SPEAKER_00
And people weren't used to receiving emails like that, that were like from Gmail account, personally followed up, targeted, personalized, all that stuff. At one point we had a 50% reply right across like a three email chain. And because we had this super targeted product just for.
..
SPEAKER_01
Hey, I'm Cody Schneider. Welcome to the In The Pit podcast. Today we're joined by my friend Damien Tanner.
He's gonna talk to us about how he started his agency called Thoughtbot and then also how he started a company called Pushr. He's an OG in the space. He's been building products for forever.
He's gonna give a ton of insight into what you should do now if you're just like starting from zero. And just honestly brain dump everything that he's learned over his whole career. This episode is brought to you by Swell AI.
Swell AI is a content repurposing tool that helps marketing teams make content for all of the channels that they're trying to do distribution through. Go to SwellAI.com to get started for free.
All right, let's get started with today's show. Damien, what is good man? Thank you for jumping on. I know it's late there right now, but it sounds like you got a flat white, so life is good.
You're caffeinated and ready. Just for this. You're coming in with me caffeinated in the morning.
So it's like, it's big energy all around. Awesome man. Well, thanks for joining me.
I'm stoked to hear kind of about everything. I mean, we chatted in the past about the stuff you previously worked on, but it's always just, I think for everybody, incredibly valuable to talk to other founders who are also building their next thing and have done things in the past and just learn from them and see what they're doing. But yeah, for the audience, you want to give just like a 40,000-foot view of kind of how you got into this world and kind of background pieces.
I think that'd be super valuable just to create context.
SPEAKER_00
Yeah, my pleasure. Yeah, thanks for having me on. I have a technical background, been coding since I was 15.
I think my parents realized that was an obsession pretty early on.
SPEAKER_01
Did they brainwash you or was this like a natural, you were drawn to a type thing?
SPEAKER_00
I think they tried to brainwash me. I hated going to school so much. Then in the area where we lived in the UK, they were like, well, you're gonna have to go to school at some point to do your exams.
And the only school you haven't been to is the Scientologist School.
SPEAKER_01
That's amazing.
SPEAKER_00
And it was so funny because I went there and I was like, everyone's a Scientologist. Why, why, why, why, why, why? And it was, it's actually a brilliant school. It's a really cool school.
And I made a lot of good friends there. But it was pretty funny because apparently in the first week I printed off all this stuff from Wikipedia and took it into the teachers and was like, I fact-checked some of this stuff online. And I've got a few questions about it.
SPEAKER_01
But yeah, so I didn't go. It might have been a big short where he's like fact-checking the rabbi.
SPEAKER_00
Yeah, that was it. But no, it was actually. So yeah, I've always done a lot of, and that's cool thing about programming, isn't it? You can just, you don't have to go to school to do it.
You just do it.
SPEAKER_01
And no one's holding you back. It's always the best people to hear yourself taught, right? My co-founder Max is just like an unbelievable engineer. And it's the classic story.
It's like, you know, taught himself the program when he was 12, built basically bots that he sold to buy shoes, you know? Like when they would do Nike shoe drops and shit, my Yeezys. And then so he was like working with basically people to do that. So he built all this automation and then went to college, dropped out, moved to the internet.
SPEAKER_00
That's the cool thing about the internet. There's no rules.
SPEAKER_01
Mostly. 100%, 100%.
SPEAKER_00
So yeah, programming for a long time. And then went, eventually did actually end up getting into a university in the UK and then dropped out very quickly after a few months. Started with my co-founder, a web development agency, Rails development shop.
That was great fun, building a lot of stuff for all sorts of big brands. Built that up to about 30 people. And then around 2009, 10, we started building our own products.
And out of that came one product, which was called Media Core, which was a video CMS for universities. And we had a third co-founder in Canada. We built that with, and we actually ended up, within four years of founding that, we sold it to Workday.
So that was like the first big exit that we had in 2015. And that was pretty crazy. It was like good and bad timing.
SPEAKER_01
Building out of the studio. Like this is like what your team at the studio was working on.
SPEAKER_00
Yeah, essentially. We did the web development agency to product thing. And I think it's, as we found, it's like hard to do that repetitively because you need the founders, right? You need founders for companies.
But there were two founders, plus another one, and we got two products out of it that went on to be big products. So it worked. I think you have an agency with a ton more founders.
You can probably do more of them. So yeah, so that was the first big success we had. That worked out really well.
Learned a lot about SaaS. It was a SaaS product. And did some great outbound emailing there.
We should go into 2015. The heyday of, God, outbound emailing was good in 2015. I just wish I sent more emails.
SPEAKER_01
I feel like that's every marketing arbitrage, though. I'm like, even now, I'm like, I need to increase volume with all this. Anyway, so why was it so good in 2015? Like, why did you feel like that was coming?
SPEAKER_00
I just guessed the tools that we have now, like Apollo and Instantly, that do the automated outbound and follow up, they just didn't really exist. And I ended up stumbling into one of the first ones, which was called SalesBeach, and started using it. And people weren't used to receiving emails like that, that were like from Gmail account, personally followed up, targeted, personalized, all that stuff.
So at one point, we had a 50% reply right across like a three email chain. And because we had this super targeted product just for universities who had all their lectures like spread all across their like internal drives and none of the students could access the videos because they were just massive video files stored somewhere. So we'd email these chief information officers, oh, hey, we know you've got this problem, like we've got this solution here.
