Bringing Back Free
21:28 Share

Welcome to ReWork, a podcast by 37signals about the better way to work and run your business. I'm your host, Kimberly Rhodes, joined by the co-founders of 37signals, Jason Freed and David Heinemeier Hansen. Well, last year in an earlier podcast, Jason mentioned we were working on a free version of Basecamp. That version is now out. So I thought we'd talk about why we're doing that, why we're offering the product for free.

something we used to do years ago, stopped doing it and brought it back. So I thought we'd talk a little bit about that decision-making process and why we're doing that now. So let's kind of go back in time a little bit. Jason, I know there was a free version, I think in 2019, if I'm not mistaken. Tell me a little bit about that and why we decided to stop offering free and then offer it again more recently. When we first launched Basecamp, we had a free one project plan. Um,

I believe, right, David? That's right. We had a free one. Yeah. So we had three pricing tiers and we had a free plan and it was one project. And that was huge for us because as much as we don't want to admit it sometimes like free sells essentially, you know, you're not making anything on it. People want free stuff. They want to try free stuff. They want to try something obligation free. They want to try something price free. They just want to give something a shot with no strings attached.

And we all want that, basically. Even when I'm looking at an app in the App Store, I'm like, it's $2. Let me just try it for free. It's $2, but I still want it for free because I don't want it to be a transaction. I just want to try something. That's all. And so we had it for a while. We had it for some more time. We went away from it. And then we recently brought it back, as you mentioned. It actually stemmed from a conversation we were having with our new head

head of marketing, he used to be over MailChimp and he was just telling us like what really kind of changed our trajectory was offering a free option, a free plan on MailChimp. And we've been sort of tossing this around internally for a bit. And I think that kind of threw us over the edge, like, yeah, of course, of course we should offer something for free primarily because for us, and we've seen this in the past is that when we, when we lose free, we might pick up more paid accounts and

But over time, we're exposing Basecamp to fewer people. And Basecamp is something, since primarily all of our growth is organic, we need more and more people to be exposed to it. It does a really good job of selling itself. But if you don't try it, you don't use it, it can't sell itself. So we want more people to try it, use it, explore it, be invited to other people's Basecamp accounts, which is why we have in our free plan, we have a one project limit, but you can add dozens of people to it.

And the idea is, of course, to sign up more people and get more people to see it and try it and feel it and get a good sense for it. And then some of those will convert to paid and others will never convert to paid. And that's also just fine because in a sense, they're still doing the job for us of getting other people exposed to it. And also, look, if you're a really small business and you can run on one project, good for you. Like we don't need to make money off of that. We're happy to provide it as a free service ultimately. So that's what we decided to do. And the idea, of course, was we need to like see significant

change or delta in the number of signups we're getting. If we're getting the same number of signups and now we're cannibalizing our paid accounts, like that's not going to work out well. So the bet, of course, is that enough people sign up for free, convert to paid eventually, we'll still get some paid signups, but overall, we're going to get more people to see Basecamp and that's going to pay off in the long term.

What's so funny about that thesis is that when we launched Basecamp in 2004, that was kind of the revolutionary thing. Freemium was a big way of turning things upside down. There'd been shareware, there'd been other things where you'd get sort of a trial of sorts. But this idea of just giving a fully functional product

in some limited capacity, but still fully functional product away for free was kind of a big deal at that time. And I think as Jason said, it really did benefit us tremendously in the early days. There's just this sense that you're not ready to buy. Like, I'm not ready to buy. I want to try the whole thing. And

When we say we went away from free, what we always had was a free trial. We always had at least 30 days. We had tried to experiment with 60 days and other periods, but there was always a free trial. So it wasn't like we were going to ask you a credit card on day one, but there's still a huge difference between those things. At least psychologically, am I even going to invest in trying this thing out if I know it's going to expire shortly thereafter? And then I

I'm okay at this point admitting that I think we got a little hubris going for us in 2015 was actually when we yanked the free plan. We introduced the $99 fixed one fee for everything approach and

Some of it was perhaps like Basecamp is well enough known. We've been around at that point for over 10 years. Everyone knows Basecamp, right? Well, maybe in some segment at that time, that was sort of kind of true. But then the thing,

happens that if you're around for another 10 years, there's an entire cohort of people who never heard of you. They literally were 10 years old when you did your thing, and now they're 20, and now they're in the market for a product management tool, never heard of Basecamp because it just did get less distribution. And we knew that when we did the pricing change in 15, it was very clear that we're going to take a distinct haircut in terms of signups,

But it was going to pay for itself because more people are going to convert to free. And I think we talked about this before, but this was a very rigorously managed experiment. It was probably the most rigorously managed experiment we've ever run for

For six months, we experimented with the pricing change and the dropping of the free and all the numbers added up within a timeframe of six months. Now we have a timeframe much longer. Now we have multiple years. And that's where you can pick up trends that aren't apparent in an A-B test.

