69: Sound, Actionable Advice with Jason Cohen

Hey everybody. Normally I would be starting off this podcast by welcoming Derek, but instead I have a guest. Derek has left me all alone to fly solo, but fortunately I have a co-pilot today and it is Mr. Jason Cohen. Hey Jason.

Yeah, you can't you people can't see this, but Jason is chugging from a large bottle of bourbon. Regularly.

Yeah, right. Exactly. Now it's just a bottle. So in just in case you have not heard of Jason, he is a four-time entrepreneur. He's done bootstrapping, he's done fundraising. Currently he is the CTO of WP Engine, which you started off bootstrapping and then raised money for and now is enormous like 700 employees or something.

Okay. So a relatively small enterprise.

That's true, yes, comparatively. So we'll go back and edit this.

100%.

Well, fortunately you know it. Although even when you know it, it's still difficult to ignore it.

Right. Yeah, that'd be great. So I saw you on Justin's podcast and live stream extravaganza and I really liked what you were saying. I appreciate your sort of semi-contrarian no BS kind of tell it like it is attitude, I guess. And also I didn't realize that you did podcast. So I was like, all right, I got to get him over here.

Yeah, that's fair. Yeah, not so much contrarian, but... by saying it's not semi-contrarian. Okay, never mind. Yeah, right. You backed yourself into a tautology, anti-tautology. I appreciate anyone that will talk straight with people. And I got the sense that like Justin was kind of hoping you would say, 'Yeah, you can work on this business about 10 hours a week and that'll be fine. Like, don't worry about it too much.' And you were like, 'Nope. No way. Sorry.'

Have you seen Paul Graham give advice to startups?

So, the reason I brought him up in particular is because I feel like he has a like 0% buffer before he's like, 'Stop talking. I have another question for you.' kind of thing. Like the questions are great and his suggestions, at least for the type of companies he advises, I think are great. It's like if you had to condense everything down into 10 minutes, he's just dispensing with all of the sort of pleasantries, subtleties, all that normal social stuff.

Yeah. And especially if you're advising like 100 companies or whatever they were doing. I have to imagine you really need to do that. You touched on this thing that I wanted to ask you about, which is advice that's any good starts with like an hour of the advisor asking you questions about what your goals are.

And did you come to that conclusion because you had bad advice in the past that wasn't tailored for you?

Right. Yeah, it might still be bad.

Totally.

I like this actually point here of even if the advice were perfect for you, like amazing advice period and perfect for you, even then that's not really enough. Like all you know is like a blueprint at this point, but everything after this is execution, which is the hard part.

Yeah.

Right? Like so the same thing except we have less skill and there's more luck.

Only like one or two because I don't have any time and I certainly haven't done any new investments in a long time because WP Engine has taken all the time and I don't want to dilute time and I am clear that I want to do WP Engine and I want to be at home. And since I'm clear on those two things, I don't do other stuff. I used to have other hobbies and I used to mess with companies at Capital Factory more and now I don't because I'm clear on that.

Yes, especially in the early days of Capital Factory, which is sort of like Austin's version of what started out as Austin's version of Y Combinator when both were small, when it was like a class of five or 10 for a summer. That's what we did, too. Now, Capital Factory is hundreds of companies. Some are co-working, some are inside of a funding program. There's a whole bunch of stuff going on with it. But early on, there were many startups that came through, some of which were in a program, so they were, you know, kind of intense advisory and investment. And others sort of just there. And so by observation, you can see some things. And just by talking to a lot of founders, you sort of get a certain sense of different kinds of personalities and what things and you kind of see what things work. None of that's very scientific, obviously, but you do get a gut feel of stuff and that's where this comes from. I don't have data that shows you should ask the following questions, right? This is just experience in trying to formulate something.

