Sean-McLeary-Audio-edit === sean: [00:00:00] If I'm signing something that says, I'm gonna buy a house for a couple million bucks and I'm gonna be on the hook for this mortgage, wanna make sure that they're not stealing my information, that this is actually legal and enforceable over time. Like, how do I. Feel that, and , we're all comfortable with that now, much like we're all comfortable with e-commerce now, but back then it wasn't the household name. I remember one day I was watching a sitcom and I remember they said, oh, I'm gonna just DocuSign in the sitcom. And I was like, oh, we're here arrived. Welcome to Launch Pod, where we sit down with top product and digital leaders. Today we're talking with Sean McCleary, VP of Product experience at Intap in this episode, we'll explore how Sean brought together macy's dot com and Macy's brick and mortar into one cohesive experience, helping to increase revenue across both sources How the team at DocuSign improved the UX going from Tech Darling to mainstream product used from everyone across Sandhill Road to Main Street, and finally, we'll discuss the current landscape of AI adoption addressing empty box syndrome, and look at the parallels that early stage internet had to how AI is being adopted now, [00:01:00] and how we can use that as a predicator to how it's gonna move forward. So here's our conversation with Sean McCleary. jeff: Sean. How's it going, man? Good to have you on. sean: Good to be here. Thanks for having me. jeff: I'm excited. We, uh, We got pushed out once or twice. you had a little motorcycle accident. How you feeling? I. sean: I'm recovering. yeah, 16 years of riding, never crashed. And then I've got a nice titanium plea in my arm now jeff: You are a bionic man now. . I'm glad it wasn't worse, but I'm glad you're here now. You seem good. But you are a digital PM living in a physical world. Which that point actually . I I'm seeing it in here. Over your shoulder. You got like the little pink notes. Is that anything? Notable? sean: yeah, it is. This is me thinking physically, right? I've been out of the office working from home for five years now, and I, and I miss the physical nature of an office space. The company that I was at before pandemic happened and we all went remote. I was going in the office two, three days a week and it was absolutely perfect 'cause there was enough time at home to think. But then there was time to do these kinds of interactions where you're exploring things in a physical space, and I just don't think digital tools have fully bridged that gap yet. jeff: [00:02:00] Yeah, no, I feel you. I definitely, I still, if I wanna remember something, I write it down. If I type it, I, it may as well be, I just read it or saw it, but if I write it down it's there pretty hard. sean: Yeah, and there's some science behind that too that, that physical nature is much more memorable jeff: Exactly. it's funny you no one can see it, but all here on the wall in front in our studio is quotes from people. I reference a lot in the show and back episodes and people think I just remember the show a lot really well and somewhat I do, but it's mostly, it's all just cheat sheets, the entire wall. It's fantastic. sean: Yeah. jeff: But cool. Back to, we talked, we said digital, but. I think this kind of, level of product planning and kinda spanning the digital and the physical really defines a lot that we're gonna talk about. 'cause like you started out, you had spells early on at macy's dot com back when people were still uncertain about e-commerce kind of, and how do you bring the brick and mortar and the.com together? Spent time at DocuSign and Zenefit as a leader in the UX side. Actually across all those, you started your product career. More on the experience side and then, Indio trust layer. Now [00:03:00] Intap more on the peroxide, but it's all come together back to experience. I think we've got some cool stuff we're gonna talk about around. Macy's and how those kind of two experiences came together. Building trust on a digital world and how do you communicate expectations and some cool stuff. Let's just dive right in here. You started out you were running experience and user experience over at Macy's, but it was a siloed world. It's back, some of us may remember the days of e-commerce being new, but paint a picture. You walked in these two realms were siloed and you're on. sean: Yeah, I started with them as a consultant. And so one of the early projects we did with them as a consultant was very much about bridging the physical and the digital. You sort of alluded to it a little bit back then. E-commerce was still very much, we were not fully comfortable with it as consumers and very much businesses were still figuring out how to have identities. Macy's, I think was still. At that point, really treated the.com as a separate entity entirely, almost a separate company. They had their own warehouses and all that, so the fulfillment was different. From the stores. And yet the customers didn't [00:04:00] look at the brand that way, right? The customers looked at Macy's as Macy's, you go to Target now and you go to online Target. That is a familiar, consistent experience across those, and I can do things where I can look at the assortment in the store from online or on my whatever. It's very holistic. jeff: I use it all the time. It's fantastic. sean: Back then that wasn't the case. And yet the expectation was there because you're, it's, it says Macy's, why isn't it holistic to me? So the project we brought into is to start to bridge that gap and bring some of the macy's dot com functionality into the point of sale for the associates. So if you think about, there's Macy's and the relationship with the customer is not always customer to technology or just customer to physical store. It's also customer to the associate, the person. Who works for Macy's in the store, right? And so they brought us in and said, look, we wanna put macy's dot com on the point of sale, which essentially is a little mini computer, right? It's, it's, it scans the UPC and [00:05:00] get collector cash