Right. You're not going to run an A-B test for five years. That's not what businesses do. They want an answer much quicker. We thought already the six months we'd been running was an incredibly long amount of time that we were so sure we got this right because six months said it was right. Yeah. OK, but what about two years? What about three years? What about five years? What about 10 years? And I think that accumulation of the fact that far fewer people got exposed to Basecamp did come back to bite us a bit.

But you know what the great thing is? You can just fucking change your mind. You're just like, oh, we're going to do this now? And now we're going to do something else. And I think actually...

to some extent, if you live long enough as we have for more than 20 years, 21 years of Basecamp, you're going to flip flop on some things, both because you were either foolish or right the first time around and you couldn't tell the difference. You couldn't tell what you were foolish about. You couldn't tell what you were right about. And then you tried something else and you sort of kind of found out. And then the other thing is that the market changes. I mean, our brand awareness, for example, in the early days of Basecamp,

was a lot higher in the sense that we didn't have any competition. Now we have a lot of competition. So the calculus is different. And if you are stuck with all your decisions made literally 20 years ago, you're delusional if you think that all of them are right, either at the time or right now. And I think that's...

what we've sort of come around to. And I think the early results are in, right, Jason? I mean, we've more than doubled the weekly signups that we have. So that part at least is working. More people are getting exposed to it. And even the early conversion numbers looking nice. We actually just yesterday, I was asking Ron, our head of finance, who now also sort of runs our A-B testing on this kind of stuff, like, where are we? And it looked like we were ahead by like 3%.

which may not even be statistically significant at this point, but also it doesn't matter. The whole point was we're not doing this to juice the paid rate this moment. We're doing this for the two-year run, for the three-year one, for the five-year run, making sure enough people see Basecamp. I actually was going to ask you that, how often you're looking at those numbers. I know we're not like a huge analytics office here, but how often are you looking at those numbers to see if you can spot trends? Yeah.

Currently weekly is how it's presented. I mean, we can look at the analytics more frequently. I've been looking at them a little bit more frequently now.

For other reasons, we're actually trying a test right now just to switch the order of our plans. So if you load up Basecamp.com slash pricing, you might see it in the order of free plus pro unlimited or pro unlimited plus free. So we're kind of seeing what the order, and I've just been very curious about that. It's like instant data to look at. It turns out putting your most expensive plan first turns out to be better for getting people to pick that plan, whether or not it's

ends up converting. We don't know, but it does change the mix of which plans people choose. So I've been looking at that, but as far as the weekly breakdown on signups and conversion, we're doing that weekly. And it's a nice way to look at it. Otherwise you can become obsessive about it. And frankly, day-to-day doesn't really, you're not going to glean anything day-to-day. The other thing I'll tell you is that when we did some A-B testing on the marketing site homepage, we ran three different variations of

For the first, I think we ran it for about 21 days until they became, ultimately they were not statistically significant, but it took 21 days to find out that they were not. Things moved a lot actually during that period of time where certain designs were winning essentially. And then it turns out they came in second or third place. So if you look at things on a daily basis, you're just gonna drive yourself crazy. You're better off kind of looking longer term just so you don't look at the day-to-day fluctuations. Anything can happen on a given day.

What's so funny about that is it's almost irresistible when you have something you want to find out. You want to peak early. You want to roll forward in the book and like, do we have the ending yet? And then the danger is that you get excited about those early results, right? Or you get disillusioned. You're like, oh, it doesn't work. Well, give it a minute. Give it five minutes. This is where we just have a data analytics person, Noah, who is always on like, don't look, don't look.

Don't look at the numbers yet. Like just whack our fingers when we try to peek at the data early. It's too early. You're getting excited about something there isn't. And I think you should sort of try to have your cake needed to. It is irresistible to look at numbers and it's okay to peek. Just don't like make any rash decisions on those early numbers. It's fine to get an indication because sometimes it's not that close, right? Like there is a significant advantage, but oftentimes it is close and you really do need to wait because otherwise you're getting excited for nothing.

