Mike McDerment changelog.com/posts

The many reasons replatforming FreshBooks should have failed (and why it succeeded)

This is a fancified excerpt of FreshBooks CEO Mike McDerment on Founders Talk #68. To get the full experience you should listen while you read.

🎧  Click here to play the episode starting where Mike begins  🎧


In 2015 we were dealing with a few technical problems at FreshBooks. One of our biggest challenges was our application architecture. Our separation between back-end logic (almost middleware) was baked in with our front-end — this wasn’t very clean. So the root cause to fully replatform was really because we believed the consumer expectations for FreshBooks had changed so much — because of mobile, and because of other advances — that if we wanted to win long-term, we were not going to be able to simply refactor our way to future greatness.

We had to get creative. So, we started out and said…

If user experience matters, ease of use and simplicity, we’ve won on that today. But it will be harder and harder for us to win and win again, and win long-term, if we don’t do something radical [application] design wise.

We needed to design where we needed be to get past these technical problems. Could we get there from where we were at? The short answer was, “Not really.” So we decided to replatform.

The problem with replatforming

When we decided to replatform it opened up a whole series of other problems for us. There are questions when you replatform. If you read Joel Spolsky, or what have you… You just don’t do it. You’re bound to fail.

Replatforming is the number one strategic mistake you can make as a software company.

So, replatforming is severely frowned-upon, and I think there’s good reasons for that.

First of all, it’s gonna take longer and cost more than you think. You may never finish. That is a very likely scenario. And actually, we tried twice before, and failed… So it was the third time that we kind of got it done.

My favorite reasons are maybe less obvious and more business-oriented, but you could build another version of your offering, and it could fundamentally perform worse from a business standpoint, and you wouldn’t even know.

What happens is, inside the building – you’re building a new platform, and your team is falling in love with doing the new thing… But that doesn’t mean the new thing is better, or more business-performant. And that should be the reason you’re doing this stuff.

Technology, as much as we all love it, is a means to an end. If you’re making the necessary investments to make the platform, the thing better perform better. So you wouldn’t know that for sure…

Then there’s the good old…

Hey, your competitors are still moving, while you’re standing still, replatforming.

That’s not good. They’re catching up if you’ve got a lead. Okay, that’s bad, because it’s probably gonna be a multi-year thing if you’re relatively established. We’re number two in America for small business accounting software, so it’s hard to do that…

And then finally, my favorite one - I like to call it the sophomore jinx. If you’ve ever had a band that you love their first album, and you go buy the second one and you’re ashamed to be a fan, because it’s so bad. It’s like…

Man, I don’t wanna be one of those.

I think that presents itself when you go and redesign a new thing. There’s no guarantee your customers are gonna love it. People are inherently afraid of change, so things are kind of stacked against you.

What if we secretly created a new competitor?

If we were going to replatform, we had a whole bunch of problems to solved for. How do we measure it and know it’s better? How do we remain stealth while we’re working on this, so our competitors can’t find us?

I was chewing on all these problems and came up with this crazy idea one weekend…

What happens if we create a company that has nothing to do with us? No one could track the two companies together. And we could use that new company as a petri dish to build the new FreshBooks.

What we would get out of that is we could build something out of our competitors’ eyes, or even our customers’ eyes, frankly, which is helpful in a lot of ways. We could scale it up to know if it was actually business-performant and better.

Because it wasn’t our brand –and I think once you have a brand and people trust you, you really cannot take as much risk– a really new, fragile, embryonic technology - there’s big risk in that, whether it’s user experience risk, or production availability, or data loss risk… All these risks are there.

So if you’re flying under the banner of your existing brand, it gets very hard to take the kinds of risks that lead to something really innovative. I think that’s why actually a lot of big companies suck at this stuff. It’s one of the reasons; there are others… But it’s less talked about, and super-important.

I really wanted to have an environment and a set of conditions where people can take risks. So by creating this new company, which we incorporated on its own, had its own website, its own brand name, did everything over there.

We ran it actually live for like 9-12 months before determining it was ready to switch it over and started moving some people over. That was the approach we took, and that’s why. That was the context. But yeah, it was definitely a different way to solve the problem.

Making the switch

And then I’ll go further and say then we had this other problem of…

Okay, we have two platforms. People don’t like to switch…

We believe in the four E’s, which is Executing Extraordinary Experiences Everyday. We have a very customer service and customer-focused orientation as a business. We recognized that a lot of migrations go badly because you’re building a platform and you force everybody to migrate, but that platform’s not ready. As much as you think it is, and your team is in love with their new thing, I can tell you - the new FreshBooks was not perfect for everyone day one, at all. So it would have been tragic if we’d forced everybody over. We didn’t do that.

We actually ran both platforms and let our customers choose when to move over, and if they wanted to move over.

We thought that was the right thing to do, because we are a serious business software, we are playing a core role in people’s lives… And having that choice we just thought was empowering to them, and would help build trust. We actually helped people roll back if it didn’t work out the way they wanted it to.

So all this – we tried to take this challenging thing that is usually a major trust-mitigating event and find a way to design the experiences so ideally it’s trust-building or neutral.

We had lots of people who were super-excited about the new FreshBooks. The vast majority.. But we definitely had some people who were like…

This is not what I had hoped.

And that might have been two years ago. And now they’d be like…

Oh, it’s better than I imagined.

This is the great thing about software. So I think not only was creating a new company novel and different, our approach to migrating customers from one platform to another… I’ve never heard of anyone else doing that before.

I think the team — for executing the complexity in those things, and conceiving of them — deserve a lot of credit.


The conversation doesn’t end there. Listen to the entire episode of Founders Talk #68 to hear more from Mike McDerment about his journey and the challenges he has faced leading FreshBooks. You can play it from the start right here 👇

Oh, and don’t forget to subscribe to Founders Talk in your favorite podcast app so you don’t miss future stories from founders and makers. ✌️


Discussion

Sign in or Join to comment or subscribe

0:00 / 0:00