:general -- posted at: 1:00am PDT

Today on the Salesforce Admins Podcast, we talk to Skip Sauls, Senior Director of Product Management at Salesforce.

Join us as we chat about how Data Cloud can make it easier than ever to roll out enhancements to your org.

You should subscribe for the full episode, but here are a few takeaways from our conversation with Skip Sauls.

The challenges of working with external data sources

Pulling data from external sources is always a challenge. For one thing, it usually requires a bunch of work on the backend to get things looking the way you want them to. What’s more, it opens the door to all sorts of potential problems when things don’t match up, not to mention the extra security challenges.

That’s why I was excited to sit down with Skip Sauls. He’s the PM for Data Cloud, and he’s here to tell us how his team has made working with external data sources easier than ever before.

How Data Cloud simplifies data management

Data Cloud allows you to combine your external data sources with what’s in Salesforce without hacking together a series of customizations. Connectors allow you to import data from external sources as direct objects, or transform it into something more useful. You can run reports with it, use it in flows, embed it in Lightning pages, and much more, without needing to write specialized code.

Skip’s goal is to minimize the customizations you need to make and seamlessly combine your external data with what’s in Salesforce. Using Data Cloud means that you’ll be able to deploy enhancements to your org without worrying that everything’s going to break, or rebuilding it from the ground up. As Skip says, “we don’t want people to feel like they have to radically change everything in their day-to-day lives just to access something new.”

Get hands-on experience with Data Cloud

Looking forward, Skip and his team are trying to further simplify how Salesforce works with external data sources. They’re rolling out tools to minimize imports, so your data lives in one place but works the same as what you have in Salesforce. They’re also working on Remote Data Cloud, which will help you consolidate data that’s spread out across multiple orgs.

If you want to learn more about Data Cloud, I have good news for you. Skip and his team are releasing dozens of new hands-on challenges to Trailhead over the next few months. There’s never been a better time to get up to speed with everything that’s possible with Data Cloud.

 

Podcast swag

Learn more

 

Admin Trailblazers Group

Social

Full show transcript

 

Mike Gerholdt:
This week on the Admins Podcast, we're talking lakes, well, not lake lakes, but I mean we do talk about lake-making kits, and I do think that would be a hilarious swag at Dreamforce. But Skip Sauls is back because data lakes and Data Cloud are on our mind, and he's got a bunch of new features that he's talking about. Not to mention, he also tells us how we can get hands-on with Data Cloud, which I'm a fan of getting hands-on anything because that really helps me understand it. That's what we're going to talk about.
Before we get to the episode, just want to make sure you're following the podcast on iTunes, Stitcher, Spotify, iHeartRadio. We're on all the podcast platforms. You don't have to follow them everywhere. Just one, your favorite one, and then the new episodes download automatically right to your phone. Every Thursday morning, you can get up head to work or walk the dog or go for a run and get your new episode just by pressing play. With that, let's get to our conversation with Skip.
Skip, welcome back to the podcast.

Skip Sauls:
Thanks, Mike. I'm glad to be back, and as always, glad to talk to you and to the admin community.

Mike Gerholdt:
I was looking at my notes from the last time we talked, and the last real podcast that we did was about a year ago. To me, a lot of Data Cloud stuff was brand new, and also a lot of the concepts around data lakes and data silos was a new thing. It still may be new to many people in the audience, but I think we're starting to become even more exposed to it by just the sheer volume of amount of Data Cloud information that's coming out, and also the number of features that now Salesforce can support. Let's start with, what's some of the new stuff that you've been rolling out in Data Cloud since we last talked?

Skip Sauls:
One of the most exciting things for the admin community is how you can now leverage data from Data Cloud in your standard Salesforce org, in your lightning pages, in your reports, in your flows. That's been a big theme for the past year, which is, we've got this great technology for unifying the data, manipulating it, doing all kinds of great stuff to the data, but we now need to make it available to our customers, to our respective users. A lot of that focus is what I think is very exciting because now you can actually make use of it, and you're not trying to write specialized code or you're not trying to export things somewhere else. You can use it in the standard Salesforce fashions. It's inside of fields on a form, it's in inside of a record with related lists, it's in a report, so it's in all the places you would expect it to be.

Mike Gerholdt:
That's good. A lot of the times when we hear stuff like that, when we're not bringing the in, but we're surfacing it, I've heard the term like a pane of glass?

Skip Sauls:
Mm-hmm.

Mike Gerholdt:
Okay. Just want to make sure that-

Skip Sauls:
That's a great way to visualize it. The trick for a lot of our customers, as you know, is that you bring things in and enhance their working environments. You make them more productive, giving them better results, better KPIs, whatever that might be. We don't want people to feel like they have to radically change everything in their day-to-day lives just to access something new. Salesforce has done a pretty good job of that over the years, of bringing things in to lighten the experience, into mobile, making things that are in a low-code, no-code fashion, and really listening to what our customers want, which is, "Make my users more productive. Give them something useful here." They're always interested in technology, but really, the reason people want Salesforce is because it makes them more productive. It's a useful application architecture. That's what, to me, is very exciting. I look forward to Data Cloud just being assumed as being part of all Salesforce, not as being an add-on or something that's on the side, so to speak, that it's actually just it is Salesforce, for that matter.

Mike Gerholdt:
Yeah, well, it is.

Skip Sauls:
Exactly.

Mike Gerholdt:
One of the things that's interesting is I learned more about Data Cloud, I go back to... This was a requirement that I got way, way back when I was an admin, 2008. I remember an executive saying, "Well, this is nice and all, Mike, but how come I can't see X?" And I remember having to explain to them, "X data is on a different server that we have on location that Salesforce doesn't have access to." The fear, for me, was them wanting to integrate, essentially, a data silo into Salesforce. Because back then, data integrations were just crazy. But I think there's often, this is how I look at Data Cloud is, but how hard is it to really set up?

