Jason Cohen: Working Out When to Break the Rules & Ignore Advice

Jason Cohen’s Business of Software talk in 2010 was a cracker. As someone who has consistently broken the rules he spoke about rules and when you might break them.

He also offers great advice about advice. He explains why you should always remember that advice, even from the best known and admired sources, should always be taken in context and offers a framework for filtering the legion sources of advice out there for entrepreneurs so that you can work out what is appropriate for you. A must view if you are in the habit of giving, or receiving advice!

Transcript

Joel Spolsky: Our last speaker of the day is Jason Cohen, who you all know if you’ve been here before. Smart Bear Software was his company, which he sold, and his current company is WP Engine. A big expert on code reviews, I hope you’re going to talk about that. Uh-oh. He said no. Good, because he talked about that last time. Please welcome Jason Cohen.

Jason Cohen: What’s up, Business of Software 2010! [Cheering] What’s up, Business of Software 2010! [Applause]. Great. Sweet. This is really just disturbing to have a time. I’m already 55 seconds in, does that sound right? OK. I feel like I’ve earned an extra 30 seconds at the end. So here’s the thing, if you can only take one thing away from this talk it’s this. When I’m finished talking we’re going to all drink and party. [Cheering] Yea! Right. So every word counts.

So, this by the way is not a litter box. I would not do that, that’s disgusting. This is actually a casserole made out of cornmeal and peanut butter. Which someone thought, I think is actually even more disgusting. Anyway, I’m a software developer. That’s still how I identify myself. I’ve started now four companies. And I’ve done every kind of thing within those companies, and I write, and I do all these different things. But still if you asked me how I felt, I still feel like I’m a coder. That’s how I identify myself.

And so when I started Smart Bear, the company I’m most known for, coding was the one thing that I was sure I could do. This was not going to be the downfall of the company. I knew I could pretty much implement any feature within reason that people wanted. So I was comfortable with that.

Sales, on the other hand, sales is kind of a different thing. In fact, I feel like, especially at the beginning, I don’t know anything about sales. I’d never done it. Forget about being trained, I’d never even really seen it. I don’t know what that entails, there’s Glengarry, there’s Glenross, I don’t know. Completely different types of people from me. So I didn’t really know what that is. And then there’s stuff like enterprise sales, I really don’t know what that is. No clue.

And in fact there’s lots of stuff that happens at a startup that I knew I was going to have to do. Scott referred to firing. I think that’s a really good example. Not only do you not learn how to do that, I don’t know anyone who’s comfortable with it even having done it thirty times. It’s never something that comes easy, I feel. In fact, although I had a very different experience with firing people than Scott related, actually after I got rid of somebody the whole office was like, “How come you didn’t do that sooner? We were all waiting like a year for you to do that. What took you so long?” And I said, “Well how come you guys didn’t say anything about it?” Well that’s because it’s my job, I’m the founder.

Exactly, that’s the problem. It’s our job as the founder, even though I think most of the people in this room are like me, a geek, it’s still our jobs. So what do you do about this, besides going into the business of software which helps a little, because now you’ve got two stories about firing. Great. Now you can go back and fire everybody? That’s really valuable. [Laughter]

Ok. So you go, you read books, you do things, you go online. Of course I learned, and we all learned, from Joel the importance of hiring the best. Not really the best but they’re smart and they get things done. And we learned from folks like 37 Signals that simplicity wins and design can be more important than features. And I learned a lot from Eric. I don’t know, Eric doesn’t blog much anymore but he’ll be speaking in a couple of days, and you should all go back and read it because Eric tells you how to do marketing for geeks. So very valuable, but I discovered something while I was trying to consume all this material. I’m sure everyone else here as fifty feeds in their RSS reader too. That for every book that says these are the rules, here’s the rules of management. “Oh, boy! At least now I won’t be a crappy manager!”

But for every book that says these are the seventeen rules of HR, right, or management, there’s another book that says “The best managers break all of the rules, and they do all this other stuff.”

Oh, great. That’s not helpful. And it’s not just these external sources of advice. You may have investors. Anybody here have Dave McClure as an investor? Nobody? Ok, one. Surely it couldn’t be nobody. I mean, God knows what Dave tells you. You know, the things he says on stage are bad enough, behind closed doors I can’t imagine. Anyway, but you have crazy guys like investors and Dave McClure telling you what to do. And the worst of all is, you’ve got customers telling you what they think you should do, and not just features, right? They tell you how to run your company too, right? And if there’s a group of really important people who cannot agree on anything, it would be your customers, right?

So you have all of these inputs of stuff to do, and yet at the end of the day, if you go forwards, you’re still kind of lost because it’s just a cacophony of inputs and stuff. And even today Dharmesh says, “Those guys at Sales Force, I don’t like how they do business. I do things the opposite of them for most things.” Of course they’re both successful. What the hell are you supposed to do with that? So which one do you pick? Or do you have to pick one? And if you’re doing a third thing, how are you supposed to synthesize? There’s just a lot of data.

So I feel like over the last ten years or so, launching four companies, that I have, I think, figured out some guidelines for how to accept this ocean of input and select for myself the things that are in fact relevant to me and valuable to me. And be OK with letting the rest roll off my back. Not feel guilty about maybe not doing something important and so on.

And so I’d like to share with you how I’ve done that so maybe you can feel less guilty. I know guilt is a strange word. But it actually, I think, characterizes how I feel about all this stuff. I feel like I should be doing all this stuff, I should be better at all this stuff, I need to be more knowledgeable about all this stuff. And I don’t feel like just getting fifty feeds in my reader fixes it. Hopefully this will help a little.

So the first story I want to tell is about blogging. I know everyone has a blog, boring. Or maybe that means it’s really relevant, I don’t know. So two and a half years ago I decided to start blogging. Remember – coder, geek , know nothing about writing, like to live in the IDE. That’s my world not, you know, Word. It was very weird, like “Word wrap? You’re not supposed… Oh, wait. No, you are supposed to do word wrap in this case. It’s fine. Word wrap.” [Laughter] Anyway, so of course, I didn’t know anything about blogging so what do I do? I go to the online centers of knowledge about blogging, which are ProBlogger and CopyBlogger, those kind of folks. Both of these guys have hundreds of thousands of readers, and they all give hundreds and hundreds of comments about any little thing that they say so people know what they’re talking about. And they do, and they have all these rules: don’t write too much, don’t write too long, post frequently so you stay in people’s brains, and so Google has a lot of stuff to crawl, break up things in titles and bullet points, and stuff like that. And that’s good. These are easy rules to follow, really. They’re clear. And so in the first year of blogging, this is what my RSS subscriber count looked like.