or your credit card. But it also is, it's a computer, right? You can do all sorts of interesting things in it. And so they really literally wanted to just plaster macy's dot com onto this little. Little tiny low resolution screen with this keyboard and this teeny tiny little track pad and make that work. And you could look at that experience and know right off the bat that just wasn't gonna be effective. There's jeff: I think they still have to this day, the little green screens and stuff. As the POS maybe, how does that even work? How would you have a modern website, even by those standards of timing on, the little green retro screen? sean: Yeah, they've gotten away, for the most part from the green screens. But the interaction, the keyboard is still the same. Like you walk into a store, the screens are, I don't know, 12 inches or something. They're not very big. They're probably 16 bit like they're pretty low resolution. They have some color to it, but they don't have, they're not like this giant monitor that I'm looking at right now. And more importantly the experience that the associates have is not a sit down experience. They're serving customers. They've got a line of customers waiting [00:06:00] to check out, and if somebody walks up to them and says, look, I want to buy this in a different color that you don't have on site. Before this project, they couldn't really do anything about that, right? They couldn't really handle what beyond what they had in their assortment. So that was the project is how do we bridge that gap? We looked at just putting the.com on it, which had all this stuff meant for a end user shopping. At their desk on it. And so all this unnecessary stuff, this marketing things, and it was gonna be really hard to use and distracting for an associate to use in the store because that was the user for this, right? So what we did is we said, look, that's not gonna work. Let's go take a step back and let's go look at this problem space and knowing what you want to tackle here. And so we did a lot of user research in the stores with the associates to understand how a furniture or a home goods associate is different from a fine clothing associate or what have you, and how they acted, and how they thought about the [00:07:00] shopping experience working with the customer. How they relate it to the.com. And that was probably the most interesting finding is they looked@the.com as a competitor. Associates are incentivized on the sales that they're getting and people are coming in with printouts of, look, I saw this online. And they didn't know what to do with it. They couldn't do anything with that. And so they were losing sales essentially to.com and that's the way they perceived it. jeff: When in reality the company, if they can't sell it right there is potentially losing a sale where you have someone ready to buy and give you cash, close the deal sean: exactly. And not every Macy's is equivalent. So you have the flagship stores like in New York or Macy in San Francisco or Miami. And then you have everything like a D level store way out in some small suburb, which doesn't have nearly the number of. Clothes or the types of clothes that you have at some of those other places. So somebody walks in with a thing of, like a Gucci shirt that they have on online and a D level store's not gonna have Gucci and so they can't sell that [00:08:00] stuff until it became this competitive relationship with things. jeff: So , you guys did user research on the actual associates. Was it like, were you in store working shifts or how'd that work? sean: Not working shifts, but watching them work shifts and then just sitting with them and just talking through their experiences and learning, being a sponge about what's it like to be an associate and not necessarily directly related to this project. 'cause you wanna understand what's the associate's experience holistically, and then how does this intersect with that. So that was just wildly in interesting and important and some of it was about the technology. they were these big keyboards. They had these T keys, which sort of specialized function keys across the top. And the whole interface that they had T key driven kinds of experiences feels a little bit like the old green screen. So you had to know the key to press to do the thing right. jeff: it's like a hotkey almost or something. sean: Exactly right. And then a little tiny track bag, which they didn't really use 'cause it was teeny tiny and they didn't the experience wasn't really cursor based. It was mainly keyboard based. So that was [00:09:00] just a whole bunch of really juicy findings. And so taking that away and coming back to the problem space of how do we bridge this disconnect between the.com and the store experience for the associates. We experimented with a bunch of things, but it became clear pretty quickly on that we needed to use what they had or what they were familiar with as an anchor point. So we actually took the capabilities of the.com and redesigned it around the experience of the point of sale that they already knew, using the T keys, using the language so that it actually, as you move between some of the existing functionality on the proto cell into the.com experiences. It was seamless. You didn't know that you were leaving for one to the other. jeff: Now, is that just because right, with any kind of upgrade or change like that? There's change management and training and stuff? Was it just. There's too much training to do and work to do to get people to wrap their heads around a new UI like that? Or was there another business reason or kinda what was the reasoning behind let's wrap this modern experience and kinda the thing they're [00:10:00] used to. sean: Yeah, the trading certainly was a key component and change management absolutely was a key component. 