Skip Sauls:
That's one thing we've focused on is making it so that you can bring data from pretty much any source into Data Cloud. I'll tell you more about something else that's even more exciting down the road, but first off, talk about the notions of connectors. You have them in various flavors where you can connect to an external source and you can pull the data in. You can do it in batch, you can do it in streaming, it can be fully scheduled, and you bring that data in either as the direct objects from the remote source or you can transform them into something more useful. You may say, "I need to do something to IT to get it in the standardized formats." Things like dates and times, all sorts of things that you may want to... Salesforce admins are familiar with this. How do you make the data get into the shape you want?
Data Cloud has a lot of really great functionality for that. We've leveraged tooling from the likes of MuleSoft, Tableau, CRMA, plus the traditional Salesforce loaders and that sort of thing, and unified that in Data Cloud. We made that part of it as simple as possible, and we're adding more and more connectors to external sites. We'll have a very rich array. In theory, an admin can say, "I need to pull data in from..." Even something relatively obscure. There'll be a way to do that, and in the future, even custom connectors will be possible. You'll be able to do one that isn't sold by Salesforce or by a partner. That all is very exciting, and that fits into the traditional model where you're importing things in, but you're now doing it into one place i.e. in the data cloud, as opposed into multiple places, or directly into Salesforce itself, which is the part that's nerve wracking, as I think you were saying.
You don't want to necessarily modify all of your existing records, so with Data Cloud, you'd bring those external sources in. You can have as many as you want to. It's highly scalable to work with almost anything. Then you'll bring in your data from Salesforce, you'll have that mapped in effectively, and then you can have that unified into a single object. You look at it as being the same person, account, contact, et cetera, across all the different data sources. And you're not having to go and manually map everything in and do all sorts of things with unique IDs and keys and that sort of thing. It's doing a lot of the heavy lifting, and it fits very much into the standard Salesforce model of making those things easy. You're now dealing with it at an app level, not at a lower level, in most cases. You're not having to do that every day, trying to figure out how to get the data in.

Mike Gerholdt:
One of the things that came up in the discussion that we had of getting the data out of the silo... To be clear, it's not that we wanted it out, it was more of we just need to reference it. I think one thing you mentioned to me that was very intriguing is because back, this is '07, we were going to copy the data and then Salesforce could see it. But with Data Cloud, we actually don't need to worry necessarily about that, right? It's a feature we can set up where, if we want to keep that single source of truth and reference that glass pane, we can do that. Right?

Skip Sauls:
Exactly. The terminology you might hear is bring your own lake, bring your own code, and that sort of thing. There's a whole class of things there. I don't know how much I can say, because there's some pretty cool announcements coming around this, but there's a lot of work on making it so that you can leave the data in the external store. It could be a lake, lake house, data warehouse, traditional database, S3 buckets, wherever. You can leave a lot of those things in place, and you reference them from Data Cloud as opposed to importing them. This gets into the zero copy, no ELT mantra that you'll hear. The basic idea is that you're not having to actually make those copies, like you were saying. You're not going to move it back into it. It stays in place. It stays resident in the external system. But to your apps and to your users, it looks like it's natively in Data Cloud, and therefore natively in Salesforce.
That's pretty exciting. There's still the case where you might want to transform something to make it fit into the shape you want, but importantly, you don't have to do that to this data. You don't have to do it every time you want to try to use it. That's what we've seen traditionally with Salesforce is we were always importing, whether it was into Core or into Tableau or CRMA, etc. You are always importing the data, doing stuff to it, making copies of it, and that sort of thing. As powerful as those tools were, they still required that copy, which is the part you were saying we're trying to get away from.

Mike Gerholdt:
Yeah, copy, sync, then you had to figure out last modified, who wins, conflict resolution... That was a whole day of meetings for me.

Skip Sauls:
Exactly.

Mike Gerholdt:
It was not good. As you mentioned all this, I'm thinking just offhand, because this is how my brain works. A really fun swag item for Dreamforce would be a lake-making kit, like from the Progressive commercials. That would be hilarious.

Skip Sauls:
Yeah.

Mike Gerholdt:
I'm the only one that thinks that's funny, I think.

Skip Sauls:
Maybe you can do that for a Dreamforce or TDX next year.

Mike Gerholdt:
Yep. "Here you go, sign up to win a lake-making kit." One of the important things I think about not having to sync data and worry about that as a potential is also, depending on the organization and how they use the data, if they're in a contract and there's PII involved, then they can confidently say, "This is only stored here." As opposed to... I remember we had to go through that with a government contract and outline all of the places that this data could appear. When we were syncing data, then it became another page and a half of documentation of how people had access to Salesforce. I think that's really cool. You mentioned ease of use, and with ease of use, to me, that is also just how do we get the word out? How do we get people hands-on with that? What are some of the things that your team's working on around that?

Skip Sauls:
One of the things we're really excited about is providing hands-on challenges where you get to actually use Data Cloud directly. There's some technology behind it, but in effect, you're getting an org that is Data Cloud-ready, and then you can go do a trail, do a hands-on challenge at TDX or Dreamforce, maybe you're in a course somewhere. Using that org, and in the Trailhead model, you're running a check, have you done the work and that kind of thing. That all works perfectly. Now we've got that working very well. You can use this in the same way you would your standard Salesforce org. You get a DE org or something, and you start working against that. That's very exciting. And the cool part about that is that also will power all of these modules that come from Einstein, things like Prompt Builder and so forth.
Almost everything that you'll see for these new technologies is actually powered by Data Cloud. Even though you're not maybe explicitly using Data Cloud for the trail or the hands-on challenge, it's under the covers, Data Cloud being used for all the data, objects, services, and so forth. The reason that's exciting is it's harder than it may sound, because Data Cloud instances are not as lightweight and inexpensive as say, a Salesforce DE org. There's a cost associated to it. They're consumption-based. So we had to do a lot of work to figure out how to get that into a manageable state so we can offer that experience to our users and not be too much of a cost burden for us. There's still a cost there, but it's worth it for us to invest in our users, our customers, so they can get up to speed on Data Cloud, they're enabled on it, and they're also enabled on, again, I mentioned Einstein and that sort of thing. That's very exciting.
We were hearing from people, "I like Data Cloud, I want to learn more about it, but these trails don't let me use it." "I don't have Data Cloud. How can I learn more directly?" As I've heard you and others say, a lot of people can learn the theory from a standard trail or from docs and that kind of thing. Maybe they can pass a test, but in practice, it's that hands-on experience that really resonates. It's like, "I actually know what I'm doing here. I know how this behaves when I click on it. I know where to go." And that sort of thing. That's a really cool thing, which you're going to see a lot of.
Our plans are to have dozens of these out over the next few months, and we have a goal of getting tens of thousands of users enabled with these hands-on challenges. That tells you the scale we're talking about. I would encourage everybody who's listening, go try out the hands-on challenges that are on Trailhead right now. There's at least a few of them there for Data Cloud, some for Einstein, etc. You can get nice, shiny new badges and get your real world hands-on experience, and you'll see more and more of these in the coming months.