And so it had to be right, it was a targeted product for a problem that we already knew that they had that was on their plate, that they hadn't worked out.
SPEAKER_01
Which was the best distribution.
SPEAKER_00
But we'd send these super personal emails where like we had maybe a bunch of customers, universities, and we'd like automatically set, select other, like existing customers we have that are similar size universities and similar based on the Carnegie ranking ranking. So we wouldn't send an email to a community college saying Yale as a customer. But like, Penn State would send an email saying, oh, and by the way, we've solved this problem for Yale.
And so it would be really personalized in that sense. And it would pick ones that are around them, near them as well and a similar ranking. So we'd kind of put in those customer case studies.
And then we just get competing against
SPEAKER_01
and also the location geography for the social proof.
SPEAKER_00
Exactly, people they would know and also either like, yeah, are on the same level. Because if we found it early, if you say, hey, Yale is our best customer to a community college, they'd be like, well, we can't afford it. Why are you telling us? So, and then we just, when we started off, which is what just couldn't believe, we get these emails back from these professors after like the second follow up.
And they're like, I'm so sorry, I haven't replied to your email last week.
SPEAKER_01
You know, I think so stat.
SPEAKER_00
And then this happened, this happened, but I'm replying now and we should chat about it. We're just like, oh, wow, they don't realize this is automated. It was just beautiful.
SPEAKER_01
And that's how I feel podcast emails are right now. Like people think they're like, you're like doing them a service. Like we're cold emailing people about podcasts.
And I'm like, the open rates are like 30% with like 5% CTRs. I keep rapping on this about it because it's just like, so ridiculous.
SPEAKER_00
And I, it's gotta be that, yeah, the alignment. It's just that it's, yeah. And the target and all that stuff.
SPEAKER_01
It's gotta be that alignment.
SPEAKER_00
Cause I've been trying outbound in other areas. And I've realized, yeah, you've got it. Like we just stumbled into this thing where we knew the persona, we knew the customer, we had the solution and they needed it.
It just worked insanely well. So I've kind of, I've tasted that. So I know what's possible.
So I can, I know very quickly now, all the ways I'm, it's not working.
SPEAKER_01
Totally. You know that feeling too. I think that's a piece of this is like, you get like almost a pulse of like when it's going, right? Like once you have one of those where you're like, oh, this is what this feels like, right? It's like almost the, once you find that arbitrage again, like even if it's in a different place, you can still like, you feel that momentum and kind of that like, I don't know.
SPEAKER_00
Yeah. And then the flip side is nowadays, I'm just trying stuff in completely different segments and being like, fuck it's not working.
SPEAKER_00
Like why can't I get it working again? We did some other cool stuff as well. We included case study links and other like bits of links in the emails. And then we had retargeting set up.
And the best thing that worked was website visitors tracked with the YouTube retargeting pixel. And then we record, we went like on site and did all these like really well produced customer case studies. Nice. And we put those on as YouTube pre-rolls, but they didn't look like ads. They're just other universities talking about how much the product had helped them and how important video is as part of learning.
And then we just like flood them after they'd clicked one of the links in the emails. And when we knew that was working, when we walked into this meeting once like a month after doing that, and the guy who was this, the sales guy who was in the video is walks in the room and they're like, oh, you're James from the videos. Cause they've all just seen this ad when they're like, I don't know if they even watch YouTube, but they're like, they're kids watching a program
SPEAKER_01
in YouTube and this B2B case study keeps coming on. Totally. It's just kept pixling their entire IP, right?
SPEAKER_01
Every device on the network. That worked really well. That's super interesting.
And just to reiterate that just for people that like the audience in particular. So what he's talking about is basically like scraping these target emails and then you can both pixel the people that clicked from the email to your site, but you can also take those emails and you can do a customer match list in YouTube as well. And so then basically you're cold emailing these people while simultaneously doing display ads.
So you're basically like touch, you know, it touching them in two different ways like in creating those touch points. So that's, that's awesome. So you sold that company.
What did you do after that? Or what was your kind of the next thing?
SPEAKER_00
Then I, at the same time we had the other product we built called Pusha, which my agency co-founder Max led. And I kind of led the other one. And so we had these two, we just sold the first one really quickly kind of by accident as the case with these things.
And Pusha went on to many successes and yeah, grew to millions, millions of revenue and used by probably millions of developers. It's a real time messaging API. So you'd use it to, we had all sorts from like news websites doing live news to most of the crypto exchanges would use it for their web UIs to update the prices in the new time.
It was like a, in real time, it was like the API you used to update an application or web app in real time to stream data. And we ended up doing lots of VC for that, including some great VCs like Bulletin in the UK. And eventually we sold it in 2022 Messagebird.
So that was a wonderful experience in building a dev tool and dev brand and definitely now I'm building dev tools.
SPEAKER_01
It's like a dev dream, right?
SPEAKER_00
Well, it's also like, I'm just noticing again, like reflecting back on it. It was slow at the start. And that's definitely the thing about like devs don't just like, they do jump on some stuff, but a lot of stuff is like, oh, that's useful tool.
I'll use it when I need to do that. And when I need to do that, it could be in six months times of your time, but customers tend to stick around though, because they'll integrate it into the product and keep it there. And then in the intervening time, while Max is leading that, I went back to our original agency and led that for a bit.
Eventually we merged that into Thoughtbot, which is a great US based agency with over a hundred people. That Max and I are still co-owners with and that's still growing. And then I did all sorts of random shit in particular.