'cause these, talking thousands of associates across, the whole country really or beyond that. But also it was just, it was from an efficiency standpoint. If I'm using these kinds of interactions on these kinds of screens and I expect things to be in certain places, why would I make it different? On a, on the next screen, why not use that experience language, that interaction language consistently. , and you also remember Macy's staffing is not consistent across the year. There's seasonal associates coming in during the holidays, things like that, going in and out. And so the faster we get people ramped up in, the more efficient we can through that experience, the better everybody feels about jeff: it's also like people aren't just sitting there at a register and a cash register job. They're usually around like organizing their section. They're doing several jobs at once, and so I. There's already a tension going away and switching cost of your attention to throw in, that kind of new experience is probably hard. I love that you guys though, you got in and you understood, like we [00:11:00] had deep, deep men on from Hunger rush recently and stuck about the same thing. They literally called it sometimes you have to just make the pizza. And they were doing, I think she worked doubles at a pizza restaurant at the drive through and making pizzas to understand not just. Oh, it's hard, but oh, when you open the drive through window, you hit your elbow a lot. When you're trying to make this change on the screen, how do we, you're answering a phone call and you're taking an order. How do you make it a one touch to edit an order when someone's got their kids at the window and wants, I actually want extra cheese, not pepperoni. But there's a lot of learnings you can't have without direct observation like that. sean: Yeah, and I think that level of empathy and understanding does not come from sitting in a room thinking really hard thoughts or even talking to industry experts. You have to get out in the field and see it in action. We actually just had one of those the other day where we were designing For lawyers and ex and anchoring off of this, understanding that when they close a matter a case essentially, that's a good point, to do the next step of things, and that makes total sense if you look at it on paper. But getting outta the field, you realize [00:12:00] quickly that they don't close the matter. They keep it open for all sorts of useful reasons. So the matter, even though it's done, is not closed in the system. And so you can't that, and you wouldn't learn that unless you get out and talk to people. jeff: It's a lot of times, right? Good to great is often in that detail of, it just works the way I work already. It fits right into my flow. I don't have to change to operate it, it just, it added value. But I didn't have to think hard about how to get that. sean: Yeah, you got it. jeff: Awesome. So you brought in now you've got macys.com integrated in a natural way for their workflow. What did that enable? What was the kind of outcome there , did it work well or how did, how'd this kind of work out from a result standpoint? sean: it was really dramatic. So it was easy to uptake, which was great. 'cause that was, that was part of it. But it actually changed it to a point where the macy's dot com was not seen as a competitor, but an asset to the associates because they could sell things that they didn't have in store. Get credit for it, essentially. So [00:13:00] somebody could walk in with a printout or they could find a piece of clothing on the rack that wasn't in their size or color. And the associate essentially had an extension of their store right at their fingertips that they could access. So it. It shot up immediately. We did it in a few test stores to see what would happen. It did very well. And then it got rolled out all over Macy's and then Bloomingdale's, because Bloomingdale's as part of Macy's, Inc as well. So we blew it up and it became this huge revenue source for the company to just bridge that gap. jeff: It's kind of like a win-win, right? Because for the associates, they're not, you're not changing their ui, you're not changing their work habits. But now it used to be, oh, we don't have that size. Sorry, nothing I can do. Yeah, sorry, lose the sale. And now it's. Let me order that for you. Let me get it. We'll ship it to you. And they get quota relief and or they get a commission. The end user, the customer gets the thing they wanted to buy and they're not going somewhere else. Macy's gets to make the sale and not have a customer go to another store. sean: Yeah, absolutely. [00:14:00] So I think when we de we develop products, we have to think about the system both within and beyond our products, right? And where, how does this intersect into the larger workflow and where is there opportunities to bridge gaps and to bring people into that workflow. jeff: It seems like from talking to you and a lot of others. Rarely is there like one major stakeholder you need to talk to. There's often the buyer, the user stakeholders and the best solutions come when you understand how that flow sits in the entire world, right? Looking at a company like Hunger Rush, which I mentioned with Deep De, they sell to, to franchise owners who ultimately have probably multiple stores, but the beneficiary is. In a lot of ways, the end users, those people who they, but they're not gonna use it if it sucks. Honestly, if it's hard to use, it makes their life harder. It's just gonna make everything else harder for the, the franchisee, the owner. But in here it's, if it didn't really work that well and made associates' lives harder, maybe they just wouldn't do it, or they wouldn't, you wouldn't get advantage of that. Or they'd take longer to check people out and actually lose sales. But in this case, everyone benefited. And that's, often the biggest wins. But you can only get [00:15:00] there when you really understand those fundamental things, which I think it's interesting 'cause this. Brings up our next topic and you were at DocuSign and this has always been an interesting problem to me. 