Giant Robots Smashing Into Other Giant Robots

thoughtbot's Incubator Program Mini Session 3: Episode 07: Knect with Josh Herzig-Marx

January 16th, 2024

If you missed the other episodes with thoughtbot Incubator Program partcipant and founder Josh Herzig-Marks of Knect, you can listen to Josh's first episode, his second, an his third to catch up.

Josh Herzig-Marx, founder of Knect, discusses the latest developments in his startup journey since his last appearance on the show. He emphasizes the program's value in helping founders like himself refine ideas and strategies. He particularly notes the program's effectiveness in addressing challenges unique to startups, such as managing professional networks and dealing with the rapid growth of online presence.

The conversation also delves into AI's technical aspects and potential applications and the practicalities and ethical considerations of using it in professional networking. Josh and Jordyn explore various AI use cases, distinguishing between beneficial applications and those they deem undesirable.


LINDSEY: We are back for our Incubator update with Josh Herzig-Marx and his startup, Knect. I'm Lindsey Christensen. I do marketing things at thoughtbot. We are also joined by Jordyn Bonds, who runs our incubator and does product strategy for thoughtbot. And today, we're going to be catching up with Josh and learn what's new since last we checked in. But before we get to that, we have an exciting incubator update: our application window has just reopened.


LINDSEY: You could be the next Josh.

JORDYN: You could.

JOSH: Don't be me. You should join the incubator.


JORDYN: Go to and apply. It's just that easy. The application doesn't take long, even though it's in Typeform, and we have gotten some feedback, including from Josh, that it's challenging to plan your application efforts because, as you all probably know, Typeform just gives you one question at a time. So, sorry, maybe we'll update that. But it won't take you very long. It's a pretty brief application. And we are looking for pre-product folks, so you don't have to have a lot. Don't worry about what you do or don't have. Just apply.

LINDSEY: Pre-product founder trying to figure out, is this problem worth solving? Who is it for? Jordyn and the team can help you out.

JOSH: And me.

LINDSEY: And Josh.

JOSH: And if, for some reason, you want to ask somebody about the program who isn't directly affiliated with thoughtbot, you should reach out to me. I'd be happy to talk about my experience.

LINDSEY: You should.

JOSH: I'd be happy to tell you what I think would be some reasons to join and some reasons that it might not be a good fit for you. And I'd be happy to chat about any of those things. It'd be my pleasure, in fact.

LINDSEY: That is a great offer.

JORDYN: It is a great offer. You all should take Josh up on that offer. He is an excellent sounding board and mentor. And additionally, if you get into the incubator, you'll just be in a Slack channel with Josh for the rest of time, inside of thoughtbot's Slack. So, that's another [crosstalk 02:05]

JOSH: Statistically, there's a good chance you already are.


JORDYN: You mean in a slack with you. That's true. Josh is in a lot of Slacks, not [crosstalk 02:14].

LINDSEY: Yeah. Once you go through the incubator, you're family for life.

JORDYN: You're family. You're here. You're with us. You can't get rid of us.

LINDSEY: And you're able to hit us up with the questions, talk to the other founders, so that's another great benefit of participating. All right, but topic of the hour, Josh, hey, how are you? How you doing?

JOSH: Lindsey, I am floating right now. We had our end of incubator session last official meeting. And we reviewed how we started, what we hope to accomplish, what we actually did accomplish, and next steps, and it feels really awesome.

LINDSEY: It does. That's so great to hear. And can you, at the top here, maybe remind folks who haven't listened before, you know, what was that beginning point that you came in the incubator or the problem that you were looking to solve?

JOSH: So, I had this Josh problem, which is that I am overwhelmed by the number of places that I am online and by the rapid increase in my professional network, professional social network, I guess you could say, but in my professional network, you know, see that comment a few minutes ago about how we're probably already in multiple Slacks together, whoever you happen to be online. Plus, if you're on LinkedIn, we're probably at least secondary connections on LinkedIn. Like, there's an awful lot of people, and it's growing really, really fast.