Yeah, so what would happen is they would say, 'Well, here's our situation.' And let's say especially in the context of an incubator, you're usually talking about raising money because that's what incubators sort of gear you to, most of them, not all. A lot of times the idea wasn't very big or the traction wasn't very good. The story wasn't really one that would excite an investor. And so I would say that. I'd say, 'Look, you know, you're saying these things and an investor's turned off when you say that because, you know, whatever the reason is.' You emphasize the team, but this team is not very investable because you don't have previous experience. Now, if you were just bootstrapping, who cares? Don't even have a team slide. In fact, don't have a slide. You get back to work. What are you doing? Yeah. The very fact that you're pitching means you're pitching to an investor. Like, I don't know why else you'd pitch. And then this team slide is bad because these are not the qualities that an investor is looking for. But there are different qualities we can emphasize that are what the investor's looking for. And you could be more honest going back to what you just said about appreciating being honest and straightforward. Just be honest and say, 'Look, this isn't the normal team that you see.' So, then how would you overcome that with an investor? If the team isn't normally investable because you don't have prior successes, blah blah blah, what would you do? Well, you could say, 'Well, we could prove that we are a learning machine. We could prove that every month we learn stuff, change it, and the business gets materially better, and we never stop doing that.' In other words, we could overcome the fact that we don't have previous success to point out by saying, 'We are such an engine for building a great company that we know how to do this.' And then you just really emphasize it. Have proof points. Show, 'Look, we saw this, so we changed that, and then this was the result. We saw this other thing, we changed this, and this result was negative, so we did this, and then it was positive.' And then you tell these stories, and no one cares what your bio is because who cares? You're able to do this, and that's obvious. So, let's deemphasize the team and emphasize these other things. Now, all this though is talking to someone in the sense that they want to go raise money. But then sometimes at the end of all that they go, 'But you know, honestly, I don't care about impressing that person.' I'm like, 'Well, what? You don't?' And they're like, 'No.' 'Why not?' 'Because I don't want to raise money. I don't want anyone telling me what to do.' 'Well, then why are you in a pitch practice?' 'Oh, they just said it would be good practice.' But see, so the whole thing is like, 'Wait a minute, the goals were different. So, none of this advice was any good.'

Mhm.

You're constantly failing. You're constantly not doing the right thing, and you have no resources at all. You don't have brand recognition, you don't have money, you don't have a big customer base to motivate you. Your social media has four followers, like you don't have enough time. You have nothing, right? So, one of the few things that you have when you're little is you can change directions really fast. Because you don't have any of those things like customers, you can make changes. You can change the product or the brand or how you sell or the pricing or what features are there or delete features. You have the ability to move really fast. So, you should do that. Like you should use the advantage that you have that your bigger competitors cannot do. One of the few things you can do. Another one is being really human. When you call a big company, every time you call them you get a different person. That's the nature of it. You're not when you're little. You should use that. You should be on the home page and you should be on social and so that people have a personal connection. That doesn't scale. That's okay. Who cares? Maybe you never want to scale, so that's fine. Or B, you do, but like first you have to get to the point where scaling is a problem. So, move in ways that a big company cannot do. So, these are some of the few advantages you have. So, to me like you must take advantage of them since you have so few advantages when small. You got it. So, one of them is this agility that you just mentioned, which is to learn fast and react to it. A big company can't even if a big company learns fast, it can't react fast. And also it probably doesn't learn fast. And so, be a learning machine that's always getting better. And that is a good recipe for since most of what you're doing isn't quite right yet by definition.

So, the issue with that and I've done it too. I did it at Smart Beer for sure. So, it's not like I'm not guilty of these things.

The problem is that you cannot in fact act like a bigger company. You will not deliver like a larger company will deliver. And that will come out. And then it's embarrassing. Because then it's clear that you've kind of lied. The usual reason why you say 'we' and act like it's huge is like but see this way people will trust us and become customers of us. And what will happen is though they expect 'we' to answer the phone and emails all the time. Because 'we' is enough people to do that. And that when they do then they talk to you they talk to different people because there's 'we'. They talk to the same person all the time it's not 'we' and they detect that. So in other words you're doing it to build trust that you're big but you've done exactly the opposite. You've destroyed trust because you lied and it's going to come out when they interact with you. How come you barely ever release features and barely answer the email? Well because there's one of me. Now on the other hand if you're honest in the first place there's one of me people are actually happy with it. Why the email take so long? Because it's just me. Oh cool. Like it's just the reverse it's like that's awesome. I'm so glad I'm supporting an independent person. That's badass.

Oh the server went down. Like yeah because it's just me. Like oh that's okay man. And that happened to us at WP Engine. Early on when our servers went down. They were like hey I know you're trying something brand new that's really cool man. Now if one server out of 7,000 goes down for you know 4 seconds there's a Twitter. Like hey what that's not what I'm paying for. And they're not wrong it's just it's a totally different thing. Again take advantage of the thing you can take advantage of which is that humanity it's great. And you can't have servers up all the time when you're one person or whatever the equivalent is for your product right? So don't set the expectation and then actually you get a lot of benefit of the doubt. You want to raise prices? As a big company there's a lot of well and you know why you raising prices? Because you just want to pad the bottom line or whatever. When you're one person and you send out that going hey everybody I'm just me. I'm just trying to make ends meet. I need to raise prices a little bit more so I can hire one person. What does everybody say? 'That's awesome. Great.' What? Great that I raised prices? You know, so I mean, talk about an advantage.

Right. It's great, and you get these loyal people because by supporting you, they're kind of part of it, and they are.

Yeah. Someday, it won't be there. Well, maybe not. Maybe it'll stay there forever. It'll be small. Like 37signals have kept it forever. So, I don't want to say it will never end. It may not end. But or maybe it will. It can end. That can happen.

Sure, but you want to take advantage of it while you can.

And they're even part of it by supporting the business.