'cause wet signing is something, I think there's a lot of people who've been around long enough to remember you used to have to wet sign documents and sometimes you get the email, print it, sign it. They take a picture of it or go to a scanner somehow re-mail it back. Yeah. But DocuSign came along and, the founder there really pushed the legislation to allow it. But even with that, that didn't miraculously turn the tide of the world because I. There are so many questions involved in this, right? Like, how do you get people using, like we can use it. We were, we work in tech, it was not just us it's people. It's you know, construction workers, it's my grandma. It's all those kind of people still have to sign contracts. And you came in to DocuSign a very important time when a lot of these questions being pondered, how do you make it accessible to people who aren't? Tech savvy. How do you think about, the signer isn't the customer, but they have to have a profile at some level. Like how do you manage all these things? So I guess you walked in, it's a couple [00:16:00] years in a DocuSign. Let's jump in there if that's cool. sean: Absolutely. So DocuSign had, Tom Ganzer, who was the founder, had seen this vision and had built this a technology and a, and almost a mindset and a thinking around how do you take what has essentially been a physical process, not just digitize it, but understand the fundamentals that make it authentic, make it. The transaction legal if you will, and enforceable, right? And so the signature is an interesting aspect of it, right? The putting your glyph on a piece of paper, GLY itself. And I know back in the day we used to sign our checks and the bank would go, oh, does that match? They'd have a signature jeff: Oh, it's wild. sean: right? That's because. Important part it is the, you're agreeing to this thing and the glyph is that, putting your stamp on the piece of paper for that. And in some cultures it literally is a stamp. It's like a chop and you put it jeff: Oh yeah. sean: Yeah. So [00:17:00] yeah, it was jeff: We have I worked in Congress for a little bit and there was the auto, signer thingy and my grandma had a stamp with her signature that she always used, and there was a million ways to do this. And it always just seemed, even back in the day, I thought, this seems pretty easy to fake. In theory, if I had ever done something wrong in school, I could have probably just signed my dad's name on the letter. Not that I would ever do that, but sean: No, course not. But, it is a very interesting time. So they had taken the idea and built out the technology. They had gotten the business running. This was a good 10 years worth of work, essentially. And then I came in to at a time when we were about 200 people, so relatively small. It was a business. They had raised a fair amount of money. So they were stable. They weren't trying to make a business. But it wasn't the rocket ship that happened over the next four years. Part of that was we needed to make that sign both the sending experience, so the person who is, sending out these agreements to people for them to have to sign and to countersign. But also for the signer [00:18:00] who may not have ever heard of DocuSign, to get this email that says, Hey, click here and put your glyph somewhere, and to feel comfortable with that. If I'm signing something that says, I'm gonna buy a house for a couple million bucks and I'm gonna be on the hook for this mortgage, right? I wanna make sure that they're not stealing my information, that this is actually legal and enforceable over time. Like, how do I. Feel that, and again, we're talking about eras because we're all comfortable with that now, much like we're all comfortable with e-commerce now, but back then it wasn't the household name. I remember one day I was watching a sitcom and I remember they said, oh, I'm gonna just DocuSign in the sitcom. And I was like, oh, we're here arrived. jeff: We're in the future now. sean: This is amazing. But that journey to get there involved a lot of. Doing user research with all sorts of people to understand how they felt about the experience. 'cause it was an emotional response. There was certainly a functional usability quotient to it as well. Like how do I [00:19:00] know what to do when, where to click and how to think. But there was also this emotional component of do I feel comfortable with this? So we did a lot of spending time on research with that and how we could make people. Feel comfortable with it, feel like this was not trying to steal my information, but actually trying to help facilitate this agreement and that I think really helped us launch into the next era of what became this big thing of DocuSign. jeff: So how, I guess let's talk through how you got there. 'cause I think from a sender standpoint, that's probably the easier of the two problems. 'cause you can, these are customers of DocuSign. You can go work with them, you can interview them. It's probably tougher. And right back to Macy's, they were associates. They were employees of Macy's or as Deep de talked about from Hunger Rush. They were employees of the customer. They were employees of the franchise. You could go and observe these people and, she could do a shift at the restaurant. sean: Yeah. jeff: These are, the signers are not, there's no relationship whatsoever to DocuSign. You can't just go [00:20:00] out and walk into these people's houses. How did you get this done? How'd you go and observe or how do you understand and gain em empathy for that group where you don't have that group direct relationship? sean: Yeah. In, in some respects, it's actually easier because you're talking about literally everybody, right? You're saying, my mother is gonna sign something. At some point, the CEO of a company is gonna sign something. Everybody's gonna sign something. In fact, that was one of the way we got some product led growth was a CEO would buy a house and sign the agreement on the side of a mountain while he is hiking. Why don't I run my business this way? So it was fascinating but it was actually a much easier problem to solve because you could just grab people off the street jeff: Yeah, that's fair. sean: yeah. So it was actually a 'cause I we're, I'm now in a highly regulated industry right now with lawyers and , investment banking and stuff like that and getting ac lawyers are especially, are really hard to get access to 'cause they bill on six minute increments. jeff: Right. And they're famously slow to pick up tech, if I remember from people I know who [00:21:00] worked in the legal industry. sean: Yeah. Yeah. Law firms and Indi and Insurance, which both of which I've worked in, are very jeff: Right. For good reasons. There's a lot of security needs and stuff, like I understand why, but. sean: And happen. So I think it's much harder for me to get access to my end users now than it was at DocuSign. jeff: Interesting. That's true. 