Mike Gerholdt:
It's one thing to do a module where it ends in quiz questions and then you read some stuff, and then it's another to do one and then get the error message be like, "Oh, really? I got to go back and read some of this. I really thought I knew what I was doing here."

Skip Sauls:
Exactly, exactly.

Mike Gerholdt:
One thing we didn't touch on, and this is maybe blinders that I have, but what about people with multiple orgs?

Skip Sauls:
A really cool feature, which is... I have to go look at my schedule and see when it's going to be out, but it's soon. Is the notion of a remote Data Cloud. I'm waving my hands here as I'm talking to you, but I'll try to describe it for the listeners. What you'll do in the future is when you have multiple orgs that you're managing, for example. A lot of people will have more than one. You'll designate one to be the home org. I don't know if that's the official terminology, but that's what we're calling it right now. That is where the Data Cloud instance will live. It'll be tied to that org. You won't have multiple Data Clouds, you'll have one, in most cases. Then the other orgs you have will be remote orgs that are leveraging that org.
There's some technology there. You can look up something called data spaces. You'll be able to say this part of this data in this data space can be mapped to these remote orgs, and in your remote org, you'll be able to access that as if it's natively inside of your org. In all cases, Data Cloud doesn't live inside of Salesforce Core, it's actually running externally. It's not as big of a hop, if you will, to have these remote orgs. It's not like they're really going around the globe trying to connect to each other. The home org is just where you're going to manage the core data, the way to do everything. But you could then have orgs that are primarily for sales or for service, or maybe you've got some that are by industry or by region. However you decide to organize yourself, no pun intended, you can still use the same Data Cloud instance.
The cool part there is because we're unifying all this data, you could have the same customer represented in multiple places across all these orgs, but they look like the same customer inside of Data Cloud. You can use this for how do you rationalize the data instead of trying to do it manually with all sorts of mappings and code and that sort of thing. You can say, "This is going to be Mike on all these different orgs." And also, importantly, it's Mike coming in from external data sources. It could be IoT, it could be social media, pretty much wherever you'd like to. But you can know this is Mike across all those, and it's a lot more straightforward than in the past, where we had to manually do a lot of work to say, "This is actually the same user across all these things."

Mike Gerholdt:
I like that. Yeah. Boy, 2007 Mike really needed Data Cloud, let me tell you. One of the things I was thinking about as you were talking through all this and unifying the data is really looking at Einstein and some of the stuff that's coming down now, and admins are seeing that. We saw it at TDX with Prompt Builder and Copilot. If you're a Salesforce admin and you're sitting there and you listen to this Data Cloud, what are some of the questions that we're hearing from customers that are really good questions to ask on what should I be looking for in an organization that should prompt me to start having these conversations about getting Data Cloud?

Skip Sauls:
There's a really good blog, and I'm going to try to find this for you. I'm going to tell you there's some great quotes in here if you're not familiar with SalesforceBlogger.com, that's actually run by some Salesforce employees. It's mostly employees posting it, but it's not our official blog. It's like some of the other semi-official blogs that has some really great content. In there, there's a whole section of what people will ask for. The reason I bring this up is a lot of times, you won't hear people saying, "We need Data Cloud." They're going to actually say, "We need to make better sense of our sales data." "My sales guy needs to be able to know which customers to target." In that example, you might have your current notion of your accounts and contacts and leads and that sort of thing. Then you've got some external data which talks about very similar things, but it's from a public source. It's not Salesforce data.
But it's information about accounts and it could be customer data, it could be company data and that kind of thing. But it tells you something interesting about them and what they're interested in, and you can actually import that data and unify it and then run some calculated insights and other about it. You might find out you're not really targeting the right people. You might say, "We actually need to branch out and target other customers." Or you're enriching the same data for your current customers, it's just data you didn't have before. It's like, "We didn't know this. We didn't know they were interested in these things, and we can have other selling opportunities." It's that kind of thing that I think is very important is that you're using it as a tool to make better sense of your data, make better sense of your respective target objects, whether they be customers or things, than you could before. You can do so in a way that doesn't require that you're manually trying to build all this inside of Salesforce Core.

Mike Gerholdt:
For the longest time, the joke was how long is your account page or your contact page, because you are having to reduplicate all these fields just to accommodate all of this extra data.

Skip Sauls:
Exactly. We see lots of interesting naming conventions for that kind of thing

Mike Gerholdt:
Probably horrible ones, too.

Skip Sauls:
Yeah, exactly.

Mike Gerholdt:
I'm guilty of that, too. Contact, last name, four, because that's how it's going to work.

Skip Sauls:
If you inherit orgs from others, sometimes that's multi-generational. You can certainly see that with like, "There's three or four different naming conventions and duplicates of objects." Because they oftentimes came in and said, "We can't really change this. We can't really make sense of it. We almost have to start over again in order to enhance something." The idea with Data Cloud is don't do that. Keep your existing data, move the source of truth into Data Cloud and operate it on it there, and you don't have to go back and rewrite everything in Core or importantly, everything off Core, every single time.