See, yeah. So cool. You can't buy that. Like Google cannot spend any amount of money on ads to make that happen for Google, right? It's cool. Like you have something even Google can't do. That's amazing. So, do it. So, don't hide behind a 'we' where you just threw away that advantage.

It depends a lot on the decision. Like a hiring decision, product strategy decision, should you move? Like there's pretty different filters you would then apply for those things. So, I don't know about favorite filters, but I would say you can't have too many, which is nice. Like you can have too many priorities. Well, this is important and so is this and so is this and so is this. And then you can't really make a decision because these options are going to have different peaks of what things they're good at and what they're not and then you're confused. So, more priorities, identifying more important things actually makes it harder to make the decision. Filters, which are thresholds or negatives, make it easier because they cut out things until there's not many options left, which is good because that's helping you focus on making a decision. So, you can't have too many. So, that's good. You can have 20, it's all right. But so like if you're hiring for someone, are there some of them based on culture, others based on things that they've accomplished or things that other people say about them when you go around. Like there's a lot of things you could sort of pile on there that you're... you know.

That's right. Yes, constraints are useful for making decisions and you know, especially bootstrapped entrepreneurs don't like constraints. It's the whole idea. I don't want a job. I don't want this. I can be my own boss and do whatever I want. And then you of course you realize very quickly like this doesn't feel like doing anything I want. This feels like I'm being crushed all the time. About to die. This does not feel like freedom at all. And then you say things like, but at least it's mine, which is true, but you're still being crushed. And then later on it's still crushing because you know, you have employees and you really feel like you're still serving them. They're all people that either have expectations or you know, you want to provide for them and so forth and so it's still crushing and because so are the customers because they give you money and of course you're trying to make them happy. Like this whole thing if you don't have a boss and in a sense it's like no kind of everything is a constraint and coming back on you. It's just not exactly what you thought it was. You know, sometimes these things sound so trite. There are two things or you know, you can like Peter Thiel says you can only focus on one thing at a time or whatever you know, all these little rules. And like now, they're probably all sort of extreme in a way that makes them a bit ridiculous. The reason why they're kind of useful anyway is exactly that it reduces options. And that is good because you sit there and you look at the blank screen and think what should I do? What should I work on? Should I write another blog post? Should I write this feature? Which feature should I do? Constraints like let's say you arbitrarily said here's my schedule for the day. From here to here I write code and then from here to here I do this like well, you don't have to restrain yourself like that. You can do anything you want with the whole day. Yeah, I know, but if I do that, it'll make sure that I'll do some of this and some of that and overall globally I want to do that. So although it's artificial, I'm going to do it anyway. Same thing with a diet. Which diet is right? I don't know. Haven't we learned that that's not just not the right question anymore? But it's also probably true that anything that says well, you can have this but not this and it's sort of arbitrary or whatever. Yeah, but if that just means you eat less and more healthily, it's probably good. Right? So yes, but the constraint is helpful because it helps you make a decision and stop thinking about it and whatever it is is probably okay. I feel that way about a lot of frameworks in... This is a framework for product strategy. Here's another framework for figuring out how to do... Like it's not really about what's the best framework or proving that this framework is right or there's four things that this... Well, not three, not five. Sure, maybe. Like it's not really about the... It's just having any framework just kind of stops a certain part of the conversation. You start filling in boxes instead of being stuck with too many ways to think about it. And that's why it's helpful. And so I do like frameworks and I do like some of these sort of truly arbitrary rules that self-impose arbitrary rules. I like that because it just lets you get to some answers.

Get to work, yeah.

There's a part I want to quote from a blog post you wrote, which is kind of getting a little bit back to this advice like what's good advice and how do you recognize it and all that and I'm going to quote you which says, 'As you accumulate a set of principles that you have a lot of energy for, you throw yourself into it and run to the extreme. Doing that is perhaps one of the truly universal rules of success that you are all in. Applying an extreme amount of energy is more important than exactly which path you're taking.'

That's good. It's pretty good. You can print that.

That's amazing.

Totally. Yeah, I've had that same experience.

Mhm. But it's way less scary to add the feature. I already know how to do it.

I'm probably your target market in that rant right there. So, I'm a tactical person. I know how to write code pretty well, but I'm in the CEO role for this, and I'm on sales and marketing duty. And try to make sure that actually sales actually happen. Assuming that I've drunk the Kool-Aid, which I have, I agree I shouldn't be writing code. That's not going to help the business that much. And I want to get better at this stuff in a hurry. Is there anything beyond just like build the learning machine, try to get awesome, pay attention to what works, try lots of stuff, see what works for you? Is it like everything else where you just learn it by doing and try to figure it out?

Can I describe where we're at and just have you like ask me questions and do a health check? I'm like I'm at the doctor's and you're just like, 'Let's check this and make sure this is okay.'