'cause you can just go out and almost, it's better the less someone knows and you can figure out how to make them comfortable. Probably the better off the entire experience is. 'cause at some level, if you were to go to signers of your customers, there's probably some inherent bias where they're working with someone who's already a little bit more tech advanced. So there's some chance of proclivity towards a little bit more advanced people. You're just literally going and pulling, my grandma off the street and some random guy and , a couple walking down the street. If you can get these random samples of humans on board and. Comfortable than the people who are have a higher chance. That earlier on being exposed to tools like this are gonna be fine. sean: We did broad research, like we didn't, it wasn't like we sat around and talked to five people and we got it. Cool. We're. jeff: Just went on the street, pulled five people. You wanna make a buck? Do you like [00:22:00] this? All right, we're done. There we go. sean: That said, there was just a lot of commonality. The more we talked to people, the more we saw the same themes over and over again. And so it really reinforced that we were on the right track there. The sender side, which you alluded to was an interesting one because we, the product that the team had really built was really, and it was one of the reasons DocuSign one in those early days was really feature rich. There was a lot of things that other e-sign companies had not even thought about yet because we're what we, we talk about. E-signature. We're really not talking about just the act of signing. We're talking about a workflow that goes into basically agreement management is really what it was. And so how do I manage all of those steps along the ways and the varying needs of a sole real estate agent who's just sending out, whatever mortgage agreements or whatever. I signing off on delo disclosures and stuff like that to Proctor and Gamble. Literally that was our spectrum of users and customers. And so that was the interesting part is how do you [00:23:00] accommodate that really broad spectrum of just size and then industry. And then eventually we got into geos and the way Japanese and the Europeans work is a little different. And so there was a lot more complexity on the sender side than there was on the signer side, but they were our customers. Whereas the signers, you are sending them something they've never heard of us, or the senders sending 'em something they've never heard of, DocuSign, they have, five minutes to go sign this thing. And we want that experience to feel intuitive, understandable, comfortable, trustworthy, all of that in that very fine moment where you may on the sender side have a little more opportunity to train them and things like that. jeff: So I wanna get into the sender side 'cause you brought up all that and that apparently I was way off base on the sender side is actually really hard. But also I, I have to think, SOEL, is there a human training here between the layers of, almost a prompt of. When you are sending a contract, part of the flow should be, at least early on, tell your signer what you're [00:24:00] doing. You will get, a DocuSign. Here's what it is. Yes, it is legal. Was there any thought or look at, how can we get ahead of this from a human process side of, can we try and even some number of people to just be better about setting expectations. I feel like oftentimes. Expectation setting is the thing that gets missed and that can solve so many problems that are hard to solve if you just go at it straight from UX or display or that. But if you just someone knows, I'm gonna get an email, it's gonna be this thing that I signed digitally. Yes, it is legal. I've had a minute to sit with it. I'm gonna just take that whole thing a lot better than what the hell is this? What am I doing? sean: , it's a really good point and it, and there's definitely things that we as DocuSign could do and did do to help facilitate that and make both sides of that equation. The sender and the signer feel comfortable with what was going on and what they needed to do. However, we have to remember that this is not a new relationship, these senders. Are doing this already. They have a relationship with the customer. And so actually oftentimes it was us getting out of the way and giving them room to talk to the sent [00:25:00] signer. And it was things like make sure the co-branding iss really clearly, so it's not coming from DocuSign, it's coming from x, y, Z company. The picture of the sender in the email the ability to have not a bunch of texts about what was gonna happen. But the message from the sender in the email there, so the, when that signer got that, they were like, oh, this is Joe. Joe sent me this because we have this relationship going on, and here he's telling me all that. There's his picture, there's his logo. That helps bridge that gap. 