I know RSS subscriber count is not the only measure. Then again, let’s be honest. That’s what you check every day and where your ego is tied to, right? So it does matter. For those of you in the back, it’s about 400 subscribers after a year. Now here’s the thing. I don’t know if you think 400 is good or bad, or whatever. I was kind of hoping to have something like a thousand. Somehow, with four digits that magically means the blog was successful in my mind. Who knows why? But the more important thing was I hated it because I don’t like to write in that style. I don’t like to write short posts. I don’t have time to write frequently. And so it was a chore, it was labor. And then furthermore, it didn’t really have the effect that I wanted it to have. Even though that was arbitrary, still I wasn’t meeting my goals, right?

So I decided, screw blogging. But instead of just screw all of blogging, I’m just going to write how I want to, in whatever form I want, whenever I want, and just ignore all the supposed rules. Now keep in mind, I was not actively disobeying. I was not saying I’m going to forge my own trail because I’m so smart. I literally, this was me giving up. I’m just saying I’m going to do what I want because it will still be useful for me to grow personally if I do it this way, that’s it.

So instead of these kind of rules, and again I didn’t write these down. It’s just what happened when I didn’t care about the rules is I wrote these long posts, and I didn’t post very frequently, and you know I had sentences this big with lots of semicolons everywhere. Ok. Here’s what happened. Today? 17, 000 subscribers. Now what’s interesting is the last graph had a little blip on it. That’s this blip. [Points to screen]. It looked like it kicked up? Uh, uh, no it didn’t. That’s where it kicked up. Again, not planned. This is just what happened automatically.

Now what does this mean exactly? I had my set of rules. ProBlogger has the ProBlogger set of rules. They’re pretty much opposites. And so does that mean, ProBlogger is wrong? Because look, as soon as I shirked the yoke of ProBlogger I finally was freed. Is that what it means? That ProBlogger sucks? No, of course not. There’s a lot of successful bloggers who use the ProBlogger strategy. Like ProBlogger, for example, right? But similarly, I’m not wrong either, obviously. This is also success by pretty much anyone’s measure.

So what does it mean that we have opposite rules and yet we’re both correct, we’re both okay. What it means is, none of those rules were important about whether the blog was going to be successful or not. None of them made or break the blog. Of course, the analogy’s coming. None of those things, none of those one things, are going to make or break the startup either.

And this is liberating. Because if all those little bits and pieces don’t actually matter that much to whether you’re going to be a success or not. If none of those things matter, that means you get to pick them. Not even necessarily consciously. It means you can do whatever is easier. Whatever comes more naturally to you. But shouldn’t you be challenged? Oh yeah, like a startups not enough of a challenge already. You don’t also need to live in someone else’s box. It’s hard enough to have a successful blog, it’s hard enough to have a successful startup without also trying to transform how you naturally feel, how you naturally are into someone else’s thing. Because there’s lots of paths there. It’s not important – any of these particular things. And the other thing that I sort of realized in retrospect about this is that ProBlogger has a particular goal. It’s called ProBlogger. That’s because it’s for people who want to be professional bloggers. Now what does that mean? Well, for example it means you have ads on your blogs, probably. And to get money from ads you need impressions. And so everything on ProBlogger is about getting more impressions, like blogging more frequently, and posting shorter things because who has the time, and so on, being active on Google and all this other stuff. But that’s not my goal. That was never my goal for the blog to get impressions on the blog. I had other goals. For example this new company that I just launched, WordPress Engine, I know that – from reading Joel and from my own experience of course – that hiring the best team is not only vital but it can be the competitive advantage. This is a WordPress company. I was able to hire one of the top five WordPress people on Earth. He wrote books, he’s active in the community, he contributes to the core WordPress stuff, and he moved to Austin to work at this company as employee number one. Why did he do all that? Because he like the blog and he wanted to do a startup with me. Now that’s why I blog. How valuable is that? Impressions are valuable too. I just had a different goal and because my goal was different (to earn, let’s say, attention, to earn that kind of ability to attract people, and other things too) of course the same rules didn’t necessarily apply.

So rules have context, and I’m going to come back to that later.

The second story I want to tell is my foray, well, not really my foray, but my experience with so-called enterprise sales.

Again, sales, to me, is sort of the opposite of developer. Developers are logical and Vulcan. And sales people are, well they say people buy from people they like. What the hell does that mean? I don’t even know if someone likes me. I’m just as good as I can code, right? End of story. Programmers are, you know, we prefer e-mail but most sales guys I know prefer the phone. It’s just a different type of human. [Laughter] And also, having never been on a real sales call I didn’t even know really what happened. But I definitely had an idea of what happened. What happens is, there’s this guy who plays golf with other people where stuff happens. And they fly around in pressed suits in First Class on the airlines. And when they get to rooms they talk about sports and they say things like, “Hey, did you see the big game this weekend?” And everybody in the room knows which game it was. [Laughter] That kind of thing.

And then, at a critical moment, the enterprise salesperson leans over the oak conference table and he does the equivalent of the Obi-wan Kenobi, the old Obi-wan Kenobi, Jedi mind trick. And he says the equivalent of, “These are not the droids you’re looking for.” [Laugher] What he actually says is essentially, “You need to give me a million dollars for this software that you and I both know will not work, and you’ll have to rip out in two years.” [Laughter].

And it works, and it works. And by the time they figure out that, yeah, those were the droids they were looking for [laughter] the enterprise sales guy is off in a plane, off to the next victim. That’s what it’s like, and it ain’t me.