I love it. I kind of, I did stop building for a bit. I think I just was like, I've done software for 10 years, let's do something else for a bit.
Yeah. What'd you do in that time though? Turns out all the stuff is hard. I built electric motorcycles.
That's fucking hate me though.
SPEAKER_01
There's not a lot of money in that.
SPEAKER_00
Still interesting. But it was fun. I mean, it was fun.
Stuff blew up and there was smoke. That was quite fun. We had to get tons of batteries and batteries are expensive.
So we ended up having to go to like a scrap yard to find a crashed electric vehicle. And then like, we had to basically bribe a garage to take the batteries out of the car. Cause everyone we took it to was like, you can't, we won't take the batteries out of this car.
Like they're like, we're an approved dealer of this brand. We won't take it out. And eventually we're like, come on guys, just take the batteries out.
SPEAKER_01
How much did you bribe them? What did you want to have to pay?
SPEAKER_00
I don't actually know. My co-founder didn't tell me, but. I didn't ask any questions.
I didn't love it. Hopefully it was money.
SPEAKER_01
Yeah, so that was good fun. Can you talk to me about the agency side of stuff? Like how, how do you think about structuring those and building those? And like now that you own it, like do you have an operator that's a business partner? Like what does that look like? And then if you were, you know, this is the, you're starting zero to one right now and you're trying to get like your first basically 10K and monthly revenues that you can, you know, pay your mortgage or whatever. Like how would you start an agency or think about going about that? Like, you know, now in 2023.
SPEAKER_00
God, I don't know about now in 2023, but when we started our agency in 2008, I remember we had a, gosh, no, even earlier, 2007. I remember Max and I had this conversation, which was like, should we do all tech that, you know, whatever people back-end tech? And we were really into Rails. And I remember us going, you know what, let's just do Rails.
Like Rails is beautiful, clean, like enjoy to work with, we don't wanna do PHP. And that was like saying no to stuff. But in hindsight, that was then a focus and we became known as the people you go to when you want something built in Rails.
And we kind of aligned ourself and was part of that community. And the reason we were able to successfully merge, our agency was New Bamboo, it was in the UK, you know, London 30 people merged with Thoughtbot and become the UK office for Thoughtbot, which is over 100 people. They had exactly the same kind of focus and vision as well.
So I guess nowadays as well, that still applies. It's having that focus to be good at one thing or one area. And if I was doing it again.
SPEAKER_01
It's almost a wedge into the market in a way. Yeah. I'm just thinking about it. Like we just got our foothold with like, being known as like the Ruby on Rails, you know.
It's like the place to go with this.
SPEAKER_00
Yeah, and also Ruby on Rails was new. So there wasn't a 10 year experience. Yeah, there was several years.
It was mature enough. But I guess the same thing applies now. Like for example, I've been building some stuff in Framer.
I mean, Framer's big now, but you know, align yourself with an emerging community or platform like that, where there are no experts yet and you can become the expert. And because then you're gonna scale with the customer base as well, right? If the community is, if the platform is early, there's not a lot of work, but hey, you're just one person. So you don't need that much.
And then in 10 years, you can grow over time as that community grows. And you've got that, that foothold, as you said, in the market, yeah.
SPEAKER_01
I think that's the way to get it. Because then you can expand from there. I totally agree.
I think honestly, it's, I think it's a good reminder for like any early stage company. I'm just over here contemplating all of my actions over the last three months. I'm like, have they been wedged enough to like get, get that foothold? Because you can't expand into anything, but it's, I think just from a marketing standpoint too, it's like, it's way harder to communicate, you know, we do these 10 different things, right? From the, in the early stages where in the beginning, it's a lot easier to say like, hey, we do this one thing.
And then as people ask for these other services, then you can start saying yes to those and kind of build out that portfolio of services like as that time goes on. But just going to the market with that single thing that you're communicating, it just, it makes it easier for you to build, like, you know, to lay a flag in people's mind of like, oh, that company does this, right? And so you can own that mental real estate a lot easier.
SPEAKER_00
I think we got away with doing basically, you know, marketing as well for our agency because of that. Interesting. We just became, we were so early on, we just became known as K-Rail's in London, there's you guys and a few other guys, you know, one of the ex-employees ended up setting up one of the other main Rails shops.
And most business just would come by word of mouth. And it was scary because we didn't have a marketing function and we didn't even have a sales function. And then when you get to like 10, 15, 20 people and you've got all these salaries, you're kind of like, ah, and then business slows down and you're like, what are we going to do? And eventually after 10 years, you're like, just wait, it'll come back.
And you just have to trust that it does. And it always does. And, you know, I thought about, I mean, now has a lot of marketing itself, but I'd say the foundation of those brands is just built from being early and doing great work and people know who to ask when they need that.
So that's the advantage.
SPEAKER_01
Do you think geography is a factor in that? Like, do you think that that's necessary to be?
SPEAKER_00
It's gotta not be, no, not anymore. Yeah, it's gotta just not. You think that's changed? Yeah.
SPEAKER_01
What was like being in London, I mean, was that it, like from a career standpoint, was that important when you initially started all this? Like did that add to kind of the trajectory of all of it? Or and then it sounds like.
SPEAKER_00
I think it can. I think it definitely, like any of these cities, it adds to your, it just has like immersing yourself in the like liquid of that energy of that place. Yeah. Of a city, I think definitely helps. And now I live outside London in the countryside, but I've also like got different goals now.