'cause it's not our relationship, DocuSign's relationship is not with the signer, the sender's relationship is with the signer. jeff: Not to go off topic again, but this is something that really gets me because you were thinking about this and you guys at DocuSign were addressing this years and years ago. This is still something that I have contracts that I have signed or tried to sign in the past few months where. I can't find the contract because they're sending it to me and it's in no way is it searchable from who it was sent by or what it's [00:26:00] about or anything like that. Or I'm searching for a contract, maybe we execute it. I can't find it again because I have to remember that we signed using this program and then sort through all the things I've gotten from that program and just hopefully remember about the time span. It was mid-April of 24. Whereas if I could just look up the company name or the person's name, I'd be able to find it so much faster. But there's just no. They don't convey that piece, that the relationship is between the sender and the signer and not, the e-signature company is just getting in the way of the relationship and obfuscating the actual intent here. sean: Yeah, it's interesting. We, I think there's a lot there. To think about when you have this relationship where your, you as a company, your relationship is with this customer, but then the customer has a third party relationship with some other person, and you are facilitating that. So you are involved but the signer in this case doesn't know you. DocuSign doesn't have a relationship with you. There is a really interesting opportunity there and you can expand it to all sorts of stuff. We can talk about some of the. With [00:27:00] Indio and trust layer. 'cause there's some similar relationships there between the people who pay me the company that's building the software, or the company or the end user, whoever that they're interacting with. jeff: , let's actually dive into this because I think this is a really interesting topic with. We've had a bunch of people on recently who work in hospitality or in, other kinds of, contracting or insurance. But there are all these industries where you have relationships with maybe a buyer, but then there's a secondary relationship that goes beyond your buyer to an end user and. AI is , a, people are already trying to like, bring in preferences and how do you kinda manage and customize, but AI is just adding this whole secondary layer of what you can do with that data. So everyone's like redoubling and tripling down on either buying third party data or getting their own proprietary. But some level who owns that? How do you make sure you as a customer are, the end user are getting the benefit of all that. And I wanna go. Maybe from Travelocity to Kayak and I want my preferences to come with me. And I think this is, [00:28:00] if I'm understanding some of the things you thought about at some of these companies. So maybe give us your picture of this world first and I think I'm under understanding and underselling it a bit. sean: I think you're on the right track, but there is this sort of opportunity for that entity at the end of the workflow chain or the one that's being brought into something to do something for them. So let's tackle trust layer really quickly because it's a little bit of a different animal. So trust layer is a risk management platform. It's a lot about how do I as a, an a, a company who has a relationship either with another company or an individual. Assure that they're meeting the ends of our agreement, specifically around proofs of insurance and licensing and things like that, right? That's a very manual process today. It's a very, it's a process that oftentimes you have to constantly check in on, or if you don't, you're putting yourself at risk, right? So there, in that circumstance, there's the company that's seeking that proof. That's our customer. And then there's the company that's providing that proof. So a good example would be [00:29:00] I am a general contractor building a building, and I've got all sorts of subs. I've got electrical subs, and I've got plumbing subs, and all of them, they need to have insurance that ensures them and the work that they do so that I'm not covering it all right? If they electrician burns down the building by accident, that electrician's gonna cover it, and I'm not on the hook for it. Right? So the whole thing with trust, there was like one thing. One is collecting that proof of insurance, whether there's certificates of insurance, or basically say, look, I've got insurance. It's at the right limit that we agreed upon in the contract yada, yada. So that's part one is collecting all of that information. It can be really complex and it has to have the right names on it and all this kind of stuff. So that's part one. And so it's very similar to the signing experience where the company that is the gc, the general contractor, is sending this over to the plumbing subcontractor and saying, give me this stuff. And so they have never heard of trust layer. They log into this portal and have all these things that they need to go do. Great. And then they've got that certificate of insurance. The second [00:30:00] piece of it is that certificate. We're getting a little bit off topic here, but it's interesting this, that certificate of insurance is dead as soon as it's printed. It's a piece of paper. It doesn't prove that in this moment you have that insurance. jeff: Right. It's when it was printed, it. sean: exactly right. And when it, hopefully, that's the other piece of it is like sometimes people, they fraudulently write up these COIs, right? So you can't really prove that thing. It's a half proof. And so we were developing a digital proof of insurance, which allowed us to touch into the carriers, the insurance providers, and even the brokers, and be able to. Real time authenticate that policy was still live, that it had the right limits, all that kind of stuff. But getting back to our point topic, right? The, that, that subcontractor may also have employees or other subcontractors. So there's a, their, and their proof of insurance is something they're gonna carry around to other jobs. So there's a real opportunity for them to own that proof. And then a, for us as a product-led growth kind of thing, for [00:31:00] us to use them as a, as an upsell to, Hey, look, you've just did it for that, that gc, how about doing it for your business? Similar to DocuSign, right? Again, you have your information, potentially your identity, 'cause you have to prove your identity. And then, um, could I carry that identity around? Could I carry