Finding Product-Market Fit and Winning Fortune 500 Clients

Intro 00:01
Welcome to Startup Success, the podcast for startup founders and investors. Here you’ll find stories of success from others in the trenches as they work to scale some of the fastest-growing startups in the world. Stories that will help you in your own journey. Startup Success starts now.

Kate 00:19
Welcome to Startup Success. In this episode, I sit down with Doug Tallmadge, co-founder and CEO of Gradial, to explore his journey from SpaceX to building an enterprise AI company. We discuss how Gradial evolved from its original idea into a platform for enterprise marketing, what it takes to win Fortune 500 customers, and the challenges of scaling a startup through rapid growth and fundraising. Doug also shares lessons of finding product-market-fit, knowing when to evolve, and building a company for the long term. Welcome Doug. So welcome, Doug. It’s great to have you on the show.

Doug Tallmadge 01:07
It’s great to be here. Thanks for having me.

Kate 01:09
Thank you. I want to walk through your background before we get into Gradial and what led to the founding, if you wouldn’t mind, because you have a really interesting background.

Doug Tallmadge 01:20
Yeah. Well, you know, I started my career at Starlink, you know, part of SpaceX, and it was pretty early in the kind of Starlink history, so it was, I think, about 50 people and no satellites, no customers. And so my first 11 months was kind of right as Starlink was getting the first satellites in space, and so you know, I think we went from a whiteboard to 60 sats in space in about 11 months. And then you know the next 12 months we’re we’re getting the first paying customer and then scaling that to you know I think about half a million customers. It was just a crazy couple years so that was really fun. And you know my background was coming from more like a physics and engineering background but kind of at the same time you know I was watching kind of AI start to pick up, and so you know, ChatGPT came out in the end of 2022, and me and my co-founders just decided to quit our jobs, and even though we were enjoying what we were doing, and then start a company.

Kate 02:13
That is incredible. I mean, first of all, what an experience that you had, right? To be at Starlink at that time of transformation. I mean, that must have been incredible to be a part of that and to see how to scale so quickly and change and all of that. And then, what gave you the courage to leave, though? Because that’s there’s probably a lot of people listening that want to hear about how you had the courage to leave.

Doug Tallmadge 02:40
Yeah, it was hard. I mean, I remember my first week at SpaceX. I remember thinking to myself, “Oh my gosh, how am I ever going to leave? This is such a cool company. And like, I entered in 2017, and so even before Starlink, it was working on you know Falcon, Dragon, seeing the factory. Like one of the cool things was you would go through the factory, and there was an espresso station basically underneath where the rocket hardware was moving. So you could literally be like getting an espresso, and then next to you, someone’s assembling the Merlin engines into the base of the rocket, and then there’s like these big cranes overhead that are moving hardware. And it was so much energy coming from that environment. And I’m thinking, well, how many companies can compare with this? So I would say, like, I mean, that was very much where I was at for about four years. And what I think made the difference for me was, with Starlink, it was very clear when I left that there was a tremendous amount of work ahead, and in some ways, probably 90 to 100% of the work was actually ahead, right? And I think that’s true of any ambitious project. However, at that point, Starlink was a scaling company, and so there is a very different feeling, and this was the benefit of getting to see it before it had any customers. Right, is like it was a very different feeling in those first two years or three years. And one of the things is you genuinely don’t know if it’s going to work. So by the time I left, it was obvious that Starlink was a great business. I would say like some people would look at it today and say it might be one of the best businesses, right? And obviously, you know, there’s a lot of consensus around that. When we were working on Starlink in 2018, 2019, there was a question of: Will this ever work? Does it economically make sense? Can you launch enough rockets fast enough to build the constellation? Can you build a satellite that’s cheap enough? Can you build the hardware that goes on someone’s house that’s cheap enough? And it genuinely was not clear to me. I was worried about this at the time, and I wasn’t a founder, obviously. I was an engineer who joined right out of college, so it wasn’t my decision, right. It wasn’t something that I had a lot of ownership over, but it was something that I would think a lot about: is have I gone and joined a project that’s going to get canceled? Is the ultimate output of this that we’re going to work on it for two years and get canceled? And if you look at aerospace and telecom in particular, there are a ton of these kind of constellations and projects that kind of did just get canceled, as like Teledesic in the Seattle area 25, 30 years ago. That feeling of operating in uncertainty and seeing the kind of product and the mission and the team pull through it to something that ends up being amazing. That is kind of a very satisfying and addicting feeling. And so, what I think happened for me that kind of partly led to the conviction is a confluence of two things. One is, as it became a scaling challenge, it was still fun, but it was less uncertain, and so there was a little bit of the adrenaline junky effect. The second was there was a sense that if I was ever going to start a company, right after ChatGPT was one of the best times to do it. And I think it’s a great time to start a company now too. But I think that it was a particularly good time in a lot of ways to start right after ChatGPT, and so that was kind of the instinct. It might be seven or eight years before another opportunity like this comes. Like if I’m going to do it, I better do it now.

