March 10, 2023

Episode 198: Aaron Podolny, Co-Founder & CTO of Scribe

Aaron Podolny is the CTO and co-founder of Scribe, a productivity tool used by over 300,000 users to automatically generate step-by-step guides. Before that, he founded a number of startups and sold Onward — an automation company — to Google in 2018. Earlier in his career, Aaron studied neuroscience and psychology at Yale and built machine learning and statistical models in economic consulting.

Julian: Hey everyone. Thankyou so much for joining the Behind Company Lines podcast. Today we have AaronPodolny, co-founder and CTO O at Scribe as Scribe. Their software let'scompanies capture and share how work gets done across their organization,analyze, optimize, and standardized processes in a way your employees willlove.

Aaron, I'm so excited to chat with you,not only about, your background, your experience, but also what you're buildingat Scribe. And I love, ideas and products. It's creating efficient and, andmore intelligent organizations, especially, with where, where the world is withfunding and, and companies are having to be a little bit more strategic abouthow they build companies like yours, at least in my mind, really create thetools that, that we need as founders to do so and to, to really run efficientcompanies.

But before we get into Scribe whatyou're working on, what were you doing before you started the company?  

Aaron: Yeah. So, prior toworking on Scribe, I was at Google. Google acquired my, my last startup. Andyeah, spent a little over a year there as part of Google Shopping. Worked withsome super smart people, became definitely a much better engineer in that year.

But then the, the startup bug kind ofbit me again and met my, my now co-founder Jennifer Smith. Yeah. And yeah, wewere sort of just off to the raises.  

Julian: Yeah. What was likethat? And it's funny, through acquihires like acquisitions in, in these areworking at a company. What was that process like, and, and how was the, theadjustment to shifting towards another culture?

Did they bring you in as their own unitor, or were you integrated into the whole Google ecosystem?  

Aaron: Yeah, for sure. So, wewere brought in sort of as a, an Independence unit, but I mean, we were alsoon, on a bigger team. Yeah, I mean, just, with Google Scale you're talkingabout hundreds of thousands of total, employees and of course they're, evenmore contractors as well.

So, I mean, it's a giant company, so. ,even if you're relatively big when you're acquired. And we were not, we weresmall, it was three of us. You're, you're gonna be part of something bigger. Sowe were, we were brought into Google Shopping. We had built a bunch of naturallanguage processing tech.

Mm-hmm. . So basically we had a chatplatform that enabled e-commerce companies to basically answer customerquestions in natural language. So someone could say, Hey, I'm looking for, Awooden dresser with at least six drawers. And we had stuff that would, surfacethat really easily. And, of course, like escalate to a support agent Sure.

If, if it was something that the, the AIcouldn't handle. So we were brought in to sort of incorporate that tech whichwe did and then also started working on, on other. Projects. But I, I think Iwas also just struck by how, like impressive my, my colleagues there seemed, Imean, I had this feeling that, that we were a bunch of, kind of like startupnoobs sort of like brought into to, to big Google and , my, my coworkers oftenhad like PhDs in like pretty complicated stuff like, math, astrophysics, like,basically just like super smart folks.

So it was cool to get to, to learn frompeople who had that kind of formal training and were doing, building at just a,a very different scale than we were used to.

Julian: Yeah. Did you bringany of the, kind of the, the systems or structure that, that Google had, wasthere anything that, say them as a company did well that you carried over that,that's made your team seem, either whether it's a little bit more mature or, ora little bit more well designed, well crafted.

What in particular? If, if anything, ,did you kind of, incorporate into Scribe that, that you learned at Google?  

Aaron: Yeah. Yeah. That's agood question. I think basically as a startup founder, and I'm sure you'veexperienced this too, you kind of have. Like everything you've done prior tofounding a company becomes almost this like menu that you can kind of orderfrom in terms of thinking about what's gonna work well in, in, in the startupenvironment.