And as somebody with a whopping case of ADD, which just feels like making an excuse, as somebody in, like, this modern world, I was feeling overwhelmed, and I felt like I was dropping the ball. And my problem was somebody must have a solution to this. I cannot be the only one. I could not find a solution myself. And I thought, well, maybe if there is no existing solution, maybe we should just go ahead and build it. And that was the genesis of my application to the thoughtbot incubator, which was that even though I've done this once before, I had never done this alone. I don't want to do this alone.

And I thought that, you know, because of my experience with thoughtbot in the past and my understanding of, like, thoughtbot's unique organizational skills and capacities, this would be a particularly good fit for the thing that I wanted to figure out. And when I say figure it out, there was really four things I was hoping to get from this program. Let's see if I can remember them all in order. Number one, is this a Josh problem, or is this a broader problem affecting more people? Number two, this is, like, a ladder of problems, right? Like a cascading set.

Number two thing I was trying to figure out: if this isn't just a Josh problem, is there at least one identifiable and addressable set of people who think about this problem in a similar way with whom I could engage? Number three, if there is such a group, are they willing, ready, and able to, like, spend money on solving this problem? And then number four, which I guess is kind of orthogonal to the other ones, it's kind of alongside, is this thing to solve even technically feasible, right?

Because you can have this, like, amazing opportunity, but you just can't build it. And, you know, is this a thing that we could build or that I could get built within the resources that I might have? And I came in with some hypotheses, with some ideas. It's not like I had never done any research in this at all. But coming out of it, we have four pretty good answers. And I would not have been able to reach those answers with the same level of confidence, certainly not within eight weeks, if I hadn't gone through the incubator, and it's a really nice way to end the year.

LINDSEY: With a bow on it. The last time we talked, you had narrowed in, I think, on your starting target market. And you had also recently introduced a prototype into the mix. How has the prototype evolved?

JOSH: It's...and this is going to be no surprise to either of you or anybody who's listening. But, like, the difference between, like, talking about something in the abstract and actually having, like, a thing in your hand is night and day. So, the prototype actually evolved pretty rapidly. You know, it allowed us to try using it, like, to put on our own empathetic user analog hats and try it ourselves and be like, "Well, this doesn't quite make sense." This doesn't actually flow right. And it allowed us to show it to a lot of people.

I'll say, we are, by far, our own strongest critics, which is good. Mostly, when we showed it to people, people are like, "This is amazing." And they would ask us, like, really specific, weird questions like, "Where's, you know, your about page? Could I see your privacy policy?" which is, like, a really, really good thing to hear. Because if the only way to interpret that is the only thing keeping them from maybe, like, diving in and using it right now, besides it doesn't actually exist as a product, is, like, some questions around privacy because it seems maybe too good to be true. Like, that's a pretty good buy sign.

You know, we were expecting, like, "The screen makes no sense. Why are we swiping here? Where does this data come from? Is this really complete?" They're like, "No, I'm pretty much ready to go." So, that was good, helpful feedback, though we evolved it ourselves a lot internally. It's really nice having a thing. Do we use the term Pinocchio prototype or Pinocchio test [crosstalk 06:58]?

LINDSEY: Yes, I did hear that.

JOSH: Yeah, I like that. If this was like, you know, this wooden toy wanted to be a real boy, like, two weeks ago, it really, really wants...I don't know, Lindsey, we should, you know, get you in front of it. You're going to be like, "Why can't I use this today?"

JORDYN: That's definitely what we're hearing from people.

JOSH: And my answer would be, "Well, you can't, but maybe in a couple of weeks." [laughs]

JORDYN: Yeah, exactly. I will say I want to say for anyone listening in, though, that that was not, getting to what Josh just described where folks weren't really...they didn't have any hang-ups about the functionality or the value prop. They were basically just like, "What's your privacy policy? And when is it going to be ready for me to use?" It's not like the first draft of this prototype that was what we jumped to. I want to be clear.

The first time we showed someone, there was this interesting problem, which is that we were still talking to the wrong people, somewhat. And the prototype hadn't evolved to be the slam dunk that it is now. So, at first, it was like, we'd have these kinds of muddled conversations where people were like, "Well, I don't really understand what this is supposed to be, and I'm not sure about that. And this seems interesting," but then their interpretation of what that thing was would be, like, wildly off from what it was intended to be. I just want to make it clear: this was work and effort.