Mike Gerholdt:
You bring up a point, so let me... Silly question, because I'm still learning this too, but with Data Cloud... This is going to sound weird, correct me if I'm wrong, but you can have multiple sources of truths. We would have a finance system that was a source of truth for address, but we had a certification system that was a source of truth for what certifications that organization held. We didn't want them all in Salesforce. We wanted each... It's a data silo, but that's its job, and it's secure that way. With Data Cloud, we can connect them, we can get that view in Salesforce, but we also don't have to pull all that data in. Am I right in saying that?

Skip Sauls:
Exactly. You're mapping the data from Salesforce into Data Cloud, and if you have the same names and same values across different objects in different fields and external sources, you can resolve those inside of Data Cloud and say which one is the one you want to use, which one is that source of truth. You can create your ideal, I think people call it the golden record, is one notion I've heard of. This is the agreed upon... I heard it called the single version of the truth, which sounds political, but it's basically you as an organization say, "This is what we all agree is the correct source of truth for these things." Instead of it being in multiple orgs or across multiple objects, you now have the single unified object and you agree that this is the address, this is the account value, this is whatever the dates might be.
That's the beauty of it is it gives you one place to do that work, instead of trying to do it across things. It's always been possible to do this kind of thing. You didn't need Data Cloud to do that kind of thing, it's just harder to do those things. People found it frustrating, and the thing we didn't want to hear, what we heard people say, "I had to pretty much export everything outside of Salesforce and do work on it in some other cloud to get the results that I wanted." So we're saying, "Let's not let require people to do that. We don't want them to leave Salesforce. We want this inner gravity to still be on Salesforce. Let's give them the tools they need to be inside of our platform instead of externally."

Mike Gerholdt:
I remember having real conversations about how this X server could do a CSV and put it on... I think it was an Outlook or SharePoint, and then how do I set up, at the time, Data Loader via the CLI to do batch imports? That was a conversation that now feels so dated. Feels like watching a early '90s sitcom where they have a bag phone in a car.

Skip Sauls:
But people still do that today. We saw that with analytics. We still see it with people exporting, and they go into Excel, and they do their work there. We have had lots of great tools for this, and Data Cloud has the best suite of these things now, and you can actually do it really well in place. There's no reason for you to export anything, unless you want to make it available to somebody to play with in Excel, but there's no reason you should be doing your work there. Importantly, you've got tools like Tableau, which are really good at this, much better than Excel would ever be. Do your work in Data Cloud, use some of these great tools we have, and not do this external manual copying, uploading type of thing. That stuff works fine in the small, but it's terrible when you have large numbers of people working on it, and really bad when you have different people coming in at different times that may not realize what was happening.

Mike Gerholdt:
Yeah, let alone the second you pull something out of a system, now you've lost all control over that data.

Skip Sauls:
Exactly.

Mike Gerholdt:
In terms of security, confidentiality, especially if it's a spreadsheet, could be emailed to somebody. That's the part that always worried me. I always had the sales manager whose second question was, "Who can export this data?" Nobody, thankfully. That's a checkbox I never check. Skip, thanks for coming on. I know last time, we talked about the Evel Knievel motorcycle, but that was just because I was fresh off of going through a world tour DC and some museums out there.

Skip Sauls:
Oh, yeah. I'm a big motor sports fan, just like you. A gear head, whatever you want to call it. If it has a motor, I'm interested in it. Getting a little too old for some of it, but I still enjoy it.

Mike Gerholdt:
Well, that's the beauty is you can always watch it. There's always somebody younger than us that'll want to do something fun.

Skip Sauls:
Exactly.

Mike Gerholdt:
Thanks for catching us up as the new Trailhead... I'll link to the Trailhead modules that we've got available on that. Then of course, knowing that there's more coming out. To me, the most exciting part with everything is the second I can get my hands on a DE org or something, that's when I can actually start to understand it. I remember that was so fundamental when I first started as an admin, the ability to get my hands on a DE org and try stuff out that wasn't a production org. The same holds true for all of our products, so I'm glad that we've overcome that barrier.

Skip Sauls:
I encourage everybody to try that out and give us your feedback. What else do you want to know? What doesn't work well? What did you enjoy? Reach out to us. You guys will see me, the community, on Twitter, LinkedIn, etc. I'm always looking for more feedback, and ping me if you need anything. Let me know how we can help.

Mike Gerholdt:
Yeah. I appreciate it, Skip. Thanks for coming back.

Skip Sauls:
Yep, thanks a lot.

Mike Gerholdt:
I'm glad we could have Skip back. Always appreciate him coming back and helping admins understand how we can break down all of the data silos that we have within an organization and make our lives easier. I wasn't kidding when I asked a few of the questions about syncing data and back and forth. I've got to believe that's some of your life, too, because I feel like everybody just one view of the customer. But everybody's got to own different parts of data, and that's fine. This really helps knock things out. I think it really makes things interesting and accessible for Salesforce admins.
Now, if you're listening, I want you to do me a favor. Click on the Share Episode button, and you can post it to any of your social media. You can text it to a friend, maybe there's a friend. You guys can both do a Data Cloud Trailhead module together, and let him know that you got hands-on with Data Cloud, which was something that the last time we were on the Salesforce podcast that Skip was on, he couldn't tell us to do.
Thanks for listening, and until next week, we'll see you in the cloud.



Direct download: Data_Cloud_Enhancements_that_Admins_Will_Love.mp3
Category:general -- posted at: 1:00am PDT

Today on the Salesforce Admins Podcast, we talk to Lizz Hellinga, Consultant and Salesforce MVP.

Join us as we chat about why product management principles Salesforce are crucial if you want to take advantage of new AI tools.

You should subscribe for the full episode, but here are a few takeaways from our conversation with Lizz Hellinga.

AI enhancements and what they mean for admins

The last time I had Lizz on the pod, we talked about why clean data is crucial for AI tools. But with everything that Einstein Copilot and Prompt Builder make possible, I wanted to bring her back to help us understand how to approach AI enhancements.