All right, so we started this company back in May. It's me and two co-founders. We're all technical. It's called Tuple. And our goal is to build an app that lets you do remote pair programming. So, two people on one computer writing code together. We think all the solutions out there are pretty crappy because you need to make something that's really, really fast, screamingly fast. And the existing things out there have too much latency and people don't like them. I started off by doing pre-sales. So, I reached out to a bunch of people. I have a bit of an audience. I came to this with an audience, which is nice. I've been working in public for a long time. So, I had a bit of a Twitter following and some email people and a podcast. And through that, started getting people on an email list and actually populated a paid alpha. So, it's pre-sold about $8,000 worth of people on that. So, we have a handful of teams and a handful of individuals who have said, 'Yeah, I want to try it. I want to be your first customers.' We're hoping to launch that alpha early next year. Get the first people in there. We have a bit of a like minimal viable versus smallest lovable product thing where if our whole pitch is like it's going to be really fast. And we know that's the most important thing, but it's really hard to make it really fast. So, we're kind of racing the clock to make it fast enough by the time that we want to hit.

It integrates with something?

So, is it like desktop sharing but you're super smart about diffs, sending diffs and compressing that and I don't know, things?

That's nice.

So, first of all, I'm kind of hung up on this thing of like, if it's not really really really good, then the business isn't going to work. I do see that in the long run. Like, why would someone love, like not just like, but love the tool? Like, latency that's so low that you don't even know what's going on? Totally is that, especially with developers. So, like, 'Well, how does it work?' You know, like, I get it. That's I totally agree. But, this whole thing of like, well, we launched in May, except it's still in alpha and we're as far away it's still going to be hard to get the stuff we want and yet everyone's asking for it. So, to me that sounds like a contradiction. You're saying the customers are saying, 'I want to pair program with it.' And you're saying, 'No, you don't understand. No, you don't.' Cuz it's not a little bit faster. So, you know, one thing to consider is they may be wrong because, like you said, you may be promised a certain thing and that's what they're thinking and they won't get it and they'll be upset. So, they may be incorrect. But, what if they're correct? What if you're right about the in the long run about having the best product? But, what if you're wrong about what people will pay for today, pay for now? So, for example, if I'm used to a workplace that has pair programming and on the other hand, the engineers in the building say we'd really like to work from home sometimes, but there's no way to make those two things work. Pair programming and working from home is hard to do simultaneously. So, here's Tuple, a way to do that. And they're like, 'Oh my god, I could transform the way that I work.' Except you're saying because the latency is 100 milliseconds, not 50 milliseconds, I'm not allowed to do that. Dude, I will take a slightly higher latency so I can work from home. Do you understand the transformation that is? The difference between not having to commute at all and having a slower than it will be Tuple changes my life. And the other part makes me go from Tuple is required to and also I love it even more.

Why would you stop one until you have the other? Bad. No, no, no, let him do it. So, I don't know which one's right because, you know, maybe you're right. Like the experience is so bad right now that actually they would be disappointed. I don't know. But my guess is usually creators are extremely self-critical.

If it's not all the things that you imagine, it's just not good enough and you're in fact not even proud of it. And you certainly feel bad charging money for it cuz not that good. But this is the wrong mentality for a new product. It's the right mentality for a new mature product and maybe even a mature company who can't really have a crappy product cuz it would be a Like Apple has to make a product that's all the way done even if it's new cuz it'd be contraband otherwise. Fine, but none of these things are relevant to you or me for that matter, right? There are those phrases like if you're not embarrassed by your first version, then you released too late and all that kind of stuff. And you know that cuz you said SLP and all that stuff. So, you know all that. But still maybe you're doing it anyways. Maybe allowing all those people to do it and charging them for it even would work. And they would say yes. And then look, your company would have a lot more revenue, which would be great. It would help you do everything. Revenue solves all problems. And so, you know, maybe maybe they're wrong cuz they don't understand how bad the experience is, but maybe they're right. Maybe they don't even need link sharing. Maybe there are certain use cases like the one I just mentioned, but there's probably I'm sure there's other more common ones. Where No, you don't understand. They have zero. You help them go from zero to something. And that's huge and they'll pay for it already. Then you can go file the edges so you can charge more or they can be more thrilled and tell their friends cuz it's that good. Yeah, then. But going from zero to something could change a lot of people. So, one is remote workers just in general like companies that have remote at all. Also, I would ask like who what kinds of organizations are obsessed with pair programming at all? Like whether or not it's remote. So, like if I'm an extreme programming organization, then I have to. I doubt there's very many of those, but just, you know, whoever whatever is, you know, maybe some famously are. If I'm just obsessed with pairing at all, I might need this. I mean, who knows? Like who knows what the use cases are? Even if I'm in the same office, it might still be there might still be a reason why we don't like it's actually nice not to sit next to each other for some reason. You know, even we only have 610 total people and we have engineering on three different floors right now. So, we have pairing rooms, but sometimes they're full. So, like even if everyone's in the building, you might need this. I don't know. So, I wouldn't arbitrarily restrict those use cases. Maybe someone would pay for it. So, I think it's good to maintain that high standard for what you think is good. That's good. But it's bad if it means like I feel bad about charging or I'm not sure if I like then it's bad.