Kate 05:44
That makes perfect sense, actually. I appreciate you walking us through that because you’re right, there is a big difference once a startup is in that scaling mode than those beginning days. Very big difference. And you were leaving to start your own company, which is a really exciting thing to do as well. Tell us what the idea was, and what prompted you. You know what you saw, what problem you were trying to solve with Gradial.

Doug Tallmadge 06:11
Yeah. So I think the big thing was one of the fun things about watching something scale is the processes all have to be upgraded, right? And so as we started to scale. You know, you had customer support ticket volume. You had just a level of operations that was missing. You know, even marketing, right? Like there was a point in time where Starlink didn’t have any marketing, and all of a sudden it has marketing. And so then you think, well, what fraction of that work could be done by AI? And so the big moment was we were having brunch actually in Seattle, me and my co-founders, and it was someone’s got to build ChatGPT for enterprise. This was a couple weeks after ChatGPT, and we hadn’t heard of anyone working on this yet. Obviously, today enterprise AI is everywhere. It’s like the biggest story, right? But it was not the case even three or four weeks after ChatGPT. Like people weren’t yet talking about it. I think something that was pretty interesting that’s come out recently is that even internal to OpenAI, people were questioning if this was a product or more of like a kind of fab.

Kate 07:07
Right.

Doug Tallmadge 07:07
Anyway, that was our kind of moment. What if I had ChatGPT, but it was connected to all of my data, and it could do work for me? And then very quickly in the first eight weeks, we realized that that was too broad, and we started to say, okay, we need to specialize on something. And where we specialized was we want to take all the work that’s in ticketing systems like Jira or Workday or Workfront, but also Workday, but all these systems that basically track to-do list items, and we want to automate that work. Then, as we started to kind of work down that path, it became clear you probably should specialize in a type of work. It’s very different to do a software ticket in Jira than it is to go do a marketing campaign in you know Workfront or Asana or Wrike or something like that. And so that led us down a path of trying to pick an area to focus on. And I think in terms of why we picked marketing, partly because every company has marketing in some form or another, and so you know we didn’t want to specialize in something that felt very narrow and niche. The other thing was marketing has this kind of mandate of growing the top line of the business, and so the work is never done. If you think about accounting or other parts operationally of the business, it’s kind of like if I’ve reconciled all my books, if I’ve done all of the accounting that I can do. What more value can I offer? So you can go from 100% of the work being manual to 0% being manual. And you can do that automation, and you can generate a lot of value. Feels like it caps the upside. And then you would be thinking, how do I build a product that captures more value and does more for customers? And I think that in marketing, the benefit is because it’s a competitive sport, where its companies are trying to reach the same 100 million households, the same customers, and they never can stop doing it because you know their competitors aren’t going to stop, right? So there’s that competitive nature, and the top line impact means that it really embraces innovation because you need to be ahead of your peers as a marketer. And so we thought that that lent itself to an AI company because we thought we could deliver that market-leading innovation to customers, and that they would be much more excited about it. Whereas in some parts of large companies, there might be more of an antibody response where it’s like I actually don’t want AI because I don’t want to cut all the work. It felt like marketing was a grow, like grow the pie, not slice the pie type of thing.

Kate 09:30
Wow, I appreciate you walking us through that. First of all, I mean I love how you explained how you and your co-founders were sitting around talking about how scaling companies need these operational processes and things because I think that’s where they struggle a lot. We see that a lot on this show. But then the whole landing on marketing is so interesting because the way you outlined it, it makes sense. But when you first talked about the problem you were trying to solve. It doesn’t automatically lead you to marketing. So, how soon did you get to marketing? I mean, was that like a few months or a year or two after you already had investment? Do you see where I’m going with that?