Now, looking back, at Smart Bear we did enterprise sales. Especially now that that defines a lot of the sales that are made at Smart Bear. So I know that actually it’s more like this. [Laughter] It feels kind of like this. [Laughter] And it normally results in something like this. [Laughter] Not knowing any of this stuff when I started Smart Bear, imagine how happy I was when… Again, it was just me by myself coding. I had this weird desk. You know how in the 80’s every desk had like 14,000 pieces of wood glued all over it and it was really heavy and had a back to it, that of course for some reason computers would never fit in. I had that weird, it was like you were in a confessional or something, kind of a thing. And I’m hiding behind this thing. That was it, that’s Smart Bear. Me huddled behind this huge, wood thing no one could move. Anyway this guy comes in. He has silver hair. This is already a good thing because see, I had been told that I needed someone that had a little more age, a little more gravitas, to help me. Because I look young anyways, and now imagine me eight years ago, right? Because just to go on a sales call, someone would probably would have to rent a car for me, right? [Laughter]

So this guy shows up. He’s an ex-VP of Sales for a big company in Austin that had just become super successful. He was responsible for God knows how much revenue, blah, blah, blah. And he wanted to work with me. He was looking for his next thing. I don’t know where he heard about me, but let’s just say his name was Frank. I was thrilled. And I was told, it’s great to be like the young looking geek, crazy guy for the tech part because this was like, 2003, so Google was still like the hot story of like, “Yeah, they’re like 22 or 14,” or whatever they are. And that’s great, although, they have adult supervision. [Laughter] So, good. That’s what I needed, I was told anyway. And so he would lend that thing. Here’s the guy with the Jedi mind tricks. Perfect.

We’re like Batman and Robin. I don’t know who’s who, but it doesn’t matter. [Laugher ]This is going to be great! Now he had a few things that he wanted. And this was great. This was like a learning opportunity for me. Just to know what he thought needed to be changed about the company. It was already kind of useful to me. First of all, he needs half the company because he’s going to take us from $100,000 a year – which is what it was, a little lifestyle kind of business – and he was going to get us $100,000 a month. Well, shoot. That’s worth half the company for sure.

Also, we needed to get a patent. You see the thing about patents is they prevent competition. That’s exactly what we need, especially when we’re going to go selling it to big companies.

Also he noticed that I worked a little bit too hard. I was always there coding and so on. You know you need room to breathe. Working more hours does not mean you’re more productive. And you should do things like knock off on Fridays, have a couple of beers, things like that. And I thought yeah, maybe I do need to lighten up a little bit.

And the big thing was, there’s no way he’s going to walk into a room at a target customer site, like Adobe or Intuit or Qualcom or Cisco. Those are the kinds of guys who made software development tools, and that’s the kind of enterprise software development tool target. And he’s not going to walk in there with a stupid name like Smart Bear in his cart. He said, “Look these guys don’t buy from companies called Smart Bear. These guys buy from companies like IBM, and BEA, CA, CSE. Like companies with gravitas and with acronyms, and stuff like that.

Smart Bear – that sounds like shareware. That sounds like one guy sitting in a loft in a tin roof building off of Highway 183 in Texas.” Which, of course, it was. [Laughter]

But, he didn’t like that. So he had another idea, Software Test and Deployment Systems, which certainly has a nice feel to it. It has an acronym as well . [Laughter] Lends itself to lots of ideas for how to market it. [Laughter] Now, you know. It’s easy to laugh now, and say what an idiot. I don’t know, me or him, whichever. But I was ready to go. Because Frank had been there. And Frank had the Jedi mind tricks, and I needed a sales guy. All these things were things that were told to me, that I read and so forth. And I almost did it. And there’s one thing that prevented me from doing it, and it was completely luck. Yet again, I lucked into it. See? Unfortunately, one of the pieces of advice is that luck has a lot more to do with it than intent, unfortunately. Anyway, the same week I talked to Frank I received a purchase order from Intuit, for $50,000. Little me, little Jason. Little shareware Jason with Smart Bear. Fifty grand. And from Intuit, exactly the kind of business that ‘would never do business with Smart Bear.’ And so it was apparent that Frank was wrong. But it shouldn’t have taken luck and an event for me to find out.

Hopefully this will help me and you to avoid this kind of thing in the future. I should have known. I had all the signs. Like, here’s this guy and he walks in and he says, “We’ve got to knock off early on Fridays.” We work seven… We? Me. But even when you’re one person it’s still we, right? We all know that. I work seventy hours a week, that’s the culture. We all talk about the culture. That’s the culture. And I know working harder, or even long, isn’t working better, blah, blah, blah. But who here does not… Who here works thirty hours a week? It’s not how you do it, right? But that’s how Frank was. It’s a mismatch. But that’s obvious, I knew that. And a patent? God. Who know a software company that doesn’t have competition because they or anyone else doesn’t have a patent? It’s silly, and I knew that at the time. And yet, I was going to do it anyway.

And the only thing I can think of why I wasn’t going to do it anyway is because he’s a real sales guy and I’m not. But here’s the thing. No one told me I couldn’t do sales. Nobody told me that I could not do sales. I decided that I couldn’t.

Maybe it was conscious, maybe it wasn’t. But I just decided that I’m not a sales guy. And that was the moment that I was going to make a bad decision, and only luck saved me. You can no longer say things like I’m not a sales guy, I’m not a marketing person, I’m not a technical person, I’m not a designer. Stop saying it. Don’t say you wear many hats. Just do it. You are not allowed to say that anymore. Oh, you’re not an expert in sales? Well of course you’re not an expert. But you’re not an expert in any of those things that was on that slide full of crap that you’re going to have to do in a startup. Not any of them. And yet, you’re not going to have to do it. And the minute that you say, “I’m not that kind of person. I don’t know how to do that. I need to defer.” That’s the moment where you start making bad decisions, or taking bad advice, or just reading something on the Internet and saying, “Well, since I don’t know and can’t possibly understand this subject, I guess I’ll do whatever’s on this particular blog page.” Which is really what you’re doing if you’re being honest.