Right. And so, when I think about thepieces, I mean, Google is just such a different scale that there, there is nota lot that like comes over one to one. Sure. But I do think. They have a verystrong culture of, of mentorship. And I did meet some just amazing managers,people who I was really inspired by there, and that was something that Ithought a lot about, especially as we grew the team here.

So we're now about 40 people. Yeah. Andso we're at that stage where, you know, we, we do have some, some managers onthe team and thinking about. , how they can really act as sort of like coachesfor the people who report to them. Sure. Sure. Think about, people up-levelingtheir career at Scribe as opposed to, like inputs and outputs.

Which I think is sort of the default wayof thinking about management. And, and that was something that I, I saw atGoogle and, and really appreciate it.  

Julian: Yeah. The fosteringthe talent piece is, is something that a lot of founders try to hack because,we're the competition for high level, high value talent is tough, but, the,the, the nurturing and, and the process of finding good people.

But then, up-leveling their skillset is,is definitely I would say more efficient, but also I, I think a little bit morestrategic when, when you're working on startup and describe Scribe, what, whatwas the inspiration behind the idea? When did you get the itch to jump back tothe startup world?

And really just abuse yourself with the,with the ups and downs of, of running a company, .  

Aaron: Yeah, sure, sure. Solet me give you some of the backstory, but, but first, just so people knowexactly what Scribe is, think about Scribe as a screen recorder that outputs astep-by-step guide. For the process you just did.

So rather than having a video at theend, you have crop screenshots, written text instructions, and that's all doneautomatically. And so the, the problem that we're, we're solving is think ofthat time when. A colleague or maybe a client asks you, Hey, like how do I dothis thing? And maybe you're like, oh, I guess I'll get on the phone, or, setup a zoom or something and show them.

Yeah. Instead you can just send aScribe. And the cool thing about that is you now have like the standardoperating procedure that's just ready to go next time. Someone needs to knowhow to do that.  

So that's what, what Scribe is. The, thebackstory here is we, we've definitely been on this journey.

We didn't start with like this productin mind and, and in fact, like if you had asked me, in 2019 when we weregetting started, would a product like Scribe. Solve a major pain point. Iwould've been like, like that, that that can't be that big a space. Yeah. Butwe were actually, so we, we had these, these really big ambitions at first totarget the automation space.

And so Jennifer and I we had had allthese conversation, especially with, with execs at big companies, and they weresaying, Hey, we, we love this new class of automation tools that we're seeing., but automations are super expensive to build and they're really expensive tomaintain. So, what we were trying to do was build a really easy to useautomation platform and one that was actually really empowering to Yeah.

Employees. One where an employee couldsay, Hey, this part of my job feels super repetitive, let me record it. Andthen sort of like, boom, like I have an automation. So yeah. super, likeimpressive, like if you can get all the way there, but it turns out that thereare some wrinkles along the way. Yeah.

To that journey. So we were initiallykind of like hacking away building automations for clients back in, in 2019,early 2020. And one of the problems that we realized was just. Enabling ourclients to articulate this is the business process that I need to, to automate.And so we built this recorder to capture the steps and let them share them withus.

And once we had that recorder, we, wesaw people were way more excited about that then actually the automation piecebecause they were like, oh, there were so many other things I can do with this.And it's easy. I get it. It's simple. And. We, at that point, we decided let'sgive this like a, a real a real shot.

Yeah. And so we did some validation and,and ultimately decided basically to, to start working on Scribe, what it istoday, sort of full-time.  

Julian: Yeah. Talk about that,that, that pivoting piece because that, there's been a few founders that youdiscussed the challenges of, of making a pivot or, or knowing when to.

Say, working on one product to work onanother feature version of a product that, that they're getting more tractionin. What kind of validation did you go through to, to then really encouragethat, that direction that you took with Scribe?

Aaron: Yeah. Yeah. So, theway Jennifer and I thought about this was, We were really, we, we believedreally strongly that we wanted a product that like anyone could use.