those documents around as I exist in this eco? jeff: Versus having to print it. You have one GC you work with who uses that and you just put stuff there. You have another GC who you have to still using, old school COIs that apparently may or may not be as a company who ran a company that collected COIs for a long time. This is terrifying to me. I just realized, but like, how do you validate that? And so you could also go the other way, right? Have these subs who, if they have a profile that they can share no, all my proof's right here. Here you go. It's digital, it's validated. sean: Exactly, and that was what we were playing with. We were talking to, VCs who were managing job sites and we were talking about QR codes where you could basically look, look, you don't get on the job site until you scan this and show that you have your for insurance. So I think there's a real lot of opportunity [00:32:00] there, both from a business standpoint to be able to sell with these people and use them as opportunities to. You know, I, I'm, I'm the COO assigned on the side of the mountain to buy my house, and now I can sell them, but also to, to create network effects, to create repositories of information. Your question about who owns that information, it gets really interesting and challenging. There's, I think there's a real underserved opportunity there in this secondary user or secondary entered entity relationships. jeff: And that starts to get into. The ma world of we, how do you carry that around and what does that, what does a world look like where you have all your data bundled about you or me or or Emma, a producer who's not on screen, but and what does that world look and how do you leverage that into a world with more ai? And how do you how do you own it versus, how do you define who owns it? Like I, as a company, gather data about you. Is your data, is it your data? Is it my data? And now we're just off of probably trust layer and more into theoreticals. But like how do you think this plays out? Like what does the world look like here where this becomes more and more common? Because that's [00:33:00] the world we're moving into. So it's probably good to talk about it even if none, neither of us have any actual decision making power here. sean: It's really juicy and I think there's ownership and then there's also access of it, right? Because, one of the things we're all wrestling with right now is, sure, I can train your AI model on my data, but I don't want you to access my data. How do I keep my data mine? And also even again, law firms. You have these walls that are created around different parts of the law firm where if a big, giant, 20,000 lawyer law firm takes on Pepsi and then takes on Coke, those two law teams shouldn't be talking to each other. The information should absolutely not be shared. There needs to be hard and fast walls. So it can't, it's not just the firm's information. It's almost this structure of. How do I protect that? So it's a really interesting challenge of ownership and access and what you can do with that I don't think we have any real good solutions for yet. I. jeff: Yeah, I mean it's almost as an opportunity, a [00:34:00] way to monetize, you know, add another layer here of is there an opportunity to run it, to have a company where they're the. They're the layer of ownership, kinda like trust layer was doing ownership of, this kind of documentation that you have valid insurance and stuff. But is there a way, you could set the middleman where, we have all your data, but we hold it in security, but you can monetize it where you want by sharing it and, we can arbitrate that and there's pre-negotiated agreements and stuff and can that happen much more at scale? sean: Yes, and I think that there's also things like, with Indio, Indio is a sort of a, an insurance tech software platform where it was a lot about kind of collecting information. There could be things where you're starting to. Potentially aggregate and obfuscate the data to give benchmarking information. So you could predict, for example, when a, what's called a hard market is coming where insurance policies become harder and harder to get and more expensive to get. So you could start to predict that information from the aggregated information that you're seeing passed through. So you're not sharing specific firm's information, but you're using that information for [00:35:00] the benefit of the whole industry. There's also lots of hard questions that get asked around that kinda stuff, even though you're not sharing my info. You're using it in this aggregate. Is that okay? And what do you do with it? How? How do you hold it jeff: how obfuscated is obfuscated enough where it's no longer yours. It's mine now, sean: And I'm making money off of it and not paying you for it. jeff: Exactly. Yeah. I mean that, that's also saying when it starts to be public benefit versus versus profit and there's a lot of gray area that we've already seen some of it, right? When I. All the AI is scraping off the web and training their set on data's out there and paying some providers and not, but this starts to get even hair 'cause no longer corporation versus corporation. It's, open AI versus Reddit's One thing open AI versus Sean McCleary is probably another one. So one more thing I wanna talk about that is just one thing I've been doing is going around the country and been able to have dinner at, with a lot of PMs kind of in, in several cities. And I. One interesting thing that's come up at every single one has been, so what are you doing with ai? Because that's just what everything devolves to. I think we had an hour discussion. At the table the other day at [00:36:00] one of these dinners when we were down in Dallas, where one of the, one of the people I was sitting with just finally went, so what's everyone doing for workflows for their team in ai? Let's just trade tips. This kind of brings up and, we mentioned open AI there, but like this is an interesting world where, how do we start to build a world where not just the corporate enabled can use it, but everyone starts to get the benefit? Because there is a bit of this empty box syndrome, right? Like when you have a, just here's an AI type O. Okay. After a while I start to think of stuff, but this is half why people are asking like, what are you using it for is 'cause, we all have a blank slate. What have you seen? Do you have any cool things you're doing for teams or workflows? That you're leveraging and what do we do to fix this open empty box syndrome. sean: the open box syndrome is interesting. We're just starting to research it more and get an understanding of how comfortable people are with it. So you're certainly seeing a lot of product management training and things like that to teach PMs to be better prompters, to write better prompts. And that's an interesting, you're starting to see people have that as a skillset in their resumes, right? Of how they then can write good prompts. 