And besides, ok, you’re not an expert in sales. But are you telling me that you don’t know when you’re being sold to by someone and you hate it? You hate every moment? They’re so, whatever, that it riles you up? You’re telling me that you don’t know that’s how you want to not be? And you’re telling me that you’ve never been sold to where you’re sitting there thinking, “I know this guy’s selling me something but I love it! This is great, I’m going to buy this stuff!” You’ve experienced that. You know what that looks like?

Same with marketing. You get, what, 200 ads shown to you a day or whatever the statistics are. And you have no feeling at all about what ads you like and not? About what techniques you like or not? Of course you do. You just have to let yourself do it. And you have to trust that what you feel is maybe good or bad about this, actually can be just as be just as valid as having data. This is hard for engineers like me to accept because again, we’re all Vulcans and we like data, and metrics, and experiments, and evidence, and logic. And that’s how we make our decisions. Or at least that’s how we’d like to make our decisions. But the truth is that you rarely have that much data and it’s rarely that high-fidelity, especially at the beginning. And things are going to change anyway, and your logic’s not that good anyway. And actually, all that crap is not that good, at least not at the beginning. Not until you have enough data points that maybe you can have a bunch of metrics. Like maybe 3,000 customers that you can watch. But when you have ten or zero, none of that is really present anyway. You kind of have no choice.

Here’s the thing. I’m going to give you one example, that I’m sure you all can agree with, in which it is correct to trust your gut and you’re comfortable doing it. Everyone here, who is a developer anyway, has worked with a brand new developer. Maybe you hired them or they’re your peer, it doesn’t matter. You know within a week or two of actually working with them – side-by-side working with them – whether they are a superstar, awesome developer or whether they’re kind of mediocre or whether they suck. You know it. There’s no doubt, and you’re almost never wrong. Ok, maybe somebody, they took a while to settle in. Almost never. You know it.

Now if you were perfect at gathering data and making a test, you would have found that out during the interview process, right? So, data and tests and stuff – that doesn’t necessarily tell you whether somebody’s a good developer, but you know in your gut without any sort of quantitative numbers, whether that person’s good or bad. And probably by now you trust that implicitly. And I’m telling you that you can trust that kind of thing, or actually you kind of have no choice but to trust that sort of thing when you’re early on in your startup.

In fact, side note, I have found, doing many, many different kinds of hiring processes and whatnot, the number one best thing you can do is pair a program with a person. I don’t look at the resume actually. I throw out the resume. Who could care less about the resume? I don’t care where they went to school, I don’t ask what they did three years ago, I don’t care. We take him, we put him in a codebase – the real, live codebase – and we work on a real, live thing we’re doing today. We’re fixing a bug today, we’re adding a feature today, it doesn’t matter. This is what they’re going to be doing, right? And your gut feel working with someone, literally, side-by-side at a keyboard for two hours – you’re going to know if they’re good or bad. Since we’ve started doing paired program based hires, we have not hired a single person we’ve regretted.

I’m not saying you can do that forever and it scales too. I don’t know. I’m just throwing it out there that your gut check is good on a lot of these kinds of things, and you don’t have a choice.

So about this name, Smart Bear. I sold Smart Bear, as Joel said, to a company called Automated QA who went on to acquire other companies. And eventually they decided, “This is too many brands, this is too much confusion. Besides we’re integrating products. We should just have a single brand.” And so they interviewed customers and other people. “How do you feel about this? Would you give us more money if we were called that?” [Laughter] I don’t know. Whatever kind of strange, you know, machinations you do to understand that.

Anyway, the bottom line is that the entire company today is called Smart Bear. Sorry, Frank. [Applause]

So. Make every mistake only once. Well, it doesn’t always work that way. Let’s go back to sort of the middle of Smart Bear.

So the Frank bullet whizzed by in the early days. Now fast forward, maybe three or four years. We’re doing millions in revenue. There are employees. We’re growing, we’re doubling every year. We have a 50% profit margin. Five-zero. Even with growth. Doing great. And I was told by several people, some of them I consider some of my best mentors. “You know what, if you’re doing this well by yourself, and by yourself I should say what that is.”

So, what did our sales process look like today? Well, I was no longer under the impression that we couldn’t do sales, or that I couldn’t do sales. Of course we’re doing sales right now. So I did the demos myself. One or two a day. That’s kind of what I spent most of my time doing, actually. And tech support was sales.

There were no salespeople, so tech support was sales. Furthermore, there were no real tech support people, so that really meant the developers did tech support, so they did sales. So everyone had to write well, and talk to customers, and stuff. That’s maybe a whole other topic for another talk, whether that’s a good idea or not. It was a good idea for us anyway, in our particular situation. Anyway, we’re doing sales. Well, we’re doing something kind of weird. We’re not doing any traditional sales, but money’s coming in. So that’s kind of cool.

But I was told by a lot of people, “Man, if you’re doing this well without a sales process. Without the things you heard Dharmesh talking about – the funnel, and numbers, and conversions, and whether you should have salespeople and testing. Man, if you’re doing this good without any of that, just imagine how good you’re going to do if you had a real sales process.” So great, that’s a good point. We should have real sales. Besides why am I doing demos all day long. That’s probably a bad use of my time. So we started doing this kind of stuff, and we had kind of a new mission. That we should be sales focused [Laughter – slide says “You’re commissioned CRM your territory up their pipeline]. Do it in a traditional way, measure stuff. And this is a long story, it goes on for years. Literally, years we tried this so I’m not going to go into it because you don’t want to hear that [laughter – slide says “Fail slowly & repeat mistakes”]. So I can summarize by that saying that it was a disaster.

We did not get more revenue. The sales guy did not bring in new people that were not coming in already. In other words, it was all inbound still even though it was supposed to be outbound. And the developers were still doing all the tech support, answering all the questions because our software was technical and so the sales guys couldn’t answer the questions. Complete failure.

Now in retrospect it’s clear that we should not have done any of this, and that we should not have attempted this big thing of real sales. And that becomes clear when you analyze and you diagram out the decision-making process that I went through to arrive at this conclusion, which looks like this [laughter – slide says “Everything’s going well.” “Great! Let’s change it.”] Remember I just said we had like, 50% profit margin, millions of dollars in revenue. We’re doubling every year. Nothing’s wrong with sales. At least, nothing’s wrong with the things that sales are supposed to do. The result sales are supposed to have – there’s nothing wrong with it. And so we decided not to like, incrementally test one thing, and maybe we should measure first. No, it’s like [makes raspberry sound – Laughter]. And again, like, when you put it that way, it’s inane that I would do it. Why is it… I mean, I’m not that dumb. Why is it that we did something so dumb. And the only thing I can come up with is that everybody told me, and the things that I read, this is how you do sales.