And the team did a really great job of iterating quickly based on, like, every time we talked to someone and showed it to them, we'd come back and say, "Here's what I heard." And it really pushed our thinking forward. Like Josh said, like, we are our toughest critics, so, like, every new version unlocked some new insights in ourselves about what it was we were actually driving toward. Really, just there's nothing like having a thing to look at and bang on to, like, clarify your thinking.

LINDSEY: There's nothing like having a thing. Jordyn, you touched on you were talking to the wrong people, maybe. How has that exploration of the core market evolved? Is it still the startup enthusiasts? Are you even more narrow in that? What are the updates there as our chief market focus get everyone thinking about this all the time, officer?

JORDYN: Yes. So, you know, startup enthusiasts is still the umbrella. What you're looking for with this is that you can guarantee pretty much every time you talk to someone in a segment or a sub-segment you will know how the conversation is going to go. And we've gotten there with two sub-segments of startup enthusiasts, which is repeat founders, key, key kind of nuance there. Founders, sure, but repeat founders really have this problem, for reasons we could talk about, and then chiefs of staff at startups, which is a relatively new role that's sort of emerged over the last sort of several years.

But those folks are really the people that you ask them about this pain point, and they immediately are, like, yes. They use the same words to talk about the pain point. That's another really strong signal. When folks are using the same vocabulary, and they say the same sentences in the same order, and you start to feel a little bit creeped out, like, you're like, "Did you see these questions before I...? What? Did someone pay you to say that?" is, like, how you start to feel [laughs] [crosstalk 09:59]

LINDSEY: Also, a marketer's dream. Oh my gosh, here comes the messaging, right?

JORDYN: Exactly.

LINDSEY: [inaudible 10:04]

JORDYN: It feels like a cheat code because you just get to reflect their language back to them. You don't have to write copy. They wrote the copy. You just show them it, and they're like, yes. And everyone's like, "Yes," and it works.

LINDSEY: Any thoughts to add to that, Josh?

JOSH: It's really good. I would say the bummer or the good thing about this point is we're getting diminishing returns from testing everything other than the actual product, which is good that we got there in eight weeks. But we're not going to learn, you know, keep on adjusting the prototype and making little tweaks and more user research. But the truth is, we're not going to get anything substantial until we get this into some users' hands.

JORDYN: Like you say, this is sort of bad news, but it's good news.

JOSH: Right.

JORDYN: It's how you know, right? When you get to the point where the thing is so clear, and the way to talk about it with folks is so clear that you're not learning as much anymore, diminishing returns is the right way to frame it. You really just need people to get in there and use it. That's the only way you're going to keep learning. That's the moment to build. Hey, everyone out there, don't build before that. That's when you build. And then you really build the smallest thing you can conceive of building, and then whatever that thing is that you've conceived of building that's very small, scope it back by 50% [laughs]. Do it.

JOSH: And it's a little humbling as someone who considers himself a founder but who had reasonable success as a founder and who has had pretty good success as, like a very, very early-stage, you know, zero to one and 1 to 10 product leader, has done this a bunch of times and actually coaches people in doing this, and came in with, I'm not going to lie, a pretty good vision in my head for how this stuff was supposed to work together. And it's so much better now.

Going through a process actually makes things better. This wasn't just, like, wasting time. Like, going through a process, a thoughtful process actually makes us much better. Like, the thing we're talking about building is much more likely to be successful than the thing I was originally thinking about building, right, Jordyn?

JORDYN: Yes. I guess it bears sort of diving into that a little bit, which is, you know, for all the founders out there or folks with a product idea kicking around your head, you're apt to have a little bit of everything we've talked about already. You have an idea of the solution you want to build. You have an idea of who it's for. You have an idea of what their pain points are. And you might be sitting there thinking to yourself, I don't need to do eight weeks of discovery. I already know the answers to all of these questions.

And it's possible Josh felt that way coming into the incubator, but doing the work, gathering the data, talking to a ton of people, what you can't understand before doing that is how much more confident and at ease you will feel once you have done it and how much clarity you'll have about what it is you need to build first because likely, you're sitting there with a vision in your head for this product that is fully featured, fully formed.