The big thing to get your head around is that these tools make it easier than ever to implement changes to your org. However, as Lizz points out, that means it’s even more important to think through how Salesforce fits with your business processes. How you gather requirements and communicate with your stakeholders is more important than ever before.

Apply project management principles to your Salesforce org

To get the most out of everything that’s possible with AI enhancements, Salesforce Admins need to brush up on product management. “It’s kind of like the operations around your operations of Salesforce,” Lizz says. She wants everyone to think through three questions:

  1. How are you taking in change requests?

  2. How are you working with your stakeholders to determine if those requests are aligned?

  3. And, finally, how do you go through the process of enabling that change and then extending it for adoption?

As Lizz points out, what you need to do hasn’t changed. You might be able to do things faster with AI tools, but big-picture thinking is even more essential so you can deliver the right solutions at the right time.

Communication with stakeholders is a two-way street

So how do you get started? For one thing, you need to figure who you’re trying to talk to. As Lizz puts it, “it’s never too late to run a report and do a stakeholder analysis.” You can look at profiles or roles to determine who the main people are in your organization and what they need from Salesforce.

You need to build trust with your stakeholders, and that means establishing two-way communication about requests and what you’re working on. Lizz recommends creating a transparent system for tracking requests, whether that’s using case objects or custom objects in Salesforce, or even (gasp!) creating a shared spreadsheet.

It can often feel like there’s a lot of heat on you to get everything done as quickly as possible, but that’s why bringing stakeholders into the conversation around enhancements is so important. If people understand why bumping something up on the roadmap will push other changes back, it can really turn the temperature down. It’s all about creating a feedback loop that turns stakeholders into collaborators.

Be sure to listen to the full episode for more from Lizz, and don’t forget to subscribe for more from the Salesforce Admins Podcast.

 

Podcast swag

Learn more

 

Admin Trailblazers Group

Social

 

Full show transcript

Mike:
So we're talking about product management this week on the Salesforce Admins Podcast with returning guest Liz Hellinga. You may remember she was on in December and really focused us on getting data cleaned up to get ready for AI. Well, now it's, how do we manage Salesforce as a product manager and also take into consideration all of the things that we've got going on with AI? And really all of the tools that AI can provide us, like Einstein Copilot, Prompt Builder. What can we do?
Before we get into that episode, I just want to point out if you're getting ready for all of the content that we've got lined up this month... So last week we had Tom Leddy on the podcast, check out that episode. Next week is Skip Solves. We're going to talk about Data Cloud updates. Skip was on last year. And then at the end of the month on the 25th, we're launching a new style of episode. It's going to happen at the end of every month, and it's... We called it Deep Dive and it's with a fellow evangelist, Josh Burke. He's going to deep dive into a topic a little bit more than I do. We're going to kick it off with our Katie Holmes, who is on our design team and talk about design and AI. It's going to be a really cool conversation.
But for now, let's talk about product management and AI and helping Salesforce admins be good stewards of the platform. So let's get Liz on the podcast.
So Liz, welcome back to the podcast.

Liz:
Thank you, Mike. I'm happy to be here.

Mike:
Last time you were here, we talked about how clean data is non-negotiable in the era of AI, and I still think it is. So let's pick up our discussion from there. What have you learned about cleaning data in the last four months?

Liz:
That it's still essential and it's ongoing. And that I really do love a good data dictionary that helps you define data and make sure that you're using it correctly aligned to your processes. But in this age of AI, it's even more crucial as we talked about before, because people are going to be making decisions on that. And we're all able to make more decisions based on AI, whether it's around our data or whether it's around how we build in our Salesforce org.

Mike:
Yeah, I think that's the thing that's changed since I started doing podcasts around AI last year is, last year we really focused a lot of the episodes on, well, how does this affect data? What can it do? Now... And I ran into you at Trailblazer DX. Now we've seen things around Einstein Copilot, Einstein Prompt Builder, which yes can do things around the data. But also a lot of the promise that we're seeing with Copilot is this will be a tool to help admins not only generate information or help users generate information, but also potentially configure organizations as well.

Liz:
Yes. And it increases the rate that we're able to make change because-

Mike:
Very true.

Liz:
... you're shortening some of that cycle to produce for those outputs. You still... Just like data, you still need to have some of these core foundations in place to make sure that you're making the best decisions for your Salesforce instance based on the output. But the scale of change is going to continue to increase, and it's going to be back to the basics for some of those things like around product management, the Salesforce, product management.

Mike:
Yeah. Well, and that's really one of the key things that admins work on because I remember way back in the days... I want to just outdate myself. I believe it was Shell Black, for those of you that are as old as me. Remember, he coined a phrase, "Order taker admin." And I think that's kind of relevant to what we're talking about because we've always talked about, "Wow, with every new innovation that Salesforce comes out with, it's easier and easier to make change."
Now, we're also have the ability to ask AI to start making change for us or to show us various flows, right?

Liz:
Right.

Mike:
And that affects our ability to manage the product because essentially the way that we're perceived potentially from our users is, "Well, it's just a field. Why can't you add it?" Or, "It's just a flow, why can't you add it?"

Liz:
Yes. And they're not always familiar with all the behind the scenes stuff that it takes. But ultimately, with your Salesforce instance, you always need to be enabling change based on what you're gathering from your end users and your stakeholders and aligning it to business objectives. And so that still hasn't changed. You may be able to do it a little bit faster with the help of some of these things like Copilot. But you still need to understand and have a lens for that decision making. Because just because you can add a flow or just because you can do something, doesn't mean that you always should. You still need to think through it from a process lens, from a stakeholder lens.

Mike:
So maybe we jumped in too quick, but to level set, I'd love to hear from you. What is your definition of product management that a Salesforce admin would do for the platform?

Liz:
Yes. Gosh, I have such a strong opinion on this, Mike.

Mike:
I know that's why I'm having to move on.

Liz:
Well, with... Product management is really sort of like the operations that you wrap around your Salesforce instance. How are you taking in change requests? How are you working with your stakeholders to determine if those requests are aligned? And then how do you go through that process of enabling that change and then extending it for adoption? So it's kind of like the operations around your operations of Salesforce a little bit. But it's just a way to bring structure and I would say consistency and continuity around how you iterate on your Salesforce org.