Especially enterprise sales.

This is just what happens. That’s it. There’s kind of no more explanation than that. No more detail than that if you really dig in. And so, “How it is done” is not an excuse. It’s the same thing as saying “Everyone else is doing it,” which is kind of an excuse my kid would use for something. “That’s how everyone else is doing it.”

“Well if Joel Spolsky told you to jump off a cliff, would you jump off a cliff?” Well, for maybe 10,000 stack reputation points, maybe. It depends on the cliff. Maybe, at that point, yeah. But it’s never a good excuse. “That is how it’s done,” in any format is not a reason for anything. Now. That’ doesn’t mean that all existing ideas are bad. Obviously that’s not what it means. What it means is just because it has age, just because it’s been successful before, is not in and of itself enough to make a decision based on it.

A great example from my career is, “How did you launch Smart Bear?” which was bootstrapped. “How did you get enough marketing to get people in?” That’s like one of the questions that everyone starting a startup has. “How do you get people in cheaply since it was bootstrapped?” And the answer is that I used AdWords. That’s not an answer anymore. You could do exactly what I did and it would cost you a ton of money. Because remember it’s seven, eight years ago. When I had a keyword for Subversion I was the only one there. When you searched on Google for Subversion you got me. That was easy. And it was $0.05 a click. Terrific. I really could build a company on no money and all that kind of promise that this kind of online marketing has. That’s not true anymore.

AdWords is so expensive for us now that at some point, anyway, I think we stopped for six or twelve months doing AdWords advertising, because we did measure our marketing very accurately and it was terrible. It’s extremely inefficient, even though it’s how I built the company. Because AdWords changed, because the world changed, whatever, because stuff changes, right? So even ideas that were right before don’t get an automatic pass, because stuff changes just like AdWords.

So I just gave you like a bunch of ways to help filter some of the stuff, even when people say, “But you’re not doing it the way you should be doing it.”

By the way, another thing about doing it the way you ought to: It could be, not necessarily, but it could be that not doing it the normal way is your competitive advantage. So, we were doing enterprise sales without the usual salesperson thing, where they harangue you, and call you on the phone, and “Well, what can I do? What’s it going to take to put you in this software?” [laughter] And this kind of stuff, right? Maybe the fact that we weren’t doing that to people, we weren’t subjecting people to that, was an advantage. Maybe we were getting customers because we weren’t like that.

No salesperson has ever told me that was a good thing to test. You know, to see whether A, B, or anything like that. That’s not what you do. It’s just “this is the way.” And so, you know, maybe getting a demo from me is part of the reason we’re getting the sales. Don’t you suppose I could do the best product demo of our product than anybody? Don’t you think that when they had some weird corner case question about “How it interfaced with the clear case, if I’m doing branching but not streaming, and I’m using this patch level, and then what do you do about the diffs?” and I can answer it? Don’t you suppose that just put him in a place like, “Ok, I can actually trust this person.” And even though we weren’t haranguing them, they knew that if they were going to get a code review tool it was going to come from us because we really knew our stuff. I mean just the demo – the dumbass sales guy in the demo, that’s me – was able to answer the questions. That’s confidence inspiring. Now maybe, maybe not. I’m not telling you that’s why we were successful. I don’t know, but sometimes not doing it the normal way is the competitive advantage. Yet another reason not to just do it.

Ok, so here’s a bunch of ways in which you can start to filter the cacophony of advice and crap that’s coming at you.

Still, I think that rules in the abstract, even with stories, may not be sufficient. So, I would like to workshop, in particular, a couple of 37 Signals ideas. Because, probably everyone in the audience is familiar with 37 Signals. And see how I filter that and maybe you can imagine how you filter it as well, to kind of get you started down that path of ignoring and feeling ok with it. So for the four people in the room who are not familiar with 37 Signals. They make a bunch of products. They have a bestselling book, a New York Times bestselling book. They’re blog is read by 18 million people or whatever. And they’re very opinionated, and smart, and interesting, and that’s why I picked them. So here’s three of the tenets of 37 Signals that I’m going to particularly look at.

One of their rules is only solve simple problems, not hard ones, not complex ones. Why? Oh, and a quote from them about this is, “Under-do the competition.” That’s interesting. Why? Because startups are hard enough. Writing code that works is hard enough. Just solving a simple problem, forget about tackling a complex problem domain. So why bother with that? There’s so many simple problems you could try to fix. Do that.

Also you shouldn’t plan. Another quote from them, “Planning is guessing.” You don’t have enough data and you don’t know how the future’s going to change. So you write something down and then you execute against it. And like Eric Ries just said, all you’ve done is – you think you’re making progress because you’ve executed against a thing you’ve just invented, when in fact you’re just ignoring what’s really happening, right? So don’t plan.

And the third thing is, don’t work too hard. A quote from them again, “Fire the workaholics.” Like me! And their argument is again, working more hours does not mean you’re more productive or better. And anyway they say, when you want to hire the best people, the best people have many interests, and you have to give them time to pursue the cooking or hiking or riding or whatever they’re doing, so that they’re nice and fresh and happy and want to work for you.

So at Smart Bear, here’s kind of my take on this and then we’ll get to why.

So first of all, we solve a really weird, complex problem. Remember our customers are big, it’s enterprise software, it’s people like Adobe and Intuit and Cisco and Qualcomm, those are all our customers. And they have these crazy workflow problems. They have 26 business units and each one wants different rules, and this person’s only allowed to see this, and they have people in Canada that are only allowed to see that. And they do some work for the DOD, which means this stuff has to be in its own silo. Of course, we need to be able to do reporting across all that, and do custom reports, and “How come your tool is also in Crystal Reports [laughter] and do all that reporting? And we have seventeen people and some of them do the review overnight, because they’re in India, and some people are there at the same time and have to go fast,” and like everything about it is super-complicated. Because the problem domain for our customers is super-complicated.