SPEAKER_01
And that's it. I feel like it's different now too, just with the internet, it feels different. And I mean, we've, you know, met on Twitter, like a couple months ago.
SPEAKER_00
Right, yeah, it's different.
SPEAKER_01
And like have become, I don't know, just like had conversations from that. Like it's almost like a dispersed, like it's moved more to digital, like after COVID, I feel like, or previously it's like you had to be in one of the hotbeds, but now more and more, it's like, it's, you know, it's way easier to find other founders. And I feel like COVID made it more okay too, to do like cold outreach.
To other founders, you're like, yo, what are you doing? And like teach me this thing or, you know, et cetera.
SPEAKER_00
But yeah, and I feel, I have a lot of appreciation for the niche areas of Twitter, because I feel like I'm able to be a part of people doing other weird niche stuff, right? Whereas if it's just in London, you know, it's definitely part of more of the venture funded startup scene. There's definitely, you know, in big, still in, you know, easy to connect with communities there. But if you're looking for people bootstrapping stuff or just doing weird stuff, the internet's always the place where that connection happens.
SPEAKER_01
I feel like there's way more documentation of it now too. Like, I mean, I just remember historically, I'd be gone like black hat world, you know, trying to figure out like, what's the crazy case studies that like people have done over the last couple months.
SPEAKER_00
Black hat forums still have a lot of good stuff.
SPEAKER_01
Dude, I still am there probably like every other week, just like reading about, you know, reading the top posts and seeing what people say. But I mean, more and more, it's like you see that in like more public settings, right? Where it's like people are posting on a Twitter or a Reddit or like these kind of larger social media channels, which is interesting.
SPEAKER_00
But yeah, and maybe that's like the nice thing about the, one of the nice things about the algorithm has been favoring these kind of how to longer form things. Everyone's incentivized in a good way to share the how to of how they're doing stuff. So I definitely get a lot of value out of that as well.
SPEAKER_01
Totally. So what are you working on now? You said your goals kind of changed. I know you were talking about like doing like large data transfers.
So that's like, was that something that you're interested in?
SPEAKER_00
Well, I've just been excited about, okay. So after I built all sorts of random stuff and made money and lost money, I eventually kind of, 2009 team was like, actually, I do love software. I do love programming.
Let's build again. So I just got back into build mode and have been building ever since. And I guess the fun thing about programming and like starting stuff is learning.
So definitely still learning a lot. I definitely have that feeling as I get older, I learn how little I know more and more. But it's good learning that circle of, yeah, getting a grip on that circle of confidence now.
So more recently in the last few years, really started to hone down on building SaaS developer tools. And absolutely love being able to build, see what new products are possible with new platforms tech. So one of the things I've been super interested in and following a lot is the products Cloudflare have been releasing because Cloudflare are massively slashing the costs of things like data transfer with the else S3 competitor R2, which basically has no transfer costs.
The work is serverless platform, which is super, super good value. And then actually they've just released this great new AI platform, which can run a lot of these smaller AI models. So I've been looking at all sorts of ideas of things that are now possible, both new ideas, but also just like different business models, cheaper alternatives now that you've got Cloudflare to be able to use.
SPEAKER_01
So yeah. Is this a strategy that you've kind of used like throughout your whole career of like look for technology changes?
SPEAKER_00
By accident, because that's what happened with Pusha. Yeah, Pusha we created because web sockets got released in Chrome. And it was like, now you can do real time stuff in the browser.
And I remember we had a meeting, we were like, hey, what can we build with that? And we just like knocked up this idea for Pusha and then realized, you know, there's a whole business here and making it scalable, supporting all the browsers that didn't support it yet, but it was enabled by a new tech. So I think that's probably good. It's probably because I enjoy playing with new things.
And then it's nice to be able to like turn that then into a profitable SaaS. But very much also taking a different tack now, like bootstrapping feels so different after having done VC. It's like, it's- What's the difference? Whenever we had VC money, I'd say we were pretty good.
You know, none of our companies went bankrupt. I think we were pretty wise with the capital. But you, I'd say VC, it buys you the ability to make a lot of expensive mistakes, right? Yeah.
And when you make a ton, just raise more VC and make some more. And eventually given enough money and time, you'll find the business model and everything. And we did.
But I look back and think, wow, okay, we spent a lot of money on stuff which clearly wasn't working and like marketing and things like that. Man, when it's your own money, yeah, you really fear if it hits differently, those costs.
SPEAKER_01
100%. It's more like that freighter flight, or sorry, freighter flight takes over. I feel like, I don't know.
I just personally, it's a, I like it honestly because I feel like you make better decisions because it's like, again, it hits your own payout like in an entirely different way. So everybody was like, oh, I think about the investor's money, like in the same way I think about my money, but that's not how it actually is. Every funded company I've ever worked with, it's never actually that.
Like, like in-
SPEAKER_00
Yeah, when it's your bank account, even if you've got some money to put it in, it's like, oh wow, there's no more, that is it. So that makes you, and I think it's healthy. You have to think long-term.
You have to think like, you have to really watch, hey, is that channel really working? Should I put more into that or not? Definitely, definitely hits different. But I like, I'm liking that now, and I'm enjoying the kind of discipline and also the long-term thinking it kind of puts in place because there is this tendency to rush when you have VC. You know, you see it in the runways, like no matter how much money you raise, it'll be gone in 12 to 18 months.