Okay. So, like one of my options is like just keep going full speed ahead on the sales. We could almost certainly close more people. Like there are people that want to sign up to like pay to be in the beta that launches at some unknown date in the future. Like am I crazy to not keep doing that?

I guess for focus because we're sort of at the point where we have like the cash in the bank is useful, but we're not hurting for it. And so like me doing like product management for example and like testing that the app out with people and things like that is seemingly more useful right now.

Mhm.

Mhm.

Yeah.

No one's paying to use it yet cuz no one has it. So people are like paying to eventually use it.

And so

No, we started working in May.

So the product is not launched.

Yeah. So yeah, and no one's using it because it's not ready enough, I guess.

Mhm.

I have Slack. I can send someone a link. It's really not hard to send a link. That's not why I need your tool.

So why would people want to give you money now? Why don't they just go use the other thing?

Okay.

That's interesting. So why do all the other tools suck? Cuz they can't get it fast enough?

So what is the typical use case? Why do people need this?

And so you're saying, 'Look, they could get on Slack video or Hangouts now. And then the problem is that it's high latency and other things.' Yeah. So if you give them a product that also is high latency, even though it's for developers, they're like, 'Whatever. I have Hangouts and it doesn't work.'

So you want to wait till it's better than Hangouts.

How long do you think it'll take before it's better than Hangouts? That the three guys who are good at engineering will beat Google at desktop sharing.

Your audience is devs and that's great and bad. The bad thing is they all want to tell you like, 'You should have this feature and it'll be easy.' And they have some reason in their mind why it's easy and they're probably wrong. The good news is they're devs, and so they geek out about the same kind of things. And they understand what alpha means. They understand a graph that shows latency and how it improves over time, and how that might imply that in the future it can continue to improve. I remember at SmartBear, of course, we also sold to developers only. And that was really helpful. When there was a bug, we could actually explain why it was a bug. And a lot of times they were like, 'Wow, yeah, now I totally see why that would be a bug.' And in a way that you could never do to another type of customer. And you really got this like not benefit of the doubt, but you got this understanding, and that was really really cool.

Same with features. We really want this feature, and then you could really geek out in crazy detail about why it's hard, or why well, but in these cases it doesn't really work, and we're not sure how to solve that, and they're like, 'Ah, yeah, I see that. Dang, yeah.' And like wow, I just said no to a customer feature, and they understood. What the hell?

So with that spirit that you can if you communicate about the geeky stuff, they actually typically come along. If you publish What if you publish the latency chart week by week? And so you're like, 'Look, we're entering alpha, and I'm telling you right now it's not good enough. Because here's our chart, and here's the line of what good enough looks like. And look, see how we're not there?'

See that? Right? Like I'm telling you right now it's not good enough. But look at our trajectory. So if you want to, you can start using a tool that we know is not good enough, but might be enough to get the job done better than not having it. And you can follow along.

It is, but I guess what I'm doing is trying to tap into what you said, which is it isn't good enough, and so you get this reputation of this tool sucks, and I'm not trying it again. So, how could you dispel this notion of the tool sucks today, and so it will always suck, which I 100% agree with you about, that people do that. Well, one way would be if you were crazy about this graph. Like, what if it's so crazy that when you install it and it comes up, it says, 'Hey all, this is our graph.' Like, in the tool, not at the blog, but in the tool. So, like when you run the tool, you cannot escape the fact that you know it's not fast enough, and it but you're working on it, and it's improving. So, hang in there. And if it weren't a dev audience, you're probably still screwed. Like, what are you even What is this? But with a dev audience, you can geek out about this crap. Learn more about how hard latency is, click. Here come a bunch of tech articles you wrote about Well, even measuring latency is hard. Is it the 99th percentile? Maybe you know, is median right? Not really, but there's this other way because when you're really using it, if it hiccups every 5 minutes, but otherwise is clean, that actually is good. Whereas if the average is a little higher, and it never hiccups, it's actually bad. Exactly the opposite is true in audio, and that's why all that shit's wrong when it comes to us. Blah blah, I just made all that up. I don't even know if that's true. But like, just geek it out about like, this is what even latency means, and how we think about it and measure it, and all these algorithms. And then, so because of the way we define it, we have these like Here's the little test harnesses we used to micro test some of these things because of this macro thing. Of course, the macro matters in the end, and you don't want to pre-optimize, but blah blah blah. You do that, and they see that graph, and it's not fast enough, but they know that. Now, do they stick around? I don't know, but it's a better bet. And if you can get that revenue going cuz revenue is this it's a commitment, they're part of it, they're loyal, they'll talk to you about stuff. And maybe they'll talk now, but it's different when they're really using it. Then you're getting real product feedback, right? Maybe they share it with others. Maybe they share the articles about the latency cuz everyone, you know, we all like to geek out on stuff like that.