What this means is, if you are Adobe you don’t have a choice but to buy a code review tool from us. But building it is out of the question it’s so complicated. We can barely get our arms around it, right? And that’s what we do for a living. And all of the other competitors are too simple to match their needs. That means we can charge them a lot of money, and it means we have fewer or no competitors at that level. That’s a huge competitive advantage to us.

We do have competitors, right Scott? We do have competitors at lower price points, and with different, simpler workflows. And then it’s everyone’s game. So if all we did was solve the simple problem, we’d be competing with Atlassian, I guess over price, maybe over features which, I guess now that Scott’s got over $68 million in his pocket, is not something that I want to do, right? [Laughter]

And he named it Atlassian, right? Even better. Scary, right? [Laughter] Yeah. So just solving simple problems, not good enough for us. Not a good strategy for us.

Planning is silly. Actually, I agree with that one.

Don’t work too hard. Of course I just said that I do work too hard. I think that eventually… Actually, it’s true. Eventually I really was working forty hours a week for Smart Bear. That’s true. I got there. I think eventually you can get there, just not at the beginning. I think that’s bad advice starting out.

So more interesting than the particular who thinks what about what… I don’t know where I’m supposed to be pointing this thing. [Laughter]

More important is how to apply these previous rules to this so that you could… I could have arrived at this already for Smart Bear, and more importantly of course, is that you can arrive at whichever combination of this is right for you.

So in particular, I’m going to take the second point, which is that all of this advice and stuff comes in context comes with goals.

Just like ProBlogger and me had different goals. So the B2B – selling to businesses versus selling to consumers – that’s pretty obvious. If you were here two years ago, or watched the video “Rich versus King?” So briefly, if you’re in the king category it means you want to run your company forever. You want to build a company that outlasts you, you never want to sell it. You want it to go forever. There’s a lot of people sitting right here, right now, that are in that category. [Laughter] One of them is split in half though. Still, that’s recent.

And then there’s people who, their company is their baby, of course they care, they put everything into it. However, for the right price they’ll sell it, right Dharmesh? That’s Dharmesh. Just said, “$25 million? Take it!” [Laugh]

That’s the rich category. I’m pretty sure if I offered $100 million for Fog Creek you guys wouldn’t take it. There you go. Maybe they would. I don’t have the ability to do that. Maybe you can, I don’t know. [Laughter]

Anyway, 37 Signals wants to run a company forever. If you read their stuff, boy do they ever, right? And they sell to businesses. Smart Bear in this category, clearly. In retrospect that’s obvious, right. And this right here explains already the differences, for example, why it’s valuable for us to solve complex problems that give us tons of money per customer and it’s not necessarily good for them. You see, remember, now they’re this big company. Everybody knows who they are. They make millions of dollars, they have lots of employees. That is not how 37signals began, let’s keep this in mind. It was guys with a consultancy, and they liked talking about design and so on. It was originally, at least, it was a lifestyle business. As a lifestyle business, why would you ever pick something hard. If you just want to feed yourself and maybe not work too hard yourself, and so on, of course you’re not going to pick a hard problem. You do not want Adobe as your customer if you just want a lifestyle business, right? And so of course it makes sense to them to solve a simple problem, and own it, and be able to continue selling that kind of thing forever, and also not work too hard. But that wasn’t my goal. And for me to be able to say, “Look at this customer list,” and be able to roll out this amazing list of logos which are all on the website, almost none of which are there with permission so when one of them complains we take it down, but [laughter]. That of course is really juicy to somebody wanting to buy a company or whatever, because clearly each time we get a customer it can mean many, many thousands of dollars.

That’s better for us. It’s more valuable for us also to not have the competition. I think it’s probably OK for me to say now, that when Smart Bear was being acquired one of the big questions was, “What are you going to do about Atlassian, because Atlassian has a code review tool too?” that they had just bought from another company it’s a really great tool as well. And you know, part of the answer was that we own this big segment of the market – the most expensive, the most lucrative, end of the market in terms of dollars per customer anyway. And you know, Atlassian can do whatever they want, of course. If they wanted to go after them they absolutely could. But it’s not in line with their culture and culture’s a big thing.

So probably after you just heard that whole speech about the ASP global, we need to have no touch – doesn’t that sound exactly the opposite of complex problem, and Adobe, and six or seven figure deals. Well it is the opposite of those kind of things. So we can co-exist. That’s a great argument I could make. I could not have made that argument had I had taken the 37signals case. I needed to have something where we had a nice niche and we made a lot of money, and that’s what we had.

But an even more devastating example of this versus 37signals is Mint [Laughter]. And the people who are laughing know what I’m referring to. Mint is first of all, they sell to customers. They also want to make money. We know this because they sold to Intuit for $170 million after being around only two years. They made a consumer based budgeting tool, where you could put all your money and see it all in one place, do budgets, and get offers for things like credit cards and savings accounts that would make you more money. Good for the customer, amazing for the credit card companies, because they have all your transactions line-by-line. You want to talk about doing credit risk properly? Forget about whether you paid your bill on time. How about, line-by-line, your whole expenditures? Oh my God. That’s why it was incredibly valuable. Ok. When Mint sold to Intuit… Yea! My Intuit. They gave me a lot of money, so I like Intuit.

When Mint sold to Intuit Jason Fried at 37signals wrote this great post. All I have to do is tell you the title: “The Next Generation Bends Over.” Oh, I wonder how he felt about that transaction? This was a great day for guys at Mint, not a great day for Jason Fried. Why? What the hell? What’s his problem, about this right? The answer is that they’re not anywhere in the same goals at all. Like, not at all. Jason wants to run it forever, so he’s pissed because they could have… Well, he says, “Taken down Intuit.” I don’t even know what that means. Intuit makes a lot of software for like, counties in Oklahoma that have to do their own weird, taxing stuff and they make that. I don’t even know how even Mint would address that. It doesn’t matter. Mint could have taken down Intuit and attacked them and fought them, and instead they sold out and gave in. Of course, this goes against his very being and he lashed out as a result. Anyway, this explains why they have a difference of opinion, let’s just say. And of course all these folks – that I’m sure of this list everyone here reads some of them, and whatever – they all have kind of a place. They’re all segmented in this way. And it’s useful to think of again, goals and context.