It is the 18th month. We just went into a hidey hole and built a really complex thing, thing. Cool, don't throw that out. But you got to begin somewhere, and you got to begin somewhere meaningful and valuable. And it's really hard to know where to begin without this discovery, without focusing on a specific person, talking to as many of those folks as you can. And really, it sort of writes itself. It does feel easy. But you've got to set aside the time and the effort to do the research, market research, whatever we call this, customer discovery.

And it thrills me to no end, Josh, to hear that that is how it felt for you, that you probably felt like you already knew the answer. But it just feels different, having talked to, I mean, how many people, 100-plus people? We were looking at the stats.

JOSH: Well over 100.

LINDSEY: Josh was talking to a bunch of people before he came to the incubator, and all the founders that we accept have been doing that. Like, we want to know that you've been doing that research. But then, I guess, coming into the incubator, you're continuing that process and maybe in a more structured or a differently structured way where the thoughtbot team is helping you, maybe zero in far deeper on the segment. Is that accurate to say? Just kind of the difference between, like, maybe some of the pre-research and then the thoughtbot-specific user interviews that happen.

JOSH: Yeah. I think they were more focused. They're both more focused from the audience, but also more focused from if it's not just you doing it; it forces you to have a more clear, here's the questions we're asking, and here's what we're trying to learn, all these conversations.

It's also really nice to have some diversity in who's asking the questions. As good or bad as I am at user research and user discovery, I am only one person. And having people with different backgrounds professionally, who live in different countries, who have different feelings about social media, basically, who are not me in a variety of really interesting ways, I think, made the entire process more interesting.

Caro, who is our lead designer on the project, handed off basically the summary document of, like, everything we learned, and she pulled out, like, little snippets from the interviews. First of all, that is not something I would have done had it been just me, like, let's be very, very clear. This is an incredibly valuable document, particularly as we consider adding additional people onto this project to be able to, like, translate insights. But also, like, this is, like, summarized in a way that, like, takes some real expertise. And I would have walked away with vibes, and instead, we walked away with like, structured learning.

LINDSEY: Awesome. So, the last time we checked in, also, you were very excited because you had just maybe started a technical spike and were starting to dig into the, okay, like, how technically feasible is this product? And I think, at that point, you all were looking at circling around this target market. Here are the main tools they use to communicate. What does it even look like to connect with those APIs? How possible is it? Can you give us an update on some of that work?

JOSH: The way that I framed the question in the very beginning was, is this a science project, or is this going to be engineering? And, for the most part, the answer is, it's going to be engineering, right? Some are a little bit easier; some are a little bit harder. But it isn't, like, reinventing new stuff, with one exception, and that is connecting up with iMessages, which has been in the news a little bit. And I honestly just hope the ghost of Steve Jobs comes back and haunts, you know, the Apple headquarters at Cupertino because, come on, guys, interoperability is sort of the future, and you're ruining it for everybody.

But other than that, I think we have a pretty clear path. I'd like to test out some of these. Like, you don't really know until you do it. I think that's kind of the next step of what we're doing is to, like, demonstrate that it is possible for a person to connect up a couple of different accounts. It is possible for us to extract data and turn that into information and insights in the kinds of ways we thought we could and then present that back in a meaningful way. I think that would be the next step for us to do. Mostly, everything seems feasible, except for iMessages.

LINDSEY: I've also, I think, heard some whispers of artificial intelligence for Knect. Is that true? Have you all looked at, you know, what AI's role could be in the solution? And how does that research look?

JORDYN: We assume it will be part of the mix. That said, I don't know how to frame it exactly. It's not like it's not an essential ingredient. I think the work with large language models and the democratization of that work recently is absolutely going to make this product way better than it would have otherwise been. But there are a lot of heuristics we've, like, been able to, you know, draw out and come up with that are, frankly, algorithmic, and they're not AI necessarily. Now, the line between big data plus an algorithm and AI in the popular lexicon, like, there's a big difference between those two things. But, like, as people talk about it, yeah, where does one end and the other begin?

But we definitely will be making use of a lot of the newest technologies, and we've dabbled in them. I've dabbled in them. I know, Josh, you've been playing around with some of them, too, to the point where we're like, okay, yeah, we can make use of this stuff. It will be a valuable kind of tool in our toolkit, but it will not be the sole basis of value. I guess that's the sort of nuanced answer. But maybe Josh has a more bite-sized hype machine answer to this. Yeah, AI to the moon, right?