That doesn't mean you release it now, but like there's got to be ways to say, 'Well, I'm worried like you just did. I'm worried that it's not good enough and we'll get a reputation.' It's like, 'Well, what if it's not good enough, but you have to have the reputation of being super awesome geeks as opposed to the reputation being slow even though it's the same product cuz of how you show it,' right?

Now, on the one hand, you've just set yourself up if you can't get it there.

But you just told me if you can't get it there, it's done anyway.

Now, the real ultimate thing would be to say, 'So, this whole thing about latency and measuring it, could that become an open source package that people contribute to? The part where you measure it.' Conversely, what about all this video codec or whatever the hell you're doing for all that part? Can that be an open source thing that people can contribute to? So, when people say it's not fast enough, you can say, 'Well, jump in the codec and see if you can make it faster against these'

Yeah, run the Docker container with the integration test and see if you can make it faster. And suddenly it's a contest. Now, that's a lot of work to go do and run and every So, I'm not saying that that's necessarily a good idea, but what I am doing is trying to say like how creative can you get in engaging what you have, which is a very specialized audience, which is other developers. That's not the normal customer. You can throw away a lot of the normal playbooks. You can talk about all kinds of weird stuff. In fact, it's better. It's better.

A developer would rather buy from a geeky developer who's Why do people buy Marco's podcast crap? Cuz it's Marco. And he geeks out about microphones. That's why.

Right? Like, this works. This works. So, geek out about the damn thing and I'm sure his software wasn't so good at first, but everyone knows Marco's going to obsess over everything and it's going to be better. You can win that. You have a platform and things like the podcast and blogging and the email list and stuff, reputation. You have a platform in which you can do that. Great, take advantage of it, do it. Take advantage of the weirdness of your audience. I would say.

Now, open sourcing is a whole 'nother thing. Like, what if no one contributes? Are you managing pull requests, which is now more work? Just making it open source is kind of you got to pull it out and I don't know what to do things. And document stuff in a way that you don't need to do now and that's just time. What if you don't want to accept someone's pull request? Now, that's a thing. People use it for purposes other than yours, so they want to make other changes that don't really make sense for you. Do you do that anyway cuz it's open source? So, like unless you're super into open source, in which case maybe it's worth it because personally you want to have open source stuff, then awesome. If not, probably it's too much work. Another thing and we think about this at WP Engine, too, cuz we do really these open source stuff is open source on purpose. One of the things we think of is, you know, whenever someone looks at our open-sourced item, they will judge our engineering prowess based on that code. So, even if this the whole point of this was it was a skunk works product, that's not at the normal standard that we use. Nevertheless, we will be judged on that standard. So, even though it's on the side, we actually have to make it really, really good in terms of everything, quality and testing documentation and CI/CD and whatever, because we'll be judged and actually So, it turns into a whole lot of work cuz we'll be judged and we want to be judged well. So, again, maybe not a great idea at this stage for those reasons, but it's kind of fun to talk it through anyway cuz it's a possibility, I guess.

Well, what if you got a tool like that and it said, 'Hey, it's not good enough.' I bet you'd use it and go, 'You know, it's okay.' Hey, if you're going to make it even better, great. Like, again, it's just like we said at the top of the podcast. You've just anchored it as like, 'I know it's not good enough.' That actually could be good. Like, people would be like, 'Oh, it's pretty good, you know.' If this is going to get better, then okay, I'll come along for the ride, but it's fine, you know. I like the link thing or whatever, you know.

Right.

Yeah. And what if you do and the latency is still not better yet? Will I come along for the ride in the latency given that you're honest about it and you're doing other things that I do want? Maybe. Like, maybe that relationship's good enough, right?

Just think of yourself if you got a tool like that. How would you react to that, right?

Which is not a great question, actually, because you're not like your customers at all. They have jobs and they write code. So, like, you're not actually like your customers. So, that's not a really good question.

Yeah.

What? What is that for? I don't know.

One other thing about dev tools, again, having bootstrapped a company that sold dev tools, I think in general it's fun to sell to devs for the reasons we just said. When it comes to the money and getting money, it's actually kind of hard. Like, people that type code into computers generally don't have a budget.

There is a budget. Yeah, there's a budget somewhere, but they don't personally control it.

You have to go ask and the front line developer Yeah, yeah. You have to go ask and it can be hard to get on the radar of that other person cuz that other person who does have the budget, they have lots of concerns and worries and things. It's hard for us like some random tool for something to sort of get on the docket.