This is part of how you can start to filter out that RSS list in particular. Of course that doesn’t mean that just because you’re in this category you’ll never read that guy. That’s not true.

But two things are true.

One is, your filter can go up. The further away you are away from somebody, the more you can be critical about what they say, about whether or not it really applies to you. How guilty are you going to feel about your desire to sell out after you read that article from Jason Fried? Do you choose to ignore that and just say, “I’m not like that”? Or, are you in his box and therefore maybe you are inconsistent and this is really valuable advice to you.

Of course the other thing is, if you have fifty feeds in your reader that’s probably not a good use of time. Not that many. So you could use this to cull it down, and if you want to cull it down, this is how you do it. And of course there’s not just two dimensions, there’s five, that was just an example. There’s many things. I know that I fall into certain categories, myself. Of course a lot of people will leave comments on my blog and send me emails and stuff. And there’s a lot of people that say things like, “Well, you say you’ve got to seek revenue from the very beginning, and you don’t like freemium,” and I don’t. I don’t like freemium.

“And you don’t believe in the value of the network,” and all this stuff. And part of that, for example, is because I am coming from a bootstrap mentality, not a VC funded mentality. VC mentalities often go big, go do land grab, and later on we’ll figure out how to monetize it. That’s simply not my strategy.

It’s not wrong, it’s just not my strategy.

So there’s lots of stuff that I write about that would be wrong for you to do in that case, right? I know that. Another one that’s common is this pleasure / pain thing. The idea is, the way you sell pain is you say, “Doesn’t it suck that your inbox is this big and there’s too much mail?” “Yeah, that’s awful! I waste so much time. That’s a pain.” You say, “Great, our product makes it go like this, and all the stupid stuff goes over here, and all of a sudden you save time.” That’s a pain argument. I have a problem, you fix the problem. A lot of B2B stuff is a pain argument. That’s also what I like to sell. I feel comfortable addressing a pain, addressing features and benefits at the same time, when it’s pain.

But then you have the pleasure argument, which is stuff like, “Why do I have an iPhone?” It’s definitely the worst phone I’ve ever had, right?

It’s one of the most incredible pieces of technology ever built, and it’s a terrible phone. But I have it because it’s cool. When the new display case came out, and it was four times denser, and when you looked at the italicized text you couldn’t see the dots. You couldn’t see the dots! [Laugher] Goddamn that’s cool! You know. Does it solves a pain? No, I can deal with dots. It’s just that it’s cool, really. Which is fine. That is of course, a great way to make money too. But that’s not what I’m going to write about because, I don’t know, I admire it but I don’t know how to do it.

So one final thought because I guess there’s questions. And also, there’s drinking. So if I finish early I don’t think anyone’s going to be too upset. You can stay in the seat if you want to keep it warm until the time expires, if you really feel the need and if you want to get all your dollar out of the conference. We can do that.

I’ll leave you with this. That all this advice is really like the glass half empty, glass half full. It doesn’t matter. Pick your favorite next line. You know, “Drink it.” It doesn’t matter. It’s nothing that you hear – not today, not from blogs –is the thing that’s going to make or break your startup. Maybe it’s the product, maybe it’s the idea, maybe it’s the timing, maybe it’s luck. I don’t know. Maybe it’s that more than other stuff. I don’t know. Maybe it’s that more than other stuff. I don’t really know.

But it’s not going to be any specific thing. So, pick the things that are naturally right for you. That’s what you’re going to be best at anyway.

Again stuff that’s too hard. Starting a startup is hard already, then you’re also going to put yourself in someone else’s box instead of doing the thing that comes naturally to you? The thing that you’re going to be best at? When in this game, like Seth was saying today, is a game about being the lynchpin, about being the best at this, or that? You’re never going to do that playing by somebody else’s rules.

So whether or not you’re right, or whether or not you have data, it kind of doesn’t matter. You kind of have to trust that whatever you naturally do is going to have to be the right thing for you. Because that’s the only thing that you’re going to be really, really fantastic about. So if I’ve been successful then, the next time you open up your RSS reader and start saying, “Ok, it’s time for me to learn some more.” Or tomorrow when you come back here, or even today, after a long day – and thanks for listening because it’s been a long day of people yelling at you about all kinds of stuff, even contradictory stuff [laughter], right. We have heard some contradictory stuff already.

Hopefully I’ve given you a new set of filters where you’re more comfortable saying, “You know what? I’m going to ignore that and do the opposite and I’m ok with that. I’m going to stop feeling guilty about it.” And this actually, this is super valuable and I’m going to listen to this. Hopefully this gives you a better feel for what those things are and makes you even emotionally more comfortable having to make those kinds of choices. Thanks. [Applause]

So questions, besides, “When do we drink?”

Audience: So can you tell us a little more about why you picked WordPress as the subject of a new startup?

Jason Cohen: So quickly, the origin is that I had the problem myself. Which is usually a bad, which is not enough of a reason to start a new company, but it’s a spark. It’s enough to spark it. So what I did first – this is very Eric Ries style – is I found thirty people who said I will give you $50 a month if you build that. And I said, “If I can’t find thirty people on my own to actually give me the money,” again with no lines of code written, no website, no anything. “Then it was a bad idea.”

Audience: [inaudible]

Jason Cohen: Meetings and email. A variety of things. I actually had another idea before this, and could not find the thirty people so I abandoned that. What was that? It was a marketing web analytics type of tool. There’s many of them of course, but this was a particular one that I thought was pretty neat but not a lot of other people did. That’s ok. [looking for more questions] Everyone wants to drink.

Audience: There’s usually a tradeoff between a lower price point and the need to have like an outbound enterprise sales team. And it seems like you figured out some way through it. You don’t have the high touch sales but your product is certainly smart there. It saves a higher price point [inaudible]. So how do you do that?