Mike:
So product management is not just a ticketing system and doing what the users ask.

Liz:
Correct. There needs to be thought around it. And I've been in orgs, right, where... We've been in those shoes where there's pressure to do something and you do it because you're just like, "Well, I don't know. I don't have enough to stop this." Or, "We're under a time crunch. We've got to get it done." But that's why it's not easy. But admins have to spend time building relationships with their stakeholders and being thoughtful about how they take in change requests.
And it could be something as simple as even just... Not that I want people to use spreadsheets and things like that, but sometimes just at least capturing that information even in a typical requirements document or building that out within Salesforce. I've seen a lot of people, and I've done this myself too, use the case object to help manage that and then review it with key stakeholders, determine what changes do we need to implement and how? What is the best method for it? 'Cause that's the beauty of Salesforce, there is usually more than one way to do something.

Mike:
Yeah, no and beauty and part where you have to really contemplate. What I'm hearing, and I'd love to know your thought on this 'cause this is something that even I struggled with as an admin. I think everybody does. The short-term change versus the long-term. And I mean that in respect of... I almost think of you know when you get a scratch on your car? Well, the long-term change is, I'm going to get it in the body shop. But the short-term is I'm just going to shoot it with a can of spray paint really quick to prevent it from rusting out. How do you balance that?
I mean, what can we do to think through product management to say like, "Okay, cool. I totally hear you need this and it's on my six month roadmap." As opposed to maybe I just invest time now and build a little bit of it to turn the heat down knowing we're going to invest in it later. What's the balance there?

Liz:
Yeah, the balance comes from understanding your stakeholders and the processes that you're using, that you're building out in your Salesforce system. So for example, if you know on the roadmap that something's coming in six months, but there's heat to get that taken care of sooner. I mean, being able to have conversations around, "Well, this is what's on the roadmap. If we pull this forward, something else is going to have to be pushed. How might we make that decision so that we can meet this business need sooner rather than later?" And it's not always an easy conversation for admins to have because it takes relationship building initially and trust building.

Mike:
What are some of the most important things that you feel should be communicated to stakeholders in order to keep that constant level of trust high?

Liz:
Yep. Great question. So there's a couple things that you can do. So initially... And it's never too late to do this, so you could be in an org for three days, or you could be in an org for three years. It's never too late to just start to do a stakeholder analysis. You can just run a report, group people by either their profile, their roles and determine who are the main people. You may be in an org where you can't talk to all 1000 users, but you may be able to get to a subset for relationship building.
And that is crucial, especially with the pace of change that we're encountering now because of generative AI. And then I tried to get a system down for gathering enhancements. Some of those enhancements may honestly never get built, but at least you're documenting them and those as request and keeping a... Excuse me, keeping a log or a history of that. Sometimes you just build it in the case object. Some people do custom objects. You can do integrations for those or a spreadsheet depending on how big your org is. Just giving a place for that and then creating some transparency around that list is helpful.
And then including those stakeholders in discussions around how do you prioritize those things? And I would start small, especially if your org is larger. If you're dealing with a lot of stakeholders, you want to start small maybe with one group. But as you can expand that, then I would probably start to do... And this is what I do in one of my orgs currently is, I do a bi-weekly update. And we don't work on a regular sprint cycle per se. We're not as hardcore with the Agile methodology, but we share every two weeks what we've accomplished, if it has a significant impact on the end users.
And then we also share in that notification projects that are in flight and their status. So if we're working on maybe implementing something from the app exchange that maybe take us a month or two to implement, we include that and we provide status. So it's creating that visibility because sometimes people... You'll be surprised that people will respond and say, "Hey, I'm interested in this," or "I have an idea, or I have a thought on this." It creates that two-way communication that is required for admins and their end users.

Mike:
Yeah, I think a lot of... And I experienced this too. A lot of the requests, I was able to dial the heat down and dial the request down when I started sharing very transparently the roadmap on what was coming and features that were coming. Because, to be frank, a ton of users, over 500, and they don't know. And when an absence of knowledge happens, they feel, well, maybe nobody's thinking of this, so I better raise my hand and put in a request when in fact it is on the roadmap.

Liz:
And then sometimes too, just getting that visibility into the roadmap, end users will kind of do a groundswell like, "Hey, we actually need this sooner." And it helps when you've got a list of 10 people, individuals, that are asking for something and you're like, you can then go to leadership and say, "Hey, this is becoming a real need. How can we prioritize this? What can we put further down the backlog so that we can push this forward and get this really great feature out that could help make the team more productive?"
So creating that path for feedback is essential. And I know sometimes there's this... People think, "Oh, we're going to just get inundated with complaints or things like that." But I'm like, "Bring it," 'cause I want to know. I'd rather have people log a case with me around something that they is annoying them so that I can analyze it and determine, can I fix this? Is this a part of something else that we're working on so that we can keep iterating?

Mike:
Yeah, I agree. I would rather them be publicly vocal than privately angry.

Liz:
Yes. Plus, when things come in the written format, it allows me to use the written format back to them versus sometimes when you're maybe on a group call, a meeting and it's hard to be eloquent. So I will say one other area that the LLMs have helped me significantly is crafting more clear and concise messages back to my end users and stakeholders.

Mike:
Yeah, I was just going to ask you about that. Because I think that's one of the things where for a long time, creating training and stuff, people can look at, "Well, I'm just not a good writer," or "I'm just not a good communicator." And I was going to ask you what specifically maybe some of the tips that you have for admins to get out there and experiment with AI and absent of some of the products that Salesforce has, because AI seems to be everywhere now. I feel like pretty soon my toothbrush is going to have AI, it's going to start talking to me while I'm brushing my teeth.

Liz:
Well, hopefully it'll tell you if it has a cavity.

Mike:
Yeah, I don't... But do you want to know that? You got a cavity here. I just might throw you away now.

Liz:
I know, right?