I find it's different for different size companies. The smaller the company, of course, the easier it is for someone to whip out a credit card and it's fine. And the bigger the company, once they decide to pay, you're on there forever cuz now it's a tool in the tool belt and they sort of have to pay cuz it's like in the list. So that's great. It just takes a little bit more effort to get in the list.

Yeah, and there it's helpful if there's more than one team. Even better is like multiple business units, but just more than one team. So in other words, if it's just one team out of 20 teams and that team wants it, that is very hard to get the senior director of whatever it is to like even pay attention to the thing. But if there's like two different teams, then it's different. It's like, you know, we tried and liked it. Someone else in the office tried the totally different team tried it and they like it. You know what? Maybe we should just standardize on this. Let's just buy this. That's a totally different thing. It becomes a thing that we do as in our processes as an R&D department or whatever. It becomes a different thing. And it's a much bigger sale, too. Like there's more seats. That's good. So typically like that helps get over the line on those bigger organizations with dev tools. Even though of course it takes more time, it's more work, but the it can be 10 or even 100 times more revenue per customer meaning company or organization. So, okay, so maybe that timeline's worth it for that. Of course, this could be a viral tool in the real sense of the word viral. Two people use it, then a third person wants to use it, then a fourth, so they get in. So I think it'd be very good if the business model maybe and you probably already did this, but the business model allowed these incremental adds like just throw someone in and later we'll come back to your boss and they won't have a choice because she'll look at it and say, 'What the hell is this?' And you say, 'Well, 10 people are using it. Wait, 11 people are using it.' And she'll go, 'Oh, damn it. What do I have to do?' And you're like, 'Well, you buy the license that lets you use it up to 20 times. And we'll come back to you if more than that use it.' She goes, 'Ah, well, either I have to go and tell everyone to stop using it, or I pay.' So, that's almost for sure the right way to go is to make it more viral that way.

You want the friends to use it with somebody else, too. Like that's how viral works. Like you've skipped the part where someone's infected.

Right?

So, another way is kind of there are other tools which come behind like Slack, like Asana, etc. And I picked those because those are other tools which are viral and work well in the business world. And are viral, which is very different than consumer viral, which obviously isn't. But the business world viral are certain companies. And so, what they'll do is they'll say like, 'Well, we're never going to stop someone from making an account and using it.' Cuz if they make an account and they don't use it, it's hard to go to the boss and ask for the money. But if you make an account and use it, then the conversation with the person with the budget is, 'You have 11 active users. So, we need to settle up.' See how different that is than like 11 people logged in this one time? No. But okay, well, they don't need to log in anymore. That's the answer. 11 people are using it, and it's growing. And actually, we're behind on licenses cuz you only bought five or zero. So, we kind of need to true up. Like we don't need to go back in time, none of that kind of stuff. It's fine. It's fine. We're happy to be generous about that. That's not the point. But we do need to true up. Let's just go ahead and get you on the 20 pack so that let's just get you trued up there. And then we can go back to sleep. So, it's a totally different thing when they're actually using it than when they're not actually using it. So, I think it'd be better if they were in the system actually using it and then you can have the payment argument.

Yeah, it's not necessarily easy to design, but these are fundamental business model things that could really really change the trajectory of the thing. So, unlike a Slack, like Slack is often used for people that are not at the same organization. So, that's an interesting way that it becomes viral. But, pair programming is probably with people in the same organization. So, it colors the way you think about what viral means. So, it could be, you know, when the first person in the org signs up, they kind of establish this company object, as it were. And then things can slide in there. And why would you do that? Because you don't pay for it. Because it just works. Otherwise, you have to pay. Or I don't know. Like there's some Yeah, it works by your work email address. So, there's a domain so you can tell. And so, like some way where yeah, like we can just keep using this and then we'll see if we really like it, then the argument is it's already here. Either pay or tell the developers they have to stop using a tool they like. That's hard. There is no manager who wants to have that conversation with their engineer. They might say there's no budget, but like they don't want that. They really would rather not have that conversation.

That's interesting. So, this is interesting because this feels kind of like a bottom-up approach where it's like let people use it in a low-friction way and then hopefully come back and we'll pick up a sale and figure out how to make it work there. I've had pretty good success so far in basically like pre-selling a certain number of seats at a pretty high price for an annual plan. And like that's pretty cool, too. It's like I mean

I'm pretty into that.

It's Yes.