Right? Like, yeah. And I mean, I knowthat sounds like, well, of course wouldn't you want that , but, but like, wewere like, given the, the choice between building something that actually mightbe. , complicated and you can sell for, many millions of dollars per contract.We're building something simple that provides a bit of value to like a broadbase of people.

That second one, we were like, that's away more interesting business, and we kind of knew that from mm-hmm. yearlydays. And I think having that kind of like. sort of bug in, in our minds.Mm-hmm. . Yeah. Then made us ready when we, when we saw Scribe emerge as, assomething that people were excited about, as like, Hey, like this is it.

Like this feels like the kind of thingwe've been searching for. We then decided to be like, like kind of scientificor as much as we could be about it, like, we were like, In a month. So we, wedid a big product hunt launch and we're like, okay, like what does it have tolook like quantitatively at the end of the month for us to decide to doubledown on, on the pivot basically.

Yeah. And. What we set, I think likeour, our goal was around maybe like, actually like 300 documents created byYeah. So like, not actually that many, like compared to of course where we arenow. But what was cool is, so we did hit that number and. People kept using theproduct like a month later.

Mm-hmm. So it wasn't just, you get a bigburst of new users. Of course people try it out. It was a month later, therewere people who had incorporated it into how they worked and actually, like,were emailing us that like, this is incredible. Like this has saved me so muchtime. Yeah. And keep in mind this is like no marketing.

There's no like gamification in theproduct. There's no notifications. It's just like people picking this thing upand using it. Yeah. And seeing. Made me feel like, okay, this is special. Andit would be like a mistake to, to not focus on this.  

Julian: Yeah. The technologypiece is so fascinating cuz you talk about the screenshots and then not only doyou input all this information, you record and, and all this is kind of thenorganized in an easy to digestible way.

What goes into building that and, and,and I'm talking about like the user experience and making it easy to identify.Important pieces of information. Is it some kind of data labeling in thebackground? What, what are you doing that, that allows for that reallysophisticated and, and, and unique but also simple display of, of that process?

Aaron: Yeah. Yeah. So let mestart actually up at the like kind of principles level and then we can moveinto, The technology. So I mean, at a, at a high level, we have this value hereof like, we do the heavy lifting for the user. Yeah. If there's ever an optionfor the software to like do the work we build that first.

And so like, to give an example a lot ofour users, it's important for them. Redact sensitive information that might bein a recording, right? So maybe they're showing something in their CRM and theydon't want, their client email addresses to, to make it in the recording. So,Like V zero of this was, well, we can let them sort of draw boxes after therecording.

That's a fair bit of work. And so wewere like, well, like what is the Scribe version of that? And so we justrecently released this feature called Smart Blur. And what it does is duringrecording it identifies all that information and like blurs it in real time. Sorather than that being something that's like on a user's mind, like that'ssomething that we can we can help with and.

So that's sort of the principle in termsof like how we actually do it. I kind of think about it as there's thesestreams of data during a recording mm-hmm. . So like one is, what's visually onthe screen. Then there's like the click stream and then there's keystrokes. Andso what we're, what we need to do is basically collect all those in sort of theright way.

And then do basically, Have some rulesabout like what's important enough to, to surface mm-hmm. . And to be clear,we're still like improving that over time. Sure. And I think we have like a lotof, a lot more work to do. But like, it's, I mean, at, at base that's like a, afairly, simple like sort of data stream that then we're, we're sort ofoperating on top.

Julian: Yeah. It's soincredible to think about the, the process behind that. But I, I love that you,you kind of talk, talked about at the principal level cuz that's a lot about,like the Airbnb founder always talked about the 11 outta 10 experience and howcould essentially I go through book, book my Airbnb, but then have everythingelse cater to me.

Almost to where I didn't have to evenand move a finger. Obviously that's that's, you know's distant in future, butworking towards that you can find little improvements, which I, I love thatphilosophy. Thinking about just like, the incumbents that were before you oreven other popular platforms similar, my companies used Lumen a lot of waysand, we use other things along as well to kind of capture processes anddistribute it.