Doug Tallmadge 10:13
Yeah. So it was probably a roughly six month journey, and I would say six months in order to really commit to, because part of it was you start to have a hypothesis pretty quickly of oh this could be interesting, but you know we had some early customers that were using Gradial for kind of SDR for rev ops for internal CI/CD like kind of technical CTO type buyer personas. Those are very different customers. We had all of them at the same time. And so we got to a point where it was, you know, we had a relatively small number, but very different customers, and we were basically realizing, look, if we’re going to be successful, we have to actually pick an avenue here. We can’t build something that’s general platform for everyone. We have to have an opinion. And that’s when we had a moment of reflection, and we said, you know, let’s go really focus on marketing, and then how do we kind of provide an off-ramp for the other customers? And so, you know, what we actually ended up basically doing was saying, “Look, you guys could keep using what we’ve got, right? Like this product, but you know, the forward-looking innovation is going to go over here. And you know, and I could get into more details on kind of the tech side. It was definitely the right decision for us.

Kate 11:17
That’s great. I think a lot of founders probably find themselves in that situation. I don’t think they all pick an avenue as well as you did. They try to be too many things for too long. I love that you did that early on. Can you explain to us how Gradial helps in marketing? I think that would be interesting at this point because marketing is challenging.

Doug Tallmadge 11:38
Well, I think that like most things, there is a tremendous amount of operational and manual work that goes into anything high quality in marketing, right? And I mean, I’m sure you know this. You know, just working on something like a podcast, right? It’s like there’s editing, there’s promotion, right? Are you going to go post it on a website? Are you going to go post it on social media? All of those steps require time and work, and the larger the company is, they require handoffs. Because especially at a larger company, it’s not necessarily one person doing those steps. A lot of times, they’re different people in different time zones, and what Gradial does is it brings an AI harness to that whole problem space and lets you basically have an agent operate across that workflow. What we uniquely do is we bring an agent that connects with all of the tools that marketers use. And so many of these tools are really just not designed to talk to each other. Where it’s like, if I’m updating my website, I actually also want to send an email. Well, those are two different platforms, and a lot of the connective tissue is done manually by someone effectively copy-pasting. And that’s where you end up with a ticketing system like Wrike or Jira or Asana or Workday or Workfront. You know, the list goes on. And that’s I think kind of where Gradial comes in is we bring an agent into that whole workflow and execute all the different all the different steps.

Kate 12:57
That’s fantastic. As somebody who’s been in marketing my whole career, I have to tell you, my early days working at Visa, I remember thinking, all we are doing is moving projects from department to department, right? Like so much of it is sign off. It’s got to go here. It’s got to go there. Then these two things have to talk to each other. Anybody in marketing, like you know, on a big enterprise level, will totally relate to what you just described. So then, were your customers primarily Fortune 500? Is that where you were kind of your sweet spot?

Doug Tallmadge 13:30
Yeah, you know, we’ve got about 20% of the Fortune 500 using Gradial today, and you know, obviously growing that very quickly. And we also have a large number of customers that look a lot more like Gradial itself. And so our marketing team uses Gradial, and so I would say it really is you know any company that has more than a handful of marketers, like more than you know five or six people, you start to get to a scale where you know you actually have handoffs and you have you know large scale campaigns, evergreen content, an editorial calendar, and it becomes work. When it’s like a five person startup, you know, marketing is often much more like I’m going to pull out my phone and I’m just going to either tweet something, right? Or I’m going to go and post something on LinkedIn. But as you get larger, you end up with logistics. Where’s the logo? What’s the approved font? How do I go and lay this out so I’ve got consistency? Where am I going to put my web content? Am I going to put on a CMS? Am I going to put it somewhere else? All of that logistics, especially when you want to do more complex campaigns. You know, if you’re doing B2B you might be doing an account-based marketing campaign based on accounts in your CRM. Okay, that kind of stuff takes a lot of manual work unless you have something like Gradial plugged in. So we do that basically for companies of any size, but basically any size where you have that kind of manual work starting to develop.

Kate 14:45
That’s fantastic. Especially a lot of companies too have a regulatory issue, right? You’ve got to make sure you have the right disclosures. You know, if you’re in financial services or pharmaceuticals, I could see where you’d be really helpful with that your software solution.

Doug Tallmadge 15:01
It’s a big use case. It was something that you know. There’s things that surprise you when you start a company, and I will say that the adoption that we were able to find from what you might kind of refer to as like regulated industries was definitely something that was very interesting and exciting for us early on. Because to your point, in some ways, the most forward-leaning companies were often ones that had that pain of the complexity of the process being even higher because they had to really think about some of those additional details like disclosures and different rules in different jurisdictions and things like that.