'cause prompt is the [00:37:00] currency for doing things in an AI world as it exists today. The challenge is that beyond that, do we want to train all our customers to do that? That's never gonna happen. So how can we meet humans where they are and how also will humans evolve to meet that? It's a little bit like when Google got introduced and there was just this box right. We did a lot of user research around that as well, and there's a lot of like, I'm not sure what to type in that box. do I type A sentence? Do I type, name? What do we type here? And yet we got familiar and now we're all great Googlers. We can all Google just fine. jeff: This is interesting because this is not just an AI problem. We had a guy named Eric Metca on, and he's at EPO now, which is like an AB testing company but he was at cameo back in the day, and they had this problem where they'd show just, what do you want, Kevin from the office to say it was just big open text box and they saw actually tons of people just bounce. They actually ended up not buying and just leaving because I don't know Hell man, what are you [00:38:00] doing? Don't put me on the spot like that. But they found, they broke it up into, we all think less forms, less entries, better. This is a case of break it up into six discrete questions and they just saw conversion rates skyrocket. But yeah. Does that work here, you think? Or, what do we sean: I do think it does. , this is not a new problem. And to your point, it, when you talk about creating something, which is really what you're asking to do here, is create something from just words. Imagine you're writing a presentation. You've gotta, you gotta, you've gotta direct report and you say, Hey, look go write me a presentation on next that. Starting point is really hard to get from essentially a zero to a draft is really hard. A lot of work that has to go into that. If you hand them, here's my rough draft of it, go edit it. Ah, done by the end of the day. So I think that there's ways that we can think about how do we help people with this empty box. It's really interesting because there's a lot of people who are thinking about, oh my gosh, AI's gonna change the world and it's gonna be like the Starship Enterprise, or just talk to it. I don't think people [00:39:00] know what to say. Right. They don't know what to say, to ask for what they want. Certainly the tech savvy of us are getting there, but the average person out there, my wife's a physician and getting her to, she just got to up to speed with chat GPT, and now she's like planning our vacations through it. But it was really hard for her initially. She's I don't know. What do I do with this? It's a whole new interaction jeff: Right. sean: and I think that we're at this stage where AI got introduced. It was, there was a lot of vision about it. There was a lot of clunkiness about it of, all sorts of things that it was doing weird. We're at that point where we're in the trough of disillusionment, where it's I haven't really gotten the value that I thought I would get yet. The winning application hasn't really shown up yet for the average person, jeff: Yeah. sean: and we're gonna start climbing that hill over the next year and two, and then it'll be interesting to see how human behavior changes. And I think you and I talked about this earlier, but the idea of this is not the first time that we've done this, right? We did this with the web. We started [00:40:00] buying stuff online and the PE companies had to figure out how to monetize the web and what that meant and what people are comfortable with. And then human behavior changed. We've done it with mobile phones that we now carry around in pocket and literally have our lives on them. When they first got introduced, it was like. For games and maybe the New York Times. And now it's a, an essential part of our lives. And I think that AI is at that early stage where it sounds exciting, but we don't entirely know what to do with it yet. And humans don't really know how to interact with it. And it hasn't changed our behaviors yet. jeff: But like the internet give it 10 years and even, your grandpa's gonna not know what they did before it kind of thing. . Sean, it's been great having you on. I appreciate taking the time. I don't wanna take your entire, I could take the entire day and just talk about AI and how this is changing alone, but probably not great for either of us. I think we both have jobs to get back to. But thank you so much for coming on. It's been a blast, man. You are the motorcycle riding pink sticky note mind mapping kind of guy who brought together Macy's brick and mortar in online. It's been cool hearing all this stuff and definitely wanna keep in touch to, to see, what's going on, [00:41:00] what's new with you. But for now, if people have, further questions or think they have something to offer what's the best way to get in touch? I know, I think we met on LinkedIn. Is that a good spot for you sean: yeah that's easiest way to, to hit me up and I check regularly and will respond quickly. jeff: Awesome. It's been a blast, man. Thank you so much for coming on. I look forward to staying in touch and yeah, talk soon. Thank you so much. sean: Thank you, Jeff.