Havoc Inspired

FreeAgent

Roan Lavery is the Lead Designer and co-founder of FreeAgent, the best online accounting software out there (note: I do use the software so I am bias but it is pretty damn good). In this interview Roan takes me through how he made the transition from Creative Director for a web agency in Edinburgh to running FreeAgent as a full-time business. If you’re building or run a web app yourself you’re going to really enjoy this interview.

And if you fancy trying out FreeAgent you can use me as a referrer: Referreral Link

Roan: My official title is Lead Designer, but what I do day to day is quite varied. Being the founder of a company, I’m having to do a lot of different things, and I wear a lot of different hats every day basically.

Ryan: OK, and of course you’re talking about FreeAgent. Do you want to tell us a bit about what FreeAgent is?

Roan: Yeah, sure. It’s an online accounting application specifically designed for freelancers and small businesses. So it helps them manage their money, does their invoicing for them, helps them track time and expenses, integrates with online banking, and then it works out the accounts and does tax predictions as well. So it’s really just all about the financial side of being a freelancer, being a small business, and sort of how we can help with that.

Ryan: How long’s FreeAgent been running?

Roan: We founded the company (myself, Olly and Ed) in the start of 2007. There had been a little bit of work done before then, with prototypes, but the company sort of kicked off in 2007. We started up just sort of making a prototype, getting something vaguely working, and we launched a beta version of the app in the middle of 2007 and ran that over the summer. So we had a couple of hundred people just trying out the app, hammering out, telling us what was right, what was wrong with it, and we were able to launch at the end of 2007 commercially. So we started charging for it then, and we’ve been running it commercially for over a year…over a year and a half now.

Ryan: And you kind of worked on the app partly in your spare time, and then you’ve took the plunge and gone full time on the app, I mean how long…

Roan: Yeah, all of us had full time jobs when we first started working on FreeAgent. I was working as the Creative Director of a web agency in Edinburgh, and Olly and Ed worked in sort of contract consulting, so we all had jobs to do. FreeAgent was very much a side project for us, and we didn’t know if it was going to work. We though it was a quite a cool idea, you know, there was nothing really focused for freelancers, so we thought we’d kind of keep building away, and we did that basically until the end of 2008. All that time, we were building this project in our spare time, trying to get it going, and the app was building, coming along, getting users on board, we were making a little bit of money, and it got to the point where it was becoming more and more difficult not to treat this as a side project. It was obvious that it was going to work, but it needed a proper commitment from all of us, and when you’re trying to do it in the evenings and at weekends – all of us have got young families, so we had to try and juggle that around having kids – it’s really tricky, and something had to give, and that something was our day jobs, basically. So, we took the decision that we were going to jack in the day jobs, and we took a little bit of investment money that would help tide us over until we thought the app would be generating enough revenue to pay us all a salary, and just went for it basically. We went for it about the end of 2008, and the start of 2009 was when we all went properly full time.

Ryan: So at the end of 2008, was it just the three of you that moved onto working on it full time?

Roan: Kind of. We’d be working with our outsourced sales guy, who was helping us with accoutants, getting them to pass FreeAgent onto their clients. So we brought on board the sales guy around the end of 2008 to help with that, and it was the beginning of 2009 when we got our first small office, with me, Ed and David (our sales guy) kind of seconded in there. But over 2009 was when things started really taking off for us, we started getting a lot more users on board, and we were just being able to bring more people on board into the company.

Ryan: So what’s an average week like for you?

Roan: Well, it’s a little bit different to when I was previously working at an agency, because obviously when you’re a designer in an agency you’re getting client briefs, you’re responding to them, you’re maybe doing mock-ups and front-end templates and stuff like that. When I’m working at FreeAgent, there’s less about designing a layer or something like that, there’s more thinking about “let’s look at a new feature in the app,” so how does that work within the application? Where does it go? What’s the user workflow? What’s the user experience of this? Or, how can we take an existing part of the app and make that work better for people? So it’s a lot of wireframing and talking to the developer guys, seeing what’s easy to do, really working quite closely with them, and just implementing those new features and making the app as polished and as easy to use as we possibly can. There’s a fair bit of my time spent doing that, and other time spent…like it’s my responsibility to track web analytics, and be on top of the SEO or the PPC campaigns that we do, so doing a fair bit of the online marketing side of things as well.