How do you view kind of building nextto. Say the looms of the world or other incumbents or other competitors that,do something similar. How do you think about differentiating selling to your,your product and, and also creating a product that's, similar but, butobviously has a unique value prop that, that the others don't.

Aaron: Yeah, for sure. Yeah,so I think the starting question here is sort of like what category are, are wein? Mm-hmm. , I think like a lot of folks, At first think of, oh, Scribe wouldbe in like the knowledge management category, right? So you think about likeproducts like Confluence, even, notion to an extent.

Mm-hmm. , but I think what I realizedpretty early on is that Scribe isn't really a knowledge management tool as muchas it's a communication tool and, and it's a communication tool thatcommunicates about a particular kind of thing, which is workflows, right? Likehow to use software to get work done.

Yeah. Once you think about the categorythat way and, and I think our users do as well. Like, at, at first you'd think,oh, there's like, this is such a crowded space, but like our users actuallydon't, don't feel that way. They're like, oh yeah, we use Loom for some thingsand we use Scribe for other things and like Confluence for a third set ofthings.

And and I think the reason for that islike communicating about how work gets done is actually like a pretty d. and,and I think somewhat unsolved problem. . Yeah. I mean it depends on like thetype of work, the system's involved. And I think what's helpful that thatScribe kind of solved early on was like, let's give people a sort of generalformat that kind of works for a broad range of tasks where they almost don'tneed to think about, oh, like, what is the thing I'm trying to, to share withthis other person?

They just turn on the recorder, do thetask, and. , boom, it's, it's done. And so, yeah, I mean, to some extent thisgets into category creation, but I think mm-hmm. category creation is oftenthought of as like a marketing exercise. And I actually think of it as more of,it's an emergent property. It like, comes out of how our users are deciding touse our tool alongside other things.

So, I, I think. In practice, it's turnedout that this does seem to be like this new space. And yeah. Anyway, that'ssuper exciting to me.  

Julian: Yeah. Yeah. Tell us alittle bit about the traction. You said tell us how many users you have, how,exciting about, not only from when you can see the product till now, but alsothis new coming year.

What are you particularly excited aboutin terms of the growth that you project for Scribe to have?  

Aaron: Yeah, definitely. Sowe're now around 400,000 users. And yeah, I mean, I feel like we've come a longway which is, which is super exciting. I think when I think about. What's next?I mean, so first of all, we, we have this we have a, a self-serve business.

We also have a, a, a thriving enterprisebusiness. Mm-hmm. So big companies using Scribe. And increasingly the questionI think about is what can happen when you have a lot of users in the platformlike working together, right? Mm-hmm. working alongside each other. Figuringout this is like the new best way of, of doing a task.

And enabling them to learn from eachother in a way that like feels very much aligned with those Scribe principlesof like, again, doing the heavy lifting. Yeah. So much of knowledge managementis super manual and, and l and d as well. Learning and development as acategory. Yeah. It's like go create this course, like go take this quiz.

Can we instead look at, at all that,that work that's being captured in Scribes within an organization or, or maybeeven, sorry, maybe even broader than the organization thinking about like, theworld and like up level. Each other. Yeah, right. Like Uplevel teams. And soyeah, that's, that's the direction I get excited about.

Julian: Yeah. Yeah. And tellus, thinking about kind of outside, looking in what are some of the biggestchallenges that Scribe faces today?  

Aaron: Hmm. Yeah. I'm like,where to start? I mean, I think about, because we're in a fundamentally newcategory mm-hmm. , there's a lot of I would say like product education that,that we have to do.

Yeah. Right. Like, and, and also, Imean, I think more than maybe education, like how can. sort of build on thisfoundation cuz we have this, I would think this really sharp wedge into thisbroader space of. Of like work, right? Which is like the data of, of how workgets done feels like it's one of the most valuable pieces of data that livesinside every business.