Yeah. It's just like, it's just the law of VC.
SPEAKER_01
It's like you're supposed to sprint a marathon. Like it's five years of you just running as fast as you can basically.
SPEAKER_00
Right, right. And then actually, some things you can't speed up with money. And I think that the thing you can't speed up with money is actually the customer development and product market fit because sometimes that just takes a certain number of iterations, right? It's like you building stuff, the customer's trying it, getting the feedback, trying a different thing, building a new feature, pivoting.
Like it just takes time. And actually, even if you throw a bunch more money into it, it doesn't necessarily, like the customers aren't gonna get back to you faster. The number of iterations you have, like it's still probably need a bunch of iterations.
SPEAKER_01
The only thing you can really get is like, in my experience with throwing more cash at it is like more customer, sorry, more users that are hopefully providing more feedback. So it's like increase the end value of users. So maybe that creates like a faster feedback loop.
But like I've never figured out a way to like, how do I get people to like what you said? Like throwing money at people responding faster doesn't work, right?
SPEAKER_00
Well, and also it's like you and your team's learning curve as well I think. Like especially if you're putting a product into a market that's new to you or is a like, you're taking a different approach in an existing market that someone hasn't tried before. It's like you've got to kind of learn the feel for that market and how that market works and how the product can fit with it.
It just takes time. There's actually a good like data-driven Harvard Business Review kind of paper, which is the sales learning curve, which applies it to sales team, which is when a small early sales team is not performing, the natural inclination is higher more sales people. But if the sales team's not performing because they haven't got the serious material, right? Or the product market fit is not right, adding more people just compounds the problem.
Yeah, exactly. And actually you just need time. You need to buy yourself time to get the product market fit right and get the existing sales person process working with one person and then you scale.
SPEAKER_01
Yeah. When you were funded, like how did you communicate that with investors? Like did you like, I mean, like, yo, we just need time. We're gonna figure this out like, but we just need time.
Like did you have those conversations? No. They just like, fuck it. Like, yeah.
It wasn't even necessary.
SPEAKER_00
I feel like with VC, I mean, it's the case with any business is like, you want, you feel like you should have more sales and be further along than you are right now. Yeah. With VC, it's like, it's still the same problem. It's just like scaled up, right? Yeah, totally.
Totally. So it's different doing that, but I'm absolutely loving it. And with Rebase Ventures now, my holding company, what we're doing is building, we'll build two or three products a year, but I think on average we'll hit a good, like we'll get one product right per year.
Nice. And that's the goal is to build one SaaS, or developer tool per year, and not rush it. We're slightly constrained.
Like we're not going to build a like massive platform that's going to be tons of features. They've got to be slightly constrained ideas that like a few people can build.
SPEAKER_01
Specific products go, that's like not.
SPEAKER_00
Yeah, make a great product out of that. Yeah. And then the goal is to then layer those up, you know, as we go on and build a portfolio of products. And half.
SPEAKER_01
Are you trying to have them be like self-service and then just like layer on?
SPEAKER_00
They've kind of got to be, yeah.
SPEAKER_01
And there's some enterprise,
SPEAKER_00
but I'd say it's on demand enterprise. But you know, with the DevTools, people come wanting to buy it and they'll, kicking and screaming goes through the enterprise sales process, but we're not out there pushing them yet. I'm sure as we scale up and they mature, we'll look at getting sales team, but unlike with VC, we're not getting ahead of ourselves.
No, 100%. We'll reinvest pretty much everything.
SPEAKER_01
Yeah, no, that makes total sense. So you're building these new things, are you building them a lot on this technology you're seeing with Cloudflare and that type of deal? Or is it like, what categories?
SPEAKER_00
Yeah, so we've got a, exactly. We've got a-
SPEAKER_01
We're talking about them publicly too, sorry. I don't want to pry too much because I don't know what you're saying.
SPEAKER_00
We've got two, we've got public. One, which is UCSV, which is developer tools for adding CSV import, like a drop-in, fully fledged CSV import process into your app. That's the first product we built.
It's going pretty well. Yeah, definitely self-serve. And really happy with that product.
Super solid. We've got enterprise customers, importing millions of rows in legal processing, healthcare, places like that. Really happy with that product.
People integrate it. There's no need to ever take it out when it's working and it's good. But again, slow.
You don't integrate CSV import unless you definitely need it. You don't do it because it could be interesting or fun, right?
SPEAKER_01
Yeah. And then you try 20 different CSV imports and pick the one you like the most.
SPEAKER_00
Exactly. And then we've also released, we're, I want to niche it down into specific industry, but we've got one of six of the, I've counted six competitors. There's probably like actually 30 chat GPT chat bot trained on your website content.
Yeah. But I think now everyone's like, got to find these niches, right? And actually apply it. So I'd love to find an industry to niche that down into.
Totally. And as is the case with the DevTools, I built the CSV import thing because CSV import pissed me off in another product we built. And actually when we were building the chat GPT stuff, scraping websites, it has always been a pain in the ass, but I felt that firsthand.
And so actually the second dev product that we're halfway through building is a web scraper. Super cool. And I think with the help of Cloudflare, we can make it way faster, way cheaper.
Oh, super interesting. And I think super easy as well. So I think we're going to be able to do it.
Some really cool stuff with chat GPT. I've seen like one or two that do this, but I think we can combine the power of like a fully kind of, you know, API driven and technical, like fully featured scraper, but also with the ability to just create your scraping and automation just with a prompt. Super interesting.