Mike:
No, you're lying at me. But I feel like that could be one avenue that could help admins both understand how to write good prompts and understand AI while benefiting us back.

Liz:
So for example, I can be quite verbose and long-winded, and so I will sometimes ask something like ChatGPT or Gemini to make... I'll just draft something. This is the one thing. It's like you can draft something and your tone could be maybe terse or it could just be long-winded or filled with jargon. And I can pop it into ChatGPT and sometimes I'll use it like explain this like you're explaining it to an eighteen-year-old. Oddly, the eighteen-year-old or sixteen-year-old sort of prompt kind of helps me because it takes out some of that technical jargon, but also softens the tone a little bit for me. That's been quite helpful. I have ideas in how I want to communicate. ChatGPT helps me kind of put a little bit of structure around it so that it's not so all over the place.

Mike:
Yeah. I also really like it. I'll ask it for different tones.

Liz:
Yes.

Mike:
That was always a... A friend of mine told me this trick, which now it feels like trying to teach somebody how to use a rotary phone. But if you need to write a difficult email, pay attention to your tongue because you're probably nervous and your tongue's on the roof of your mouth, so you need to lower that. That'll also help lower your stress, but also pay attention to your eyebrows. And it was always referred to me as eyebrows up, because it's really hard to write something angry if you have your eyebrows up.

Liz:
Oh, I never knew that.

Mike:
Yeah, I mean, you can of course, but if you're trying to not write something terse per se, eyebrows up because it kind of pulls your whole face into that happy smiley. And it's very non-verbal and it's telling your brain, "No, we're happy. Let's write this in a non-confrontational way." But the AI can do that.

Liz:
Where was that advice pre-ChatGPT, Mike-

Mike:
Sorry.

Liz:
... when my brow is furrowed and I'm thinking I've reset your password 10 times in the last week.

Mike:
Yeah. Well, there are some things you can write with your eyebrows up that still come off terse, but that was always the trick that I was told.

Liz:
Well, it helps too, 'cause sometimes you can play around with formatting using ChatGPT. It's like how would you format this for a Slack message versus an email? And it is helpful. It even adds an emoji sometimes, which to me, it could be a little bit of overkill. And sometimes when I ask it to write in a friendly tone, it's a little too much. So I like balance between professional and friendly.
And then obviously, I'm going to still make changes to it, but it just gets me closer to... It saves me a significant amount of time. It gets me closer to communicating effectively, and it allows me to continue the positive relationships that I do care about and that I want with my stakeholders. But sometimes in the moment, emotions can get the best of you.

Mike:
Well, and you bring up a good point, and I can always edit it. And I think that's very relevant to a discussion we had last week with Tom Leddy on decision-making and throwing things at Copilot and Prompt Builder and then just taking what it gives you. One of the ways you can always think about that as a product manager is, "Okay, so is this thinking of ideas that maybe I didn't come up with?" And I think for me, a lot of times Copilot if it can build me a flow that I didn't think about or in a way I didn't think about, that gives me another option as a product manager for how best to manage all of the platform.

Liz:
Yes. Well, and if it's giving you an option that you didn't think about, you still need to spend time thinking about it before you select that as your option, right, to solve that problem or to solve that request. What's nice about is it shortens your learning time. You're not having to build and fail, build and fail as much, but you can't take the human assessment out of it.

Mike:
Right, right. There's no one right way to product manage. And I say that and then oof, somebody's going to be like, "No, there is," because the internet. But I think from your perspective, having worked with a lot of admins and seen orgs and seeing various different ways of product managing, rather than asking you, "What's your preface for product management?" What are attributes that admins should think about when creating a system like a ticketing system, regardless of what it looks like, that would be a good attribute to help them manage Salesforce as a product in their organization?

Liz:
Yeah. So I mean, you think fields that you would ask?

Mike:
Happy with fields or outcomes or... I've definitely... We've discussed using, you said, service cloud and cases. That's one way.

Liz:
I mean, I love a good Kanban view, right?

Mike:
Sure. I mean, we can have that.

Liz:
That is one of my favorites, and it's a great screenshot for a slide if you have to.

Mike:
Not that you've done that before.

Liz:
But for me, it's the level of detail that you're willing to ask your end users to put in. I mean, I would say for about every couple of tickets, most of the time I still need to have an additional conversation to understand. And this is purely around enhancements. I mean, definitely if there's fixes that need to happen, there's conversations automatically. But I love that it creates this opportunity for me to reach out to people and continue to build that rapport. So it's never just like, "Oh, I get this request in from a case for an enhancement." And, "Okay, I am just going to do this." It's never that. I always want to understand a little bit more context.
So I try not to create too many field requests on that enhancement request, but I do want to understand what process is this supporting? What is this hindering you from doing? How will this help you? Things like that. I try to get a little bit of what I would say, just contextual information if I can. But sometimes people just put stuff in. It's like, "I want this field so that I can do this reporting." And that's their enhancement request.

Mike:
Yes. I need six check boxes, one for every color of the product that we sell.

Liz:
Yes. Or they want five multi-selects.

Mike:
Oh.

Liz:
But it's also too... The one other thing that I like about creating the space for people to submit enhancements, it allows me to create groupings. And that's one of the key things that I look for when I'm building the system is, can I group... Well, for lack of a better term, tickets or requests together to form a mini project? Because then you're really thinking through the process and how it relates to Salesforce beyond just a field.
And so typically when things are going on at the company and people are all of a sudden you're getting all these kind of related asks, that means there's like, is there an initiative that I haven't been informed upon? Do I need to know more than I know? Because maybe I can build a better solution if I get more looped into these internal initiatives or objectives.

Mike:
Yeah, I really like that idea. I mean, I think that's something that I would love to see in a future roadmap for some of the Einstein products is helping if you set up a ticketing system, using cases, helping you kind of clump, for lack of a better term, those cases together into these kind of... It's almost like, do you play connections in the New York Times?

Liz:
I do.

Mike:
Okay. So don't get started on that. But it would be like, what are these all have to do together with each other kind of a thing?