And in most businesses it's notdocumented at all. Right? It's like it's in people's heads. And it gets handedsort of from person to person and oftentimes it gets lost. Yeah. And so as wesort of expand beyond what we have today, how do we keep that? Because it's soeasy to lose. I mean, I think that's probably the, like when I think about whatthe journey feels like now versus what it felt like a year or two ago, like twoyears ago, it was like, well, you have a cool idea and you're like, let's buildit.

Okay. Like, let's just get it out infront of people. But now it's like, well, how do you incorporate it into aproduct in a way that fundamentally is compatible that like yeah. Is part ofthat, that path forward. And. It's, it's tough.

Julian: Yeah. Yeah. Ifeverything goes well, what's the long term vision?

Aaron: Yeah. Well I do thinkthere are a number of ways that this, this gets big. I really get excitedthough about this world of. people basically being known for what they'rereally good at. Yeah, sort of within the workplace. So it's like, think aboutLinkedIn today. LinkedIn is essentially just resumes.

I mean, yes. I mean, there's a feed inall this other. Stuff, but like the way we identify other people on LinkedIn islike, well, it's like, what schools have you gone to? What jobs have you had?Yeah. And then they had this like, skill chip, but like that, that is sodisconnected from mm-hmm. , the skills that people develop over the course oftheir career.

And what's interesting about Scribe isit becomes a way of demonstrating. your area of expertise. Yeah. And I think,anyone who spends some time in the workforce, they, we all develop expertiseand right now, It's like that, that's actually really difficult to, tocommunicate. And so I think about a world where that's completely transparentand then that actually becomes what we're known for.

And so rather than being so resumefocused, it's like we become, as a society, as a working culture expertisefocused instead. Yeah. And I think that's what we, we can enable here. If wejust keep pushing in the, in the right.  

Julian: Yeah. Yeah. I lovethat. I always like this next this section I called my founder FAQs.

So I'm gonna ask you some rapid firequestions and then Sure. We'll see where we get. But first question is what'sparticularly hard about your job?  

Aaron: Hmm. For me, so, soI'm the cto. I built a lot of the early product and I think like one of thechallenges for me is knowing. When to let other people kind of like take over,because I still love writing code myself.

I mean, it is a really meaningfulexperience for me. It feels like this creative act. So, so yeah. I mean, tosome extent it's about delegation, I guess, which, is a classic founder toomuch.  

Julian: Yeah. Yeah. How, and,and where do you go to hire the right people?

Aaron: Hmm. Well, I mean, Ithink one of the things that's cool about having a product that is increasinglypopular like ours is a lot of folks reach out to us, actually.

Yeah. But But yeah, I mean, I, I thinkin some ways there's, there's also some, some boring answers around, well, wework with recruiters, um Sure, sure. Yeah. But, you know, we do try to hire notjust from, traditional resume driven type big firms we're we, we we're lookingfor basically like scrappy people who are excited about building somethingfundamentally new.

Julian: Yeah. How do youmaintain remote working culture? What, what, what do you do say, saystructurally as an organization that essentially keeps everyone at least withtheir compass line towards the vision of the company when, it's difficultbecause you sometimes get lost in, task lists or communication.

I think the, the, the essence of why wehave values and principles as companies is, is to fall back on. Alright, whatam I doing today is focusing on this particular objective. How do you reinstillthat in your team?  

Aaron: Yeah, I mean, I don'tthink anyone has cracked the code for, for hybrid work yet. I mean, I'll tellyou what we do, which is one, we effectively have two.

One is in San Francisco and the other isin Belgrade, Serbia. And so, so rather than being totally distributed we, we dohave sort of two places where we, we hire more than others. And then we, forthe people who are remote who aren't in those areas, we do try to fly them outpretty often. Just so we can be together.

Together or in one place. Yeah. Yeah.And I. That makes a big difference.  

Julian: Yeah. For, foundersout there who are in a similar position or, working through their buildingtheir company, what advice would you give something that maybe you knew, on inyour entrepreneurship career?

Aaron: Yeah, I thinkparticularly for folks who are in the early days, I think it's very temptingbefore beginning work to try to be like really process driven about like,picking your idea. And I have a lot of friends who, who have done this, right.They'll have a list of like 70 ideas and they'll have columns and they'll tryto score everything.