Kate 15:39
Yes, yes, I started my career in that. It’s a lot of work. That’s so exciting. What has been the moment, like some of in this journey moments for you that made you validate your decision to leave Starlink and do this?

Doug Tallmadge 15:55
I like to say I go to sleep easy every night because I feel like I’m learning as much as possible and kind of challenging myself, and that’s like something that I value tremendously. There were definitely a lot of periods of time where I felt that way at SpaceX. There were also periods of time where I felt a little bit more complacent, and it was something I wasn’t looking for, right. I wanted to feel like I was really learning, and nothing teaches you something like necessity, right. You become aware of the skills that you simply don’t have. The first time that you start doing cold outreach to try to reach a customer when you have zero customers and an empty calendar, that’s a unique feeling. And then you learn like okay, maybe you can get you know 3, 4, 5, 7% response rate, right? And that becomes your lifeline. And so that’s been the common thread for me is I just really enjoy that, and the the beauty is like as a company grows, the job description changes completely. You know, I might have said every three months, I might say every six months, but three to six months, you’re doing something completely different if you’re growing quickly, and that becomes you know part of the fun. It’s like how do I become what I need to to to be to kind of do my job today.

Kate 17:02
That’s so true. That that evolution is so fast. When we talk about scaling and growing, and you have this platform that helps with that, is there advice you can share with founders around scaling, just from your own challenges at Gradial, and then also around like the problems you’re solving for your clients?

Doug Tallmadge 17:23
Advice is always tricky, right? Because circumstances are always different. But I think the one of the things that I feel strongly is that you want to try to dampen the emotional roller coaster, because there are going to be amazing good days and there’s going to be really hard days. And reality is often somewhere in between. Even if you’re doing really well, you probably should be focused on what comes next, right? And this is tough because you know when you’re working hard, you want to celebrate everything. But I think the challenge is, let’s say you’re growing 4x, you know, 5x, 3x, whatever it is. If you want to maintain that growth rate, it means that you actually have to grow more in the next six months than you did in all the years before in the business. And so the challenge of the bar is always rising. I think that is something that requires you to just really try to stay as stable emotionally as possible. Where you know there’s going to be things that feel like a big setback. Honestly, in hindsight, once you get past three months, even weren’t something that you really register in the long term. And then there’s going to be things where you feel like you’re on top of the world. But reality is, you should be worried about the next hurdle that you need to overcome, and immediately get started on that, right? And you’re going to wish that you had been even more thinking about the future and less excited in the present. So, I think that’s probably one thing is like I picture kind of like a sinusoid, and it’s like okay, just try to flatten that out, you know.

Kate 18:42
Yeah, that makes sense. And you, because otherwise it could really be a roller coaster in so many ways. What about AI? I mean, you obviously are leveraging it in such an important way. Has the pace of change around AI has that been a challenge or benefit for Gradial?

Doug Tallmadge 19:01
Well, I think we were very purposeful in trying to steer the business in a way where, as AI improves, Gradial is able to do more for customers, and so it’s been a tremendous tailwind for us. Where I think we’ve been able to, you know, offer so much value because we had the harness designed for kind of what we call like a model cocktail, right? It’s like it’s not one model; it’s many models, right? Especially in marketing, where it’s like you’ve got models for doing content, you’ve got models for orchestration, you’ve got models doing a lot of different roles. Certainly, it’s interesting to think about what’s moved fast and what stayed the same. And there’s certain parts of the models that have gotten tremendously better, and there’s other parts that kind of, you know, there’s been a jagged frontier. And I mean, this is kind of an interesting example, but I can’t use AI to send emails in my own voice. That’s a really interesting bottleneck, right? It’s like if I’m going to email someone – try to send something like a respectful note, you can see and hear and read AI, right? You can taste it in the model, and of course that has implications from a marketing perspective. And so we spend a lot of time thinking about when you’re producing content, a high fraction of what you’re doing is actually reusing and remixing content, especially when you’re at very large enterprise scale. You know, for example, I’m assembling a product description page. Think about how many 1000s and 1000s of pages of words have been written about that product by humans internally to the company – I’m talking emails, I’m talking blog posts, I’m talking PowerPoint decks. And within the existing knowledge base, there is incredible context to go make that product page. And similarly, there’s language that you already have across your web or email or other marketing presence. And what I found is that oftentimes the highest quality content comes from the remixing of that strong base versus asking even the smartest models. Like even if you go to you know GPT six, Fable, Opus 5.5, there’s something about the writing that doesn’t quite get there, and it is interesting, right? And it’s a little bit puzzling, and we’ll see how long that stays true. I wouldn’t expect it to stay true forever. I would expect the models to get superhuman at writing, but I certainly think that there’re surprises like that along this journey that have been you know interesting. Who would have predicted that emails would be arguably harder to get to kind of parody versus like high complexity software. It’s very interesting.