But what I also find interesting is that that AP test we're running right now, aren't humans just fascinating? Why the hell do humans care whether the most expensive price is first? There aren't that many. We're just shuffling it around. And what I find so fascinating is it's not like you're affecting one individual human. You're affecting a group of humans. There's something in the hardware up here that's just pattern matched.

to something's working. And then you try to derive things from it. And as Jason said, well, it turns out to presenting your most expensive plan first is the thing that works. Yeah. In the context of this site at this time in this category, I can guarantee you there's some gaming sites somewhere that go like, hey guys,

We figured it out. You put your most expensive plan at the end. It really converts so much better. And that's one of the, I think, most important lessons of all this A-B stuff is that so little of it transfers. It doesn't transfer between industries. It doesn't transfer between sites. It doesn't even transfer over time. You can run the same A-B test

six years earlier, and you may get a totally different result six years later because the market changed, because the product changed, because the customers changed. So I think the most honest thing to do is just accept that you don't know shit and you need to ask reality for an answer at a regular cadence. This is one of the things I think I was quite liable to fall into this trap. Oh, we've already tested that. Like I remember a result from six years ago. We did this test and that said X. Yeah, then...

And then I think I forget what the last one was that really humbled me, where I thought we had the answer. And then we ran a test. Three users free or something? Yes, maybe that was the one. Like we were trying around with the different number of free users you should have on a promotion set and we did. And I thought we already knew. I thought we had an answer that I could dig out of the archives.

And the thing is, it's just it's easier to be dumb in some ways. Like you can be less clever and just ask the market if you have enough traffic as we do. We have plenty of traffic to run these tests in a reasonable amount of time. You don't need to speculate so hard. You don't need to come up with all these theories about why things are the way they are. You could also just like ask the market, would you like to pay more for this order? And the market will tell you.

Okay, tell me this because I would imagine that free isn't really free. Like when we're giving away product for free, there's some cost involved, whether that's support costs because there's more people trying it or server costs. Tell me kind of your thoughts about that and what free really costs us.

Or maybe it doesn't. It does. It does. I don't think we analyze it so closely. I think primarily a lot of the cost is around support. Like we have a good sense of what it costs to support a customer if someone emails us.

So if you have more customers emailing you, it costs more. And if they're not paying you, it actually costs way more in a sense, because now you're technically losing when before you had margin that you had some room where, you know, someone emails you three times, you're still okay that month. If they, if they email you 18 times, maybe you lost your money. If we're giving something away for free and they email you twice, like you've lost for sure. But I don't like to look at it that way because I don't think someone asking us a question is a loss. I think it's a gain.

I think if we can present a really wonderful experience to somebody, if we can help someone in a way they were shocked by, because most companies suck at support. I actually, in many ways, I want to invite people to email us and ask us questions. And so I don't see these moments as losses. I see them as gains. Obviously, technically, if you were just to count the numbers, perhaps they're losses. But I think there's way more to it than just doing math. Math is not what business is all about.

Uh, some of it for sure, but there's way more to it. And there's a lot of stuff you just can't measure, but you have a feel for like, if someone comes away with a good experience, how do you actually measure that? You don't really know, but maybe they tell someone else, maybe they stick around. Maybe they just put a smile on their face that day. Maybe that's worth something. And so I feel like, you know, that's the bet you basically make an all in the aim for us is like all in out of everything we're doing. Are we coming out ahead in a given year?

That's how I've always looked at things. And it's super nonspecific. And I'm sure there's lots of people who go, that's a crazy way to look at it. But for me, it's like, out of all the things we're doing, are we coming out ahead? That's enough. And so, yeah, we're losing on some, we're gaining on some. As long as we come out ahead, I almost kind of don't personally really care.

care. I know to some degree, but I also don't know completely what matters, what doesn't. Obviously, if we're charging people a lot more money here and giving something away for free, the people we're charging are paying for the people who are free. I get all that. But

I don't want to look at a free customer and go, they're costing us $7. I don't think that's a good way to look at things personally. Unless of course you have no margins, you're losing money already and you're not making any money. And then you've got to really figure that stuff out. But if you have some room, you can be a little bit more, I think, open to just sort of, you know, playing fair in a sense. David, do you look at it the same way?

Yeah, this is one of those fundamental delusions of traditional business management. This is why MBAs have such a bad reputation. It's why no one fucking likes bean counters as a stereotype, because the only thing they can count is the fucking beans. Well, there's a little bit more to it than just the beans. There's all the stuff that isn't visible, as Jason says. How do you measure a good experience? How do you measure a free customer who gets a

surprisingly good customer service experience, then tells five friends who then signs up. Yeah, there are ways you can sort of try to approximate some of that word of mouth stuff, but it's always going to be lousy. It's always going to be full of gaps. And I think to some degree, if you have the margin, you also develop the confidence that like, we're just going to do what feels right.