And I think there's, there is value tothat, but I, I mean, personally, Have just always started building . Yeah. I, Ithink there's like so much that you can't know in advance, even if you do allthe classic validation stuff. So, say you pick a target market, a persona, yougo ask all the right questions.

Yeah, you hear that there's demand forsomething. It can turn out that like you build it and maybe it was like thereis demand, but it's not at the relevant price. Or like maybe something'sactually changed about the world since you had the conversation. It's tough,but I think once you are engaging with people around something that they cansee suddenly that feedback is really concrete.

They're like, this isn't working for me.Or you see that they're not using the product and you're like, huh. Like, whyis that? . I, I guess like the, the advice is sort of just like, go for it andgo for it a little earlier than, than you think you, you're ready? Yeah.Yeah.  

Julian: If you were to wave a,if you had a magic wand and you were to have, a wish and, and wave a magicwand, what's one thing that you would like your company to have now versushaving to wait for it or not having it at all?

What's one thing you wish that you wouldhave now?  

Aaron: Hmm. . Well, I willplug a role we are hiring for right now, . So we are looking for a head ofsales specifically someone who has experience doing product led sales. Soworking at, a PLG company. Yeah. And that is I mean, we, we know there are alot of people out there who have done it, but it's definitely like a subset of,of sales leaders.

So, I like, I wish we had someone heretoday cuz like, people are knocking on our door like being like, Hey, want totalk about. About an enterprise product. And, and also we have a lot of ofusers across a lot of big companies who are using our free or self-serveproduct. So, we would love to have someone in seat who can help us figure outsort of how to take those, those engagements to the next level.

Julian: Yeah. Yeah. I alwayslike to ask this question, just really get to the I, I think founders areexceptional at extracting knowledge, and it's fun to see what types ofknowledge that they extract from. But what books or people, whether it's earlyin your career or now, what books or people have influenced you the most?

Aaron: Hmm. Yeah. I, so, soin, in my, in my last company onward I, I was lucky enough to meet Elis Torresand David Cancel Of, of Drift. And, only had a few conversations with them, butI was really impressed both at like those conversations, but also what theybuilt with Drift, I think.

Yeah. I mean, for, for those who don'tknow, drift is I mean, it's, it's basically an automated conversationalmarketing platform. So, chatbot sits on your site and basically drives leadsfor you. Yeah. But more than the product, they built such a strong culture,like Yeah. All over LinkedIn if their employees were telling their story and,and from, I mean, this was years ago before that.

Sort of a, a more common thing. Yeah,there's just such a, a level of excitement also just that sense of like hustleand motivation and like, we are gonna make this work. So I think those twofolks just made a, a big impact on me. And, and I, they've, they've written afew publications. I'm kind of like blanking on, on the name.

I think we've actually written a fewbooks. I think I've seen it in the airport, . But I know I've read a coupleand. Yeah. Just really, really impressed by what they've done.  

Julian: Yeah. Yeah. Before we,we, we give you a chance to give us all your plugs. I know we're coming to anend episode here, but is there anything that I didn't ask you that you would'veliked to answer or that that I should have asked?

Aaron: No, I mean, this wasgreat. I really enjoyed the conversation.

Julian: Of course. Thank you,Aaron. Well, I appreciate you not only going through your journey as a founder,but how you've built Scribe, how you, how you kind of, it wasn't your originalidea, but through the process of building you, you were able to identify wherethe need was and I, I love that because, it, it is such an underrated part ofthe founder journey as.

Is building and adapting. And there'sone founder once that said product market fit is just a moment in time. Andonce that time changes, you, you, you move on to the next thing. And and it'samazing to see where Scribe is now and, and where you're going to take it forthe near future. So, hope you enjoyed yourself on the show and, and thank youagain for joining us.

Aaron: Yeah, of course.Thanks Julian.  

Julian: Of course.

Other interesting podcasts