Yeah. Yeah, yeah, right. But I'm with you. Like if that's the mode is like we're going to have a sales team. We're going to marketing is going to generate leads for the sales team and the sales team is going to call in and make annual chunks and like 50k at a time or whatever kind of things. I think that's awesome. What Slack does is both. And again, I'm not saying you have to do what Slack does. I'm just This is just ideas, right? And things that have worked in similar context. So what Slack does is both. There's an easy way to just get in there and go because that's part of how anyone even knows about it in the first place. Again, even if you're having that conversation about the whole thing with someone, if they have one team that's actually using and saying, 'Oh yeah, yeah, we like that.' How much easier is the sale now? So it's pretty easy to argue that some kind of grassroots bottoms-up thing is helpful, you know, even if you're selling from top-down. And then so what Slack does is that's easy and then maybe you just buy the number of seats that you have and that's fine. Maybe they go to the top and say, you know, 'We could do that, but then we're going to be coming back to you all the time and you're going to keep paying a different amount and your budget's going to keep changing. And you know what? If we just go with an annual license and cover up to X people, well, the unit rate goes way down to this and we'll just talk again in a year.' That's exactly what Slack does, right? That is also what some other companies that do dev tools in particular will do. So there's this grassroots and this like, 'Hey man, we can see where this is going. Why don't we go ahead and get this all organized?' And that's a good argument. And then you can sell either way. You can go in from the top and at SmartBear we did both. At SmartBear sometimes, we would go in at the CTO level or VP of engineering level. Or sometimes at a big company like a Qualcomm, there'd be an entire group just about programming quality for the whole company. Like a whole group devoted to those things. So it depends on the structure, but there would be someone who in that sense that you're saying is like kind of above it all, either in the org chart or cuz they're tasked that way. Which this could be that too, by the way, right? Pair programming could be something that a group like that would consider, right? And so sometimes they would go in there and they're like, 'Great, let's buy 2,000 seats and that's the trial.' And if that works, there'll be 8,000 seats. Like you're like, 'Whoa, trial? That's What? Yes, right?' So there is that we had that motion, but also we had the motion where in our case it was code review, so just like you would be a group of people So we'd have a group of people try it. And there were just three of them or maybe 10 or something like that and if it works, they go get the money. Then another team looks at it. So 4 months later we see another sale for a little bit more and then eventually we sort of go high up the chain and go, 'Look, you now have four different teams across the country using this. It seems like this is what you need. Why don't we' And so sometimes we'd go that way. So it did both in that case. I know it does both in Slack, too. I'm pretty sure Atlassian is the same way. Right? Like someone sets up a Jira, someone sets up a Jira. It's free or it's cheap. Eventually they go, 'You know, you need one Confluence for everything cuz having wikis is not good, you guys.' And everyone agrees and so then they make the one big Confluence and charge you whatever. So again, I'm not using those to prove that this is right, but just to show, well, there's a lot of successful footprints along this path. So it's and you do have a viral component. Therefore, maybe although that's work, maybe that could really help the business be successful because as we all know business model is just as important as a feature. And the business model is just as tuned into product as features is. It's not marketing, sales, and business model on the right and product on the left. We all know that's not true. And here you have a product that's naturally viral, which is amazing. Most dev tools aren't. And so a lot of these things don't apply to most dev tools. It's more like that one engineer has to want it or like you said, like it has to be mandated from on high, which is not very common. And so that's hard and that's like so it's super hard to sell an IDE, for example. So, they're all free. You notice that? Like, so some aren't. Like, there's like two that have figured out how to do it and they're still around. It's hard. Like, it'd be very hard to launch an IDE, wouldn't it? Like, Oh my god.

Right? But you have a naturally viral thing. Wow. And I hate that word, actually, cuz people typically use it and it's not really viral, but you really are. It really is like we need you need to use the software for us both to use the software. Like, it actually is viral.

So, wow. So, I would like lean into that and even if that means product features that'll allow for certain pricing things to happen. Man, though, that could mint the business in terms of the business model. So, that doesn't sound like a bad use of time. Now, is right now the time? I don't know about that, because no one's paying anyway, so maybe it doesn't matter, but maybe architecturally you could think about that in terms of what does it mean to sign up? What does it mean to be part of a company? How do you account for that? How do you see that? That's where you could think about that.

And is the key thing to think about there making it spread easily? Is that the play with the viral situation?

Mhm.

Yeah.

It's tough.

Yeah.

Yep. Feel the same way. Awesome. This is really helpful. I don't want to take any more of your time, but it was great talking to you.

Yeah. I didn't even realize that overlap there, but that worked out nicely.

Too bad they have to pay. No virality allowed.

Yeah, cool. Well, plus I just told you that the product is too slow, so I don't know. We'll see how that turns out.

[laughter] Yeah.

Awesome. Is there anything you want to plug, by the way, while you're here? Point people towards?

Yep. I've been a customer in the past, and it works beautifully. So, nothing but good things to say there. And I'm going to throw you another plug, too, actually, which is your designing the ideal bootstrapped business talk at MicroConf is one of my all-time favorites. So, folks should for sure go watch that.

Beautiful. Well, thanks, Jason, for coming by. I appreciate it.