Kate 21:42
So true. That’s such a great example. We talk about that a lot on the show. All the things that AI can do, but then there’s these core things that it hasn’t mastered yet. That is super interesting. Email and and writing content is definitely two of the big ones that come up here a lot,

Doug Tallmadge 22:00
But the other piece of that is, I also see that it’s very easy for people to get in a mindset of well, that use case doesn’t work, right? And the velocity of what is working is changing so fast that you know if you’re not retesting and rethinking about it every week, you’re not iterating enough because you know it’ll be like a flip of the switch. The ability to lip sync a video, right? Whereas, like you look at it six months ago, you might be like, “This doesn’t work at all. Boom! Suddenly, it works. And I see in the market a lot of people, even who are very very technical, over-index on the biases of what didn’t work for them three months ago because they’re not used to having to retest things so quickly.

Kate 22:45
That’s a great point. I’ve never, yeah, that’s a good point to point out. Yes, that makes a lot of sense. So overall, we’ve kind of been wrapping up the show with this lately because there’s a lot of people on the show now who there’s a lot of doom and gloom around AI. Are you still primarily positive?

Doug Tallmadge 23:04
Yeah. No, I think it’s going to be the most tremendous force for good in human history, and I don’t think it’s even going to be close. I mean, I think that there’s an interesting challenge, and you see this in so many places you know, in history, right? Not just with AI. Which is the way that our discourse as a society works, it rewards very loud and extreme positions on these issues, and so I think it’s a genuinely good thing that there are people who are saying very, very worrying things about safety, because if they didn’t say those things, no one would listen. So the reality is somewhere again. I feel like it’s kind of that same roller coaster thing where it’s like you want to put effort into AI safety and alignment. Like that is a good thing for a society to do. I wish you could do that without having to escalate straight to the rhetoric of the most extreme rhetoric that you can come up with. And I think it’s kind of in some ways, not dissimilar from something like climate change, where if you just take a 30,000 foot view and you were to time travel or something, right? So you don’t have any sort of politics or kind of rhetoric around it. If you were to just ask, is it a good idea to change the concentration of one of the core gasses in the atmosphere by two 2x go from 200 parts per million to or you know 200 to 400 I don’t remember if it’s parts per million or parts per billion, I think it’s parts per million. I think people would say that’s probably a bad idea. You know, it’s like if you’re in a fish tank, you don’t change the fish tank pH, right? You don’t mess around with the basis. But all of a sudden you end up in a spot where the only way to have a dialog on some of these issues now is the extreme positions, and otherwise people just don’t take it seriously. And I think there’s going to be very extreme narratives that are much more complex and probably less extreme, like in reality than they sound. But we’re right to take it seriously, just like we’re right to take you know some of these other issues seriously. The question is, how do we unpoliticize the issues, in my opinion.

Kate 25:03
Gosh, so well said. I appreciate you explaining that so well. You really touched on that perfectly. We’re at time. You are so fascinating to talk to. Where can listeners find more information about Gradial? As somebody in marketing, it sounds like a fabulous solution.

Doug Tallmadge 25:21
Yeah. So our website gradial.com or LinkedIn or yeah, shoot me an email.

Kate 25:25
Thank you so much. You should write a column or someday for your next thing, think about being a professor. I’ve had 150 episodes of this show, but you explain things in a really great manner.

Doug Tallmadge 25:38
Well, thank you.

Kate 25:38
You really do. Thank you for being here today, we appreciate your time.

Doug Tallmadge 25:42
Thank you. It’s great to meet you, and appreciate the conversation.

Outro 25:45
You’ve been listening to Startup Success. To make sure you don’t miss out on future episodes, subscribe to the show in your favorite podcast player. Like what you hear? Tap the number of stars you think the show deserves in Apple Podcasts. For more tools and resources for your own startup success. Check out BurklandAssociates.com. Thank you so much for listening. Until next time.