JOSH: Um, no. My only answer would be more cynical. Would anybody rightfully start a company in 2023 without having AI in there someplace? Maybe I'll say something different. One of the things that we've wondered is, there's more than a handful of companies that are adjacent to what we're doing that are definitely looking at similar kinds of problems and that aren't building the solution that, clearly, some market is, like, desperate for. And these are not, like, wildly successful companies that have grown astronomically and changed the market. And, like, trying to figure out, like, why is that?

And one of the reasons is...I sound like a tech bro, right? There has been a paradigm shift in the technology world, but there really has been. What do, you know, publicly available LLMs like, you know, OpenAI's ChatGPT, like, what have they done? They have taken a whole set of problems that were once really, really complicated and allowed you to do a reasonable job of solving them much more easily than you ever could before.

And it takes some amount of imagination, to realize that, to realize that these things are more than just, I mean, every product I have on my computer has some kind of OpenAI ChatGPT-style thing in there, right? It's, like, 16 different variations on give me a prompt, and I'll write your essay for you, and they all kind of suck. But those aren't the really exciting uses that I've seen. It's the more subtle things.

There's a company called Booklet, which tries to replace, like, noisy email lists or noisy communities to something more calm. And one of its features is it'll send you a summary of what's been going on in the community since, like, the last time you checked in. And it gives you, like, two paragraphs to read, and they're really chill and really informative, and they don't make you feel FOMO. They don't make you feel stressed up. Like, okay, stuff's happened in the community. This is really neat. And it's all powered by OpenAI's APIs. And it's really kind of magical. And, like, you have to have a slightly different perspective to imagine these kinds of magical moments. So, that's what I'm excited about.

There's a set of things that we would have had to do with, like, terrible, complicated queries and, like, pattern matching, and freaking grep, or whatever old-school tools we would have had, you know, for doing things in the past. And now you just get to, like, shove text in one end, and say how you want the results structured and get the results back in the other end. And it doesn't have to be perfect, but that's okay. Like, we're talking about human relationships, which are inherently imperfect. So, I'm fine with this. And it's kind of exciting.

But we'll see in, you know, if we end up continuing going down this path. Like, that's the goal of the next stage is to be, like, okay, what are the easy things which we can generate out of this? Is there an intersection between, like, easy and meaningful? And if there is, this is pretty exciting.

JORDYN: Can I add something to that? Which is that the problem Knect is trying to solve and the way that we're trying to solve it, the way we've thought of solving it that's differentiated, lends itself really well to the current landscape of AI tools in that, and you were kind of getting at this, Josh, but I feel like it bears drilling into a little bit, in that what we are proposing here is not a set of deterministic things. We're not going to give you a to-do list. It's not, like, a linear...deterministic is really the right word. Like, there's a to-do list. There are things that make the cut. You got to go address them, et cetera.

We're way more trying to approximate the way a slightly more put-together person with more time would approach nurturing their relationships, which is just to remember more of it more of the time. It doesn't mean we need to remember all of it every time. That's not the kind of task this is, which makes it a really good task for the place that AI is at right now.

And I think where folks have failed in the past is that they've either tried to turn it into a deterministic set of tasks, which then just feels like another to-do list, another inbox in a series of to-do lists and inboxes that you have in your life that just make you feel guilty and inadequate. That doesn't seem fun to us. We don't think you need another one of those. Or other places we've seen this fall down, which is that it takes the current sort of state of AI and tries to actually do the deterministic thing for you, but it doesn't do a good enough job right now.

But where we've kind of landed in the middle is that, again, what we're trying to solve for is solvable in a way more probabilistic way. Like, can we get more of this accomplished more easily for you? It's never going to, like, completely, you know, do the task in this perfect deterministic way. But it is going to make you feel more confident and more relaxed à la Booklet, it sounds like, how to do that for this particular problem, which is a different bar and one we think we can clear. And that really does provide value. People are really longing for this.

LINDSEY: Jordyn and Josh, building on those descriptions of, like, kind of maybe bad AI use case, good AI use case, could you give some specific examples of, like, what that might look like for Knect, like, how AI could be used in a good way or maybe what you're trying to avoid, more specifically?