Liz:
Oh, that would be awesome.

Mike:
I know, right?

Liz:
It'll be here before-

Mike:
And then purple would be like-

Liz:
...we know it.

Mike:
... here's the craziest cases that we've found a through thread. I can't figure out what the through thread is.

Liz:
That would be incredible. I'm sure that will be here before we know it. I

Mike:
I mean, next year maybe we'll be on the podcast together. I'll be like, "Liz, hey, remember we had that podcast we talked about? Now it works."

Liz:
Yes, that would be a dream because I think I do spend a bit of time grouping things together and trying to figure out like, "Okay, is this related to something? And maybe I need to look more." But that's the beauty of it 'cause sometimes you get the cases in and then you're like, "Oh my gosh, we may have something that we may need to fix or we may need to re-engineer," and that's okay. You want that. And again, the more you can get that and iterate on it, the more your stakeholders and your end users are going to trust what you're building and doing in the system.

Mike:
So one thing I want to touch on, and unfortunately we're doing it at the end, but a lot of this we talked about product management in terms of what users want and what's being requested of you, absent of the release schedule that Salesforce has. What is your philosophy or how do you think about adding that as... Because that's an additional layer that we have to consider, right? Three times a year there's a brand new release, there's new features, and they may or may not be on any roadmap.

Liz:
Well, if you're in your org and you know what's going on with it, you're usually waiting for those features. I usually get pretty excited about some of the things that come out. So usually I feel that I'm always eager. And I love all the stuff that the, admins blog and the Salesforce plus sessions that you all do for release readiness. That's so incredibly helpful because I sometimes listen to and have an 'aha' moment like, "Oh wait, we were just talking about this and this could help me solve it."
And it's just being aware. I think having that rigor around your enhancement list and then reviewing it on a regular basis helps you when you're starting to review the release notes or attend release readiness live, you'll get those light bulb moments. And you'll just be like, "Okay, well this is coming, but it's going to have to now wait until the summer release." Or should I say fall? Just joking.

Mike:
Yeah, right. That's always a quiz question to get wrong is what's the season that's not a release?

Liz:
Right.

Mike:
Or I guess I always think of it, here's a feature that's coming that I don't have to build.

Liz:
Yes, yes. It saves time.

Mike:
Which used to be the case a lot. So thanks for coming on and refreshing us on product management because I feel this is a core responsibility that admins focus on, and especially given the speed of innovation now that we're thrust into with AI.

Liz:
Yes. And it's even more important for them to flex these types of skills around product management.

Mike:
Oh, absolutely. I mean, product management skills as an admin will translate to other skills across the organization and within technology.

Liz:
Yes, exactly. And it's a great way for them to build their career and trust.

Mike:
Absolutely. Thanks for being on.

Liz:
Thanks so much. It's always a pleasure to be here.

Mike:
Of course, that was a great discussion with Liz. I love having her back. I'll also link to that previous episode down in the show notes so you can check it out because clean data with AI is super important. But love some of the points that she had to bring up, especially around thinking through different features and really managing all of the requests and roadmap. I hope that's something that you're thinking about, too.
Now, if you enjoyed this episode... I hope you did. I had a lot of fun recording it. I would love for you to share the episode. And if you're listening on iTunes or really any platform, it's usually super easy. You hit like an up arrow or in iTunes, you hit three dots and you can click share episode, and that'll allow you to post it to social or text to a friend. Maybe you got a friend that's getting started as a Salesforce admin and they want to learn how to manage the product.

And of course, as always, I appreciate you listening. So until next week, we'll see you in the cloud.

 




Today on the Salesforce Admins Podcast, we talk to Tom Leddy, the Product Director of Well-Architected and Decision Guides at Salesforce.

Join us as we chat about decision-making in the age of AI and why cleaning up your data is more important than ever.

You should subscribe for the full episode, but here are a few takeaways from our conversation with Tom Leddy.

Decision Guides and the Well-Architected Framework

Almost exactly a year ago, we had Tom on the pod to talk about the Well-Architected Framework. I’ll link the episode below but Tom gives us a quick summary: “It tells you how to build healthy solutions with Salesforce and what a healthy solution should look like,” he says.

Making your org healthy comes down to looking for patterns and anti-patterns. Essentially, you want to do things in a way that sets you up for long-term success.

Tom and his team are hard at work rolling Decision Guides into the Well-Architected Framework. These walkthroughs are designed to help you decide which Salesforce tool is right for you when they have overlapping functionality. The answer is going to depend on your specific use case, so looking at a Decision Guide can help you understand the full picture and make the best choice for your business.

Understanding AI as a tool

Looking forward, Tom sees a lot of potential in combining AI tools like Einstein Copilot with the information in Well-Architected and Decision Guides. There’s a lot of potential to make things more interactive or quicker to digest, but you’ll still need to do some critical thinking and make your own decisions. 

In terms of incorporating AI tools into your org, Tom is working on decision guides for that, too. “A lot of the cool AI features are not going to work very well unless you have a good underlying data strategy,” he says. Working through the Well-Architected Framework will help you create a solid foundation to get the most out of these new tools now and in the future. 

Why AI needs clean data

If you’re a frequent listener, you’ll know that we can’t have an episode about AI without mentioning just how important it is to have clean data. As Tom points out, this extends to patterns and anti-patterns as well. It’ll be easier than ever to roll out code to your org and create new customizations, but you need to be sure you’re doing it the right way and not crippling yourself with technical debt.

Luckily, Tom and his team are working on tools to help you make sure your org is, well, Well-Architected. Stay on the lookout for a Data Strategy Decision Guide, coming soon™, and new ways to assess the health of your org with Einstein Copilot. The future is bright, and hopefully a little more organized.

Be sure to listen to the full episode for more tips from Tom, and don’t forget to subscribe for more from the Salesforce Admins Podcast.

Podcast swag

Learn more

 

Admin Trailblazers Group

Social



Direct download: Can_AI_Enhance_Salesforce_Architecture_and_Decision-Making_.mp3
Category:general -- posted at: 1:00am PDT

1