And you know what? In a long enough time frame, if you have margin and you do what feels right, you know, it's going to work out. Right. Like the margin alone part, make sure of that. And then you can just enjoy yourself more, which is like we don't have to count everything. And this is something I've sort of had to internalize over the years. I used to be a little more on the bean counting, just the beans side of it.

But I've really come to appreciate this fact that many of the most important things, not just in business, but in life, you don't get to count because they're not quantifiable. They're not even specifiable. You don't even know what you don't know. So you can have this general theory. Do you know what? If we give really good service, I bet that's good for business. How am I going to prove that with a statistical rigor? That's actually quite difficult. Now, caveat, that's Basecamp.

Hay? Hay is actually a bit different. When we were on the cloud, especially with hay, I calculated that every damn customer cost us $30 a year to maintain just in terms of our operating costs for the servers and so forth.

And that's in part because email is very different from something like Basecamp. It's just a massive amount more data. It's per user. And especially when you run it in the cloud, it's really expensive. Now we've gone out of the cloud, so we've dramatically lowered the cost per customer. But it's still not insignificant. Like one additional user account in Basecamp...

costs so little the vast majority of the time that it almost isn't worth counting. One hey customer who actually uses the system legitimately as their primary email does cost real money. So sometimes it also depends on the kind of business it is, the kind of application it is. And

Basecamp is just a little more amenable. There's just more margin in the operating costs of it. Also, because we charge more, pay is $100 once a year versus Basecamp is very frequently $100 a month or at least $50 a month. There's just more room in that number combined with the cost that we have it that we get to experiment more. So that's

That's why you always got to build this stuff around. You're sitting listening right now and you're like, oh, I should just do more stuff. I should give more stuff away for free. Yeah. Not if you're like a 4% margin. If you're Walmart, you really do need to count like the plastic. Yeah, it'd be kind of more nice if there was more plastic around. Yeah. Have a bean counter. Count the plastic beans. Right. You can't at that margin just...

run and shoot from the hip. But that's exactly why you must do so if you have the margin. It's actually an insult to people who have to count the beans. If you're in a position where you don't have to count the beans and you're like, I need all the beans counted all the time. Okay. Last question before we wrap up, and you might not have an answer for this one yet. I know we've talked about two new products that we're working on. Do you imagine there'll be free options with those? Have you even gotten to start thinking about pricing yet?

Yeah, a little bit. Yes, I would imagine. I mean, don't quote me or hold me to it because I don't know. We haven't decided. But ultimately, yes, I think that, you know, this lesson applies, especially with something that's new.

You want people to try it. And I think we're making some things that are kind of novel. So you definitely want people to get in front of it and try it out. We might even be just extremely generous with one of the products in terms of categories of free usage. I mean, there's a variety of ways to approach this. It could be everyone gets something for free, or it could be this type of business gets it completely for free, or it could be a first year for free. I don't know.

But certainly we're going to explore that for sure. I think coming out of the gate with just like, you can't even try this unless you pay for it or you only get 30 days to try it. I don't think it's going to work. Now, one of the other things that David didn't mention, but is related to Hayes, part of the other reason we were careful about Hayes because you actually get an email address. So-

If you're a free user and you take an email address, that email address is kind of yours forever in a sense. And so we didn't want the market flooded with people just taking email addresses from the namespace and then never using the product. So that's another reason why we say you only get to keep your email address if you pay for it.

Even if you cancel out for the first year or cancel out for the first three months, you get that email address permanently. But if it was free right off the bat, free trial, it'd be kind of tricky to do that. So there's other issues with products and deciding what you can give away and what you can't. But these other two products aren't like that. They're more like Basecamp in terms of the way we could give them away for free. So we'll see. We'll see how it plays out.

Yeah, and then everyone would have a Kimberly at 8-7-6-5-2. Right. Like you do on Gmail. Yes, exactly, exactly. Okay, well, with that, we're going to wrap it up. You can find your free Basecamp account at Basecamp.com. Rework is production of 37 Signals. You can find show notes and transcripts on our website at 37signals.com slash podcast. Full video episodes are on YouTube. And if you have a question for Jason or David about a better way to work and run your business, leave us a voicemail at 708-628-7000.

7850. You can also text that number or send us an email to rework at 37signals.com.