That's going to be super useful in, yeah, marketing sales, but also one of the other main use cases now. I want that. Is chat GPT like and AI agents and things like that.
Yeah. A lot of the time you want to get results from the web and turn them into markdown or some other kind of data and have that done really quickly. Totally.
So I think.
SPEAKER_01
Are you guys experimenting with agents like pretty deeply? I've been more and more looking at these, what was it called? Like chat dev. I just saw this demo. I haven't played with that.
Dude, it's crazy. They basically, there's a paper that was released. I think it was two weeks ago, but it was basically talking about how like agents perform best at completing tasks when there's like multiple agents with specific designated roles and they work together.
So they have like a directory of agents and then they kind of work as a team, right? It's and so they basically showed that like the performance output would like increase. It was just, it was way better. So chat dev, I believe is what it's called.
I'll look it up and out in the show notes, the actual, but basically how it functions is like you, you can create like a team of agents. Like for example, like I can have a social media marketing org and it's like, here's this marketing director that makes proposals and then like there's an actual like specialist that goes and does this. And then they like the analysts analyze the content to see how it's doing.
And then they're like, you know, actually publishing it to your, to your social media profiles. Like all you're providing them is like the raw materials that are just, you know, imagine it's like, we take this podcast and we give it to them and like a video format and then they can just like, you know, pull out the segments and their job is like, get the most views on videos, right? And you can like designate that, that basically that goal. And then it like, they have this word just like runs, right? And so initially the, the first things that they're, people are starting to build are like, like little software programs, like for example, they made like Snake and Tic Tac Toe and like, you know, these other little like almost like games, right? But it was like this entirely autonomous, like Dev shop that basically like built out this whole thing.
And you can see them like go through the iteration process where like Q and A is sending back to like the actual like engineers like, Hey, here's, here's problems. You need to fix this. And it like goes through four iterations to like fix the actual thing.
That's wild. It's crazy, right? And like, I saw my, I don't know, the marketer in me is like, I'm imagining, okay, well, what if I have like a social media agency group? And I think with all this, it's like, it's all still crazy hype. But like the fact that we're even at the like toddler version of this, like the kind of early beginnings is to me super interesting, because I especially for startups, right? Like, you're talking about bootstrappings, like, Oh, imagine I can like hire an SEO firm.
And like, I can hire a social media marketing company, and I can hire a paid ads company, right? And all the like, they're just agents, like it's just like a group of agents that basically work in this way that run my entire marketing department, like based off of these defined goals that they're trying to accomplish. And so anyways, just brand of random stuff.
SPEAKER_00
But there's a back to the agencies. There's got to be a great opportunity for someone to form an agency around that because I've chatted with a bunch of SEO agencies about working together. And that is kind of like, we do, but we don't use AI.
SPEAKER_01
They all say they are. And then they all are is what I'm saying.
SPEAKER_00
They kind of are. They say they are. But I just want an agency which is like, we use AI and therefore it's like 10 times more or this much cheaper.
Yeah. And that's why it's awesome. Yeah. Yeah.
SPEAKER_01
There's definitely something that I'm sure like they're in, you know, they're, you know, publicly saying, hey, we're powered by AI and like, that's why we can give you 30k worth of service for three grand, you know, because all the existing
SPEAKER_00
agencies, they're too, yeah, they've got too much to lose. Yep. So they're ripe to be. Yeah.
SPEAKER_01
Well, that's the cannibalized themselves, right? Like it's like, if you know, no one wants to do that. Totally. If you're charging $100 per article and now you can get them out the door for seven cents.
Like, like, of course you're not going to just pretend it's not
SPEAKER_00
there until it's too late.
SPEAKER_01
I was just like, you're no evil, see no evil, the whole thing. So I mean, but that's the reason they're going out of business. But I think that for like people that are trying to start, I mean, we were talking about this a little bit earlier, like, how would you start an agency now? Like, I think those are the things that are huge opportunities like Pika Niche, like even clip production for like fire, you know, people that are doing YouTube videos or fireside chats or podcasts.
Yeah. Right. And all you do is just like, you know, unlimited clip production for them. And it doesn't even have to do like be animation.
Like it can just be like, we find viral clips and pull them out for you. But I mean, that, that can be entirely AI powered at this point. And, and you can build.
SPEAKER_00
I wonder if the, the, the part of the, the value of doing it wrapped in an agency that's not obvious is the trust and the brand of the agency. Because most people that hire an agency, it's not just to get the task done. It's that they don't know how to do it.
Like they don't even know, like, you know, if you, if you go out and you know nothing about video, you hire a video agency. It's because you actually have no idea what is a good video and what is a bad video. Totally.
You don't, not even at that level. So you've got to trust someone else. And so even with these powerful tools that'll do it for you, if you don't know how that world works, whether it's marketing or making clips, the tools that do it, the AI that does it magically actually might feel too kind of like, like scary.
It's like this machine that'll do it for you. But how do I, like that's terrifying. They're actually having just a brand in front of that.
SPEAKER_00
Who can judge it and control the machine and make sure the output's good. I think that that's them. That's the margin for the agency.
Cause that's all an agency does. It's just, it's just a salary arbitrage, right? With people right now. It's like, people want, people want reliable jobs and agency will pay you a salary and then the agency takes the risks of the work coming in and takes the margin and it's kind of wild how big the margin can, can be.