Jason Cohen: Number one. We didn’t reach our high price point for a while. Early on it was low. And then we sort of raised it over time as we sort of earned the ability to do so. Another struggle that we went through. Another thing is that though we didn’t have a high touch sales force in the sense of having quarterly numbers to hit and all that sort of stuff, we were doing sales. It was just that we had a lot of tech support, which really would help with sales. So I was on the phone with people. And so it’s really not that we had zero touch. It wasn’t zero touch, it’s just that we didn’t use the traditional process. It’s actually very similar to how Dharmesh was talking about with vacation time. “This is how you do it.” And he was like, “We’re just not going to do that.” That’s more fortitude than I had at the time. Hopefully I now have that fortitude to just decide no vacation time. Well, we don’t actually have vacation time so actually, sounds like we did do that. But in the same way, we just said like, “This is how we do it.” Well, we’ve discovered that other way was not better. But it just morphed literally without thought. It just grew into that and that was ok.

Audience: What does Smart Bear do for sales today?

Jason Cohen: Smart Bear now has a sales force in Boston.

Audience: Is that working for them?

Jason Cohen: Yes. Revenues are going up and yeah, they’re profitable. They’ve bought other companies.

Audience:  But the model didn’t change the whole box. The tech support guys, the developers still support the sales guys. You know they organized it a little bit, the tech guys do their thing. You were the glue – you just pulled it all together.

Jason Cohen: No, no, no. That all happened after I left.

Audience: Ok.

Jason Cohen: I failed at the sales.

Audience: That’s what I’m saying. You were the glue. You actually sell.

Jason Cohen: Oh yeah. Yeah.

Audience: [Inaudible] You were just doing it your own way.

Jason Cohen: Yeah. But then it wasn’t just me. We had a wonderful woman who was also doing sales. It just, it wasn’t commissioned. She wasn’t making numbers. Genuinely, she just wanted people to be happy and so did the kinds of things that people would like. And so, we had that too. There’s other ways to offload some of the real problems, like it’s true that I cannot always do it myself. There are ways though to offload that. And the point is that whatever we decided to do seemed like it was right. But, I mean you can’t look at it now. It’s a global company, it’s very big, there’s different management, there are different goals about how big the company should be and what it should do. So it’s very difficult to say, you know today, and extrapolate anything from that.

Audience: Was there thought, when you went from the low price point to the high price point? And I assume you had to give up some of you early clients to moving prices around. Was there thought that, one, what kind of wrestling did you do over that? And two, did you think, “Well, let’s also make four things with a simpler product and compete on that since we already have those customers,” and then go this other direction?

Jason Cohen: So the question is, as you raise price don’t you re-segment the market? Don’t you lose your original customers? Maybe lose your way, as well, since that’s what you built it around? And then was there any angst about that? And then the second part was… oh maybe a light version or something. The answer is yes to all of those things. Everybody, I’m sure everyone here has angst about price, and that never, ever goes away. One of the cool things that happened, one of our customers told us to raise the price. They said, “This is way too cheap. I want to know you guys are in business. I want to pay more.” The first product $29.95. Now it’s $1,200.00 for a floating license. Of course, we went there like this. Every single time we lost some customers, there was angst, and “How high should we go?” Then there were all these kind of things like, “Make it higher and do a coupon, because then you can test different thing.” Also if you say it’s high then give a discount, they feel like that’s a better price than this.

There’s another thing here, I’ll give you another data point which I do think is true across a lot of industries and stuff. Which is, if you have two price points. The people that you’re targeting for the bigger one will still go for that but they’ll feel much better about it because they’re like, “No, I’m getting the primo.”

So if you just take away the lower one… We did this as an A, B test by the way. That’s how I’m confident about this. And we did it more than once, at different price points. So I’m pretty confident that when you insert a lower price thing, people will still buy the higher price thing. Of course you build the features accordingly, but they feel more comfortable. Because well yes, you’re paying that but of course you don’t want to get the little, light crappy one that doesn’t have LDAP support [Laughter] or whatever the thing is. Or a T-shirt. [Laughter] Yeah, that was a great story, I love that.

Tons of angst. Never sure whether it’s right or not. Still not sure. Pricing isn’t about one thing. But one thing I would say is, if you have no pressure on the price then it’s too low. You never really know if it’s too high, because maybe you’re just segmenting a different population. A little bit of back pressure but they still mostly buy is, I think, the right place to be. They should be going, “Well, I’m not quite sure.” Then like 80% of them do it anyway, there you go. That’s the best I got. One more question.

Audience: Do you do ever regret selling?

Jason Cohen: No. The question is, “Do you regret selling?” Exactly what Dharmesh said. When you do sell, what you’ve bought is the freedom to do, not just the next thing, but anything you want. When I sold, you know a year later, you know when my time was over, I got to spend months and months and months at home with my new baby. I would never give that up. Never regret that, never, ever, ever. Or now I’m doing WordPress Engine and other things. I could work on my blog some more, and you see what happened. Even if those things had not happened or whatever, there is always the next thing. That’s true. Not wrong, either. Don’t want to ruin it for everyone. But you make a good point which is, “You’d better figure out which one it is, because there ain’t no turning back once you’ve passed through that door, that’s true, yeah.” [Applause] [Beer].

Jason Cohen

Jason has  built four software startups, both bootstrapped and funded, both alone and with co-founders. All of them grew to more than $1m annual revenue.He sold two, and currently serves as CTO of the fourth, WPEngine, with 380 employees headquartered in Austin, Texas.  More recently, he has also been an angel investor and was a founding member of Capital Factory, an Austin incubator and co-working space. He writes about software and startups at ASmartBear.

More from Jason.

Next Events

BoS Europe 2025 🇬🇧

🗓️ 31 March – 1 April 2025
📍 Cambridge, UK

Grab early bird tickets until 15 Nov

Spend time with other smart people in a supportive community of SaaS & software entrepreneurs who want to build great products and companies.

Lock in Early Bird Pricing

BoS USA 2025 🇺🇸

🗓️ To be announced soon
📍 Raleigh, NC

Learn how great software companies are built at an extraordinary conference run since 2007 to help you build long term, profitable, sustainable businesses.

More Details