JOSH: Yeah. First, I'm going to start with what I want to avoid, which is, there are tools out there, and these may be interesting to some people listening, and if so, go find them. Good luck. But there are tools out there that say things like, "Keep in touch with your network at scale." And will use AI to write a message which you can send out to people without you ever having to, like, review it. That seems like creepy, futuristic sort of, you know, there's, like, a Black Mirror episode about that.

Like, the whole point of having, like, a professional network of people who you care about is actually interacting with them. And having some service, like, write some prompt, maybe in its own voice, maybe if it's really good in your voice to, like, let them know that you care about them, let them know that you're thinking about them is, like, that's just bad. I think that's bad. And we don't have any plans to do that kind of thing, even though most uses for AI in the products that I use are writing three or four paragraphs in response to, like, a prompt. So, certainly, that's the common use case.

It's not very appealing to us, and, frankly, in the people we were talking to, that wasn't one of the things that anybody ever suggested. It's obvious, but as far as we can tell, uninteresting, right? Just because it's obvious and just because it's straightforward doesn't mean it's interesting.

The things we're imagining, for example, is, talk about Jordyn. Jordyn and I have known each other since 2020, I think. And we have, like, a whole history of text messages going back and forth, which, by the way, we actually could integrate because we both have Android phones, you know, shout-out for Open Internet. It might be interesting to, you know, summarize some of that, like, I know Jordyn pretty well, but other people who I might have not talked with in a while, sure, you could present me with a whole timeline of our communication. But that isn't necessarily useful. I'll have to read every bit of it.

Why not, like, take all that and summarize, here's things you guys talk about. Here's things that, like, prompted your past few conversations: job change, got laid off, started a company, got a cat. Whatever those topics happen to be like, share some of those things. Bring me up to speed a little bit faster without having to literally review every word that could have been going back multiple years. That's a pretty good use of it.

If you think about the way that messages work, right? Like, my kids are now at the age where they have phones, and I can now text my kids during the day. I will just tell you, like, this is, like, an incredibly joyful thing for me to be able to send, like, stupid memes to my kids or, like, what's exactly the right emoji to, like, send to them or for them to send to me.

If every one of these things were, like, pushed to some kind of timeline, and I'm like, "What's going on with my kids?" Like, that's just, like, going back and reading through, like, your WhatsApp thread, which is something that isn't interesting necessarily, at least not from, like, a professional perspective. And there's, like, thousands of these things.

Like, why do I want, like, a record in my, like, database of people who I talk to that says, "OMG," or "K," or "lol," or those sorts of things? Like, that's, like, a phrase. It isn't a conversation. And we could use an LLM to go summarize what the conversation was all about, which is, by the way, a way more interesting thing to persist over time than, like, my daughter typing "JK, JK, JK," which I think is 15-year-old for laughing at me, but I'm not entirely sure.

LINDSEY: [laughs] Okay, so as you mentioned wrapping up, and you did your last meeting, and you've got your kind of takeaway docs. You know, one, I'm curious, like, if there's, for your last, you know, days, hours of the program, if there's any final morsels you're trying to get out of it, and then how that kind of leads you into, like, what's next. What are you planning?

JOSH: Let's do another one of these things in two weeks.


LINDSEY: Oh, okay.

JOSH: Yeah. I'm inviting myself back on your show. We have one more day of school then, like so many folks, we get in a plane or get in a car and go do some travel and try to disconnect a little bit from our professional networks. So, I'm consciously not trying to say what's going to happen next. I would love to have this conversation again, maybe in two weeks, in the new year, about what comes next. I don't know that I could have a meaningful one right now.

JORDYN: I will say what we are trying to send Josh off with into his R&R is what's it going to take to get to a viable MVP, not merely viable, but actually viable? Given what we know, given all this, you know, work that we've done in the last eight weeks, we now have, you know, the ability to envision what version one of something might be. And so, making that kind of argument: here's why it is what we're imagining it to be; here's what it is; here's what it would take to build that thing, gives Josh a lot of stuff to think about in the meantime in terms of how to accomplish that.