And so now with AI doing a bunch of the work, there's going to be an even bigger margin there and or a lot of competition. If you can lower the price by using AI and compete against the legacy agencies.
SPEAKER_01
I totally agree. I totally agree. I think it's just like, it's changed kind of that whole model.
And I mean, again, it's, I feel like it's with every kind of big technology change, there's opportunities that come up. Like what you're talking about with Cloudflare, like, you know, all of this LLM stuff is the same thing. Like mobile was one of them as well.
I mean, everybody's saying the exact same. Like there's no revelation that I'm talking about, but I think it's like.
SPEAKER_00
The thing that has slightly surprised me in AI is I only have recently appreciated how much of a consumer driven phenomenon it is. Okay. I, you know, my head's really always been in B2B and having now been in the B2B world a bit with AI kind of realized actually, you know, there is so much talk about how businesses need to use it and everyone in the business will say, yes, we're working on how to use it or researching. But actually there is many, many reasons why they can't use it right now, whether it's at the level of their employees being out having authorization to use it or even just using AI in any part in their product.
And so all of this usage is really consumer driven and all of that successes, whether it's chat, GPD or, you know, the surprising success is like character AI. Right. I was just about to say this.
SPEAKER_01
I, okay. Their subreddit has like 700,000 subs and like it was on page one of Reddit yesterday because it went down. I'm old.
SPEAKER_00
I don't understand. It's crazy.
SPEAKER_01
You know what I think it is? Okay. So I think it's, I think it's the people that like are obsessed with fandoms and they go there and they get to chat with their favorite characters. So it's like, oh, you're obsessed with XYZ fandom, like Wheel of Time or something, whatever.
Right. And it's like, you want to go chat with like your favorite character from that, from that series. That's what I think is drive.
Like personally, that's what I think is driving this. Cause it's like, I mean, people are obsessed, right? Like it's like to, I had, I had no idea that it was like, but it kind of makes sense though. It's like, oh, I admire, you know, whatever, like I want to talk to Hermione and like I read fanfic about Hermione and, you know, or whatever.
And like, this is this way that I get to interact with this character and I'm holding it in. I feel like that's like this reason, but I don't know. I'm curious to see it.
It comes into my friends working on this company. Actually right now it's for, it's like for HR. She just raised and, but basically it's like, we create like a HR agent that like you can interact with and have conversations with to help you like navigate situations that are happening like within the organization.
And so it's almost like this like counselor within the org, right? And I was like, that's crazy. And like, imagine that for every department, like, you know, the customer service people, like they have a, you know, they have somebody they can enter or a character they can interact with. That's like the, you know, the North Star goal of like what you're trying to basically portray like externally.
So it's almost like a coach that, that exists like within the organization. So anyways, I think it'll be interesting to see how that kind of gets adopted, you know, within these companies, but that I've been shocked by character AI. I'm glad you brought it up.
Honestly.
SPEAKER_00
Like, that's the, that's the, yeah, the growth, the beach head and the businesses are so far behind. I just kind of had to remind myself that the other day I was like, man, okay, like that we haven't really even started the business rollout of this stuff yet. And, and the open source models and platforms like Cloudflare now, workers AI, which makes easy to run the open source models, that'll hopefully remove a bunch of kind of barriers that businesses have against using the, the hosted centralized AI's.
But yeah, there is so much to happen. So there's, I'm excited about that. There's going to be a lot of exciting B2B companies that are being built and will be built.
And, you know, once as businesses get more comfortable buying AI stuff, whether it's via an agency or directly SAS with that's built, you know, in front of the models, there's, there's actually a lot to happen there. We're like coming off that top of the hype cycle into like, okay, now we're going to actually deploy the stuff. And, you know, if it's agents, you know, there's, there's a hell of a lot of work just turning that into a reliable thing.
But once you get all those details right and all the edge cases and get it working, it's going to be hugely valuable.
SPEAKER_01
I totally agree. So question for you. I'm just thinking about this with the, this like the chat bot that's trained off of the website.
I'm thinking about this like in relationship to knowledge base. Like, okay, so have you seen intercoms fin? Like they just launched this. I don't want to say it's too like, so Chris, like chat, which a ton of these indie devs use, like doesn't have a version that's as competent.
Like they have something that's similar, but it's just like quality is not what I think it like. Like in comparison to what intercom has put out, like I'm just imagining like you could just go to built with, you find every, you know, company that like is using Chris on their website and then cold email them, right? Or like even just cold message them of like, Hey, you can like add a chat bot like in minutes. Like, and it's, you know, you just basically like, I think they have an application like an app store that you could basically like deploy through.
So anyway, just a random thought for you on that, but I think something like that.
SPEAKER_00
I like that idea. I'm slightly, cause I have emailed with Chris, but yeah, that this is the this is the challenge with this. A lot of this stuff right now is how much is it going to be a feature in existing products? Yeah, totally.
Where the existing distribution you have and customer relationships you have is your advantage versus how much is going to be in a new, in a completely new area. And I guess like coming back to the character AI thing, like that's a kind of, I mean, it's a consumer product. It's a very standalone, unique experience that no one's doing.
And this HR examples, maybe a great example where it's like, it might be too risky for existing HR companies to offer something like that. It seems a bit too scary for them to do that. Cause what if it's, what if it's wrong or it, whereas as a startup, maybe that's, that's kind of one of your angles is you can, you've got less to lose, right? So you can do stuff that others.
So I've seen that in the chatbot world. Yeah. Is it a feature? Is it a new product?
SPEAKER_01
I wonder if you can augment search like e-commerce search too.
SPEAKER_00
Cause like, well, this is, I want to get into, I want to niche down. So I've been chatting with customers in the various kind of segments we've got and trying to work out. Yeah. Let's just focus. I think I want to focus on one industry.
Yeah. And build out additional features for that industry. So if it's like a travel, let's have features where actually the chatbot can access the search system and the booking system and just like guide you through the whole thing.
And that would be specific to travel e-commerce is exactly the same. Although in the case Shopify has got some cool stuff coming out.
SPEAKER_01
Yeah. I feel like they're going to deploy like their own version of it at some
SPEAKER_00
point. But there's other little worlds that, you know, industries, corners of the world where there is not a Shopify 100% out there and their websites in WordPress and other random stuff called together. And actually we can build something that integrates all that and works for their industry.
Also for the case of the cold emailing too, you know, like the cold emailing works really well when you've got a specific solution, specific industry and you know, their problems. Otherwise, I mean, you see, it's the same in agencies, right? How many of us get a thing? Hey, we're an agency that does everything. Like that cold email is never applied to.
SPEAKER_01
No, 100%. But if you like solve this specific problem, it's like, that's when you actually get that feedback. And we can, yeah.
SPEAKER_00
And like in the early days and in media core, it's, and we've done it for these customers that you probably know. Yeah. And then they get that feeling, hey, hey, if they're doing it, we need to keep up and do it too. No, that makes total sense.
SPEAKER_01
Man, this is great. Thank you. I'm a gear turning just thinking about a ton of this.
So where, where people find you if they want to ask you questions and then any, any final like thoughts, like I'm just like, here's the tactic, like here's tactical things that I'm seeing work right now or like I'm excited about on the marketing and distribution side that you'd want to pass on.
SPEAKER_00
People can find me DC Tana Twitter. Cool. And Tana.me. I think I've probably got an email address on there as well. God, final, final thoughts.
I just the general feeling I, I really think I'm noticing now as I build things is like building costs, building products for customers that you really want to serve and you enjoy working with. Yeah. I think that is, that is so important. It just, that's what being able to build stuff and enjoy that experience.
And then having built stuff for customers where it's been like uphill or like they're in like super stressful industries. I just realized like actually choosing the customer first is the important thing. That's why I like doing DevTools.
Devs are the best customers. Marketing distribution, man, just, it is, you just in the pit, isn't it? Because the, the stuff you do that works for one product doesn't work for the
SPEAKER_00
So you just got to try it all again. Yep. And I definitely feel like the places where I find something like that, that gets me kind of some cheaper, cheaper channel or traction is like finding the weirdest stuff that's going on and then just taking that little bit further that someone's not talking about. You pointed out uploading customer lists as targeted audiences.
It's like not everyone will, we'll talk about that stuff. But yeah, being able to get a massive email list of your target customer and upload that as a customer audience, targeted audience in Twitter or Twitter.
SPEAKER_01
And it won't match, it won't match all of them, right? Like I, I, no, but something you're not,
SPEAKER_00
Pissing money into the wind in a random to a random people. You're like, okay, well, at least I'm spending money on people who are so cool, who at some point are probably going to become a customer. Totally.
So I love stuff like that being able to like actually, that's how we did.
SPEAKER_01
Well, I just like a ton of podcast emails. And then we just did a customer match list because like there was no way to target them and we couldn't figure out like the other distribution.
SPEAKER_00
It's messy, right? It's like scraping, like cleaning up data and then uploading places. And like, I, that's when you were like talking about like a promptable,
SPEAKER_01
like scraping application. Like the, the marketer in me is like, here is $100 a month. I will pay you right now.
I've got so many cases.
SPEAKER_00
I want to use it in as well. Like just being able to say, Hey, yeah, go here, take out all the records and then press next page until you're done. Yeah. I'm really looking forward to being able to get that working in the new product. Totally.
SPEAKER_01
Cause there's versions of it, but it's garbage, right? Like I've used like web pilot for this as an example. And it just like is not consistent. Like in it like also doesn't want to do what you want cause it like doesn't really want to scrape.
SPEAKER_00
There's, there's a few new stuff like functions and stuff in GPT, which make it possible. Yeah.
SPEAKER_01
To do it way more effectively. Yeah.
SPEAKER_00
And, and reliably. Yeah. Cool. So that's what I'm, that's what I'm working on. Love to hear about any, any, uh, people who are, yeah, dealing with hairy web scraping problems.
But, um, I think that's, that's a great, that's a great thing to end on, which is, uh, if, if the marketing you're doing feels messy, you're in the right place.
SPEAKER_01
Totally agree. Totally agree.
SPEAKER_00
All the easy clean channels, they're all diluted. You've got to be in the, the, the messy place.
SPEAKER_01
There's the clip right there. So awesome. Man, I appreciate it.
Thank you for your time. Likewise.
SPEAKER_00
That's been fun.
SPEAKER_01
We'll have you on in like six months. They learn more about like kind of what you found with all this new tech that you're building and stuff. I can't wait.
SPEAKER_00
So I'm all learning about what I've, what I don't know. Yeah. What I've learned. Awesome.
SPEAKER_01
Thanks again. We'll talk to you soon. Cheers.