And the thing that will happen in two weeks is understanding a little bit more about, like, the actual, okay, here's the actual plan. But the ingredients are there, which is super valuable and is a thing we have done every time at the end of every incubator we've done. It's essentially is that what's next plan and why, why that thing. What's the ultimate upside of pursuing this product, and what's the near-term upside? And what's it going to take to get there?

Because that's often a thing that founders, especially for some founders, which Josh is not, but what they often can't get their heads around is there's this little feeling if you've got this big vision over here, and you've got, like, the set of things you could do tomorrow, really tasky things really, like, operational things, oh, I need to, like, set up a C Corp, but I need to...whatever those things are, right? What's in between? What's that near-term path that's going to directionally head in the direction of that big vision? It's, so far, always, what we have sent founders off with.

LINDSEY: So, if you weren't here at the very beginning of our session, we mentioned that the applications are now open for session 1 of 2024. I'm curious, Josh, what kind of founders would you recommend for the thoughtbot incubator? What's the profile of someone you might send our way?

JOSH: I'm going to say something, and I don't think I match that profile, which is interesting, and folks should think about that, what that means. But I would say that if I had to, like, pick a profile, having gone through this, I would say somebody with an idea, of course; ideally, it's one that they have some connection to. They have some personal passion for but, not just because it's an abstract idea but a personal passion that comes from their own experience.

And it's really great for somebody who hasn't been inside of a tech company before, at least on the tech, half the business. Tech companies have three halves: one half is, you know, the product building side of bit of it or the tech half, which is engineers, and product designers, and product managers. And the other half of that is the go-to-market side, like sales, and marketing, and customer success. And the third half would be, like, operations like HR and finance.

So, if you have experience in, like, the sales, or the marketing, or the customer success side, or the HR, or the finance, or corporate operations or that part of it, and, you know, you're familiar with tech coming from that perspective but maybe haven't been on the actually building stuff side of them before, this is a really, really good process. Because what does thoughtbot do? It does the building in tech side of things: designers, product managers, and especially engineers.

And it has this, like, legacy and this history and expertise, therefore, with, like, the journeyman program where they help, like, level people up in those areas and now are applying this to founders. Because as the founder, you do need to develop some ability to converse around engineering and technical stuff. And you really, really, really, really need to get good at the discovery side, especially of, like, product design and product management. And those are the things you're going to get to do and you're going to get to do with people who are themselves really, really good at it. And that's awesome.

The flip side is if you're, you know, a founder who is super attached to every bit of your vision, and you think you have the strategy all laid out and you're just looking for, like, warm bodies to build it, I mean, is it the insight team? What's the right level at thoughtbot? I forget the names of things, but, like, thoughtbot has, like, a startup program where you can give thoughtbot money, and they will build things for you. And they're also really, really good at that, but that's not the incubator program. The incubator program is probably a step earlier.

So, I think it is worth thinking, are you at the I'm so confident of my vision; I'm so confident in my strategy that I just want to get this thing built, then maybe don't sign up for the Incubator.

But if you're at the stage of I think this is a problem; I'm pretty sure this is a problem; I really want it to get solved; I have some vision, but I know it's going to change, then I think the incubator is really ideal, especially if you're looking to upskill yourself, too, because you're going to walk away with the ability to be conversant around the technology stuff. And you're going to walk away with a crap ton of experience with the discovery, qualitative discovery, like user interviews, quantitative discovery, like, you know, running ads, and landing pages, and all that stuff. Like, you're going to be really solid with that stuff after eight weeks because you will have done it.

LINDSEY: Jordyn, any thoughts?

JORDYN: I love all that. I think it's accurate. I would only say to those of you sitting out there who are thinking, I'm in that other camp; I'm very confident about what it is I want to build; I would ask you to do a little soul-searching as to whether that's actually true. Like, what evidence do you have? If you needed to stand up in court and defend your conclusions and your vision, could you?

And I say that as the person who, as a first-time founder, was deluded in that way. I thought I knew exactly what I was doing and for whom and why. And, boy, howdy, could I have used a program like this to actually get me to sit down and, like, talk to people, listen to them, figure out what was valuable and what wasn't, what a valuable, you know, initial market offering was going to be like. Ah, I wish really, really badly that I'd had something like this because I was pretty deluded. I don't even know, like, what the right word is. I just didn't know what I didn't know.

So, like the way you described it, Josh, I know Jordyn of 2017 would have been like, "That's me. I know this thing that I need to do.

LINDSEY: [laughs]

JORDYN: So, I don't need to apply to this program because I don't need to do any of that discovery work." But I was wrong [laughs]. I was absolutely wrong. I was wrong to the tune of, you know, two years and $150,000 of angel investment. So, consider, it is not idly that I say this to you, person sitting out there who feels very confident in your vision right now. Perhaps you have done all those things already; in that case, [inaudible 33:43] you don't need this. And you just need to [inaudible 33:46] with the thing you already know to be true. But ask yourself, how do you know what you know?

LINDSEY: Yeah, even if you...we can help you build the thing. But we're probably, also, still going to push you on [laughs] some of those things we [crosstalk 34:01].

JORDYN: Yeah, we're still going to ask. We're going to ask to see the receipts.

LINDSEY: Yeah [laughs].

JORDYN: And maybe you have the receipts, which is great, but we're still going to ask you for them, I guess, is my point. Every team at thoughtbot will ask you for the receipts, by the way, not just mine [laughs].

LINDSEY: The other interesting thing you touched on, Josh, was, I think, where we kind of started the incubator was with that target profile that you just described, which is, like, the less technical founder, and maybe even, like, a first-time founder. And then over time and seeing, like, applications, we broadened that as we saw, like, oh, you know, actually, also, technical founders and repeat founders do still need, like, help with this and can use guidance. So, we've expanded a bit, and maybe that is still, like, the person who gets the most value at the end of the day is the non-technical who hasn't really done this before. But yeah, we've kind of expanded to those other profiles as well.

JOSH: There's a reason that repeat founders are no more successful on average than first-time founders, and it's something really important that Jordyn said, which is, you may think you've done all this, but we're going to ask you for the receipts. Just because you've done this before doesn't mean you're going to be good at it. Chances are, if you've done this before, it's mostly because you got really, really lucky; ask me how I know.

So, it's nice to have. I mean, I described a profile, and I said that wasn't me. But I'll just tell you, as somebody who, like, spent his entire career, almost his entire career, in the tech side of tech companies, and I think I'm pretty good at it, I'm certainly not the worst at it, thinks I'm pretty good at it, it's still really nice to have a team backing you up in this early moment. It's really nice to have a team.

JORDYN: Yeah, I will say another thing that we've heard from every founder we've worked with is just how much more real and actionable their idea feels when they have a team sitting there with them taking them seriously, which is another thing, you know, I really would have benefited from is, like, suddenly, when you've got three or more industry professionals sitting there in a Zoom call with you, like, okay, what are we doing? Why are we doing this? How do we know?

The feeling of being taken seriously in that way and then having a bunch of people working full-time with you for eight weeks, they're in it with you; they're asking the questions; they're talking to people; they're coming back and saying, "I just had the most amazing conversation with someone. Here's what I learned," it just takes your project to a different level of reality.

Like, we're humans. We're social beings. We create reality together. And when you're working alone, you know, through force of will, you can do a lot, but with a group, it really feels like you're creating something together. And, like Josh said, having those other brains with other experiences in other contexts percolating on your idea it's like bringing a team to bear on something. There's just nothing quite like it, and it's a huge value of the program. Like, we can give you the programming and, in fact, you can go run the programming. It is published in our handbook. The things that we do together you can go do, but it is a whole other matter to do them with a team. It just feels different.

LINDSEY: Great. Well, I think that's where we're going to end today. I mean, Josh is leaving us hanging a little bit. So, we might need to...we're going to figure out a way to get your final thoughts, conclusions in a few weeks because I know everyone would love to hear what the plan is for Knect. Josh and Jordyn, as always, thank you so much. Any final thoughts or farewells from you today?

JOSH: I've really enjoyed it. I'm going to miss these folks. Though, apparently, I get to hang out in a special Slack channel forever.

LINDSEY: Yeah, you get to hang out.

JOSH: Which is nice.

LINDSEY: Exactly. You can't get rid of us just yet.

JOSH: Good. I wouldn't want to.

LINDSEY: All right. Thanks, y'all. And thanks, everyone, for tuning in.

Support Giant Robots Smashing Into Other Giant Robots