Ryan: So rewind a bit, let’s go back to the start of your career. How did you start out?

Roan: Well, I did a degree at university in physics, and when I finished that I didn’t really know what I wanted to do. I didn’t want to get a job for one thing, so I did a PHD and ended up staying at university for another 4 years, and I quickly realised that I didn’t really enjoy being an academicm, but I wanted to stick out the PHD because I thought I’d started it so I might as well stick it out, but I always knew I was going to be doing something else. During my last year of doing my PHD, I wasn’t getting any funding because I’d run over my grant, so it took a year to write up, so I had to sign on – I was on the dole for a year. By the time I actually finished my PHD and got my thesis written, the dole were getting pretty stroppy with me, wanted me to get a job. So they suggested I go to this course, which was a web design course, which I thought was great because I’d been interested in web design when I was at uni but never really got a chance to pursue it, so I ended up going on this short course, this 6 month course teaching web design, and I loved it, it was an absolutely great course. I ended up staying with that company for 8 years, because after the course they took me on as an intern, and I worked there a bit just learning all about web design. Then I ended up, a little bit later on, teaching the course that I had been on, and that company did some training and also had a web design side to the business as well, and basically over the years I went from working on the training side to working on the web design side, and I ended up running the web design side of the business and leading a small team of developers there. That gave me a huge breadth of experience, everything from training through to designm, front-end build, online marketing, co-operating – because it was a really small agency, I got a chance to try out all the different aspects of it. So I was there for…that was where I was up until I started FreeAgent.

Ryan: So, one question I ask a lot of people is: what are you most proud of in your career? Do you have something you look back on and think I’m really proud of what I achieved there?

Roan: I think FreeAgent’s the one I’m definitely most proud of. It’s not a written story by any means, there’s still a lot more for us to do with FreeAgent, but I think quitting your day job and working for someone else, starting up something which is your own project and trying to see that through…you’re always reading some blog articles about people saying how you should start your own side projects up, and you get a lot out of it, and I totally relate with that, but being able to take it and actually turn it into a successful business, where in a day you get people emailng and saying “I love your application so much, it’s transformed the way I run my business” – getting that feedback is incredibly nice to hear and it does make it all worthwhile. So, hearing that is something I’m really really proud of, you know, I’m proud of what we’ve managed to achieve with FreeAgent, and I still think there’s a lot more we can do.

Ryan: So from building FreeAgent, what lessons have you learned that you would tell other people about – what advice would you give about starting a successful app that you could do full time?

Roan: Yeah, I think the first thing I would say is: just do it. It really is an incredibly worthwhile thing to do, so if you’re at all thinking about an idea, if you’re considering a start-up, I would really recommend it as a worthwhile venture. It’s a really rewarding thing to do, even if it doesn’t turn out hugely successful, you’ll learn a massive amount in that process. In terms of how you make it successful, it’s a difficult thing. I think have an idea about how you’re going to make money from it from a very early point. Even if you don’t make money from it straight away, at least have an idea for a business model of how you are actually going to make revenue from this, because if you leave it until much later down the line, I think you’ll find it quite tricky – it’s hard to retrofit something like a business model, despite what Twitter might tell you. That would certainly be one of the things I would advise.

Ryan: OK, and what mistakes have you made that you would tell people: avoid that completely?

Roan: Well, I’ve made loads of mistakes. I think the biggest one is…it’s quite funny because you read a lot about people doing web apps and things like that and they always say “build an application for you, build it to solve a problem that you’ve got”. Now I totally do agree with that, but there is a flipside, and that is if you build it for the way that you work, and it just so happens that everyone else works a little differently to you, then you’re kind of screwed. That’s what happened with FreeAgent – there’s a lot we got right, but the stuff we got really wrong with FreeAgent was making assumptions about the way other people would work based on how we worked, and those assumptions were wrong. So, the kind of key thing is that you build something to solve your problem, but get it out into other people’s hands as quickly as you can, and realise where you’ve screwed up quickly. So, get other people using it, even if you don’t think it’s that well polishedm, get it out there and wait for them to come back and say “er, this is wrong, you’re going to need to fix this” or “this is confusing” and then just kind of improve it from there.

Ryan: What do you do to manage that feedback, because obviously you must have thousands of people using FreeAgent, how do you keep track of it all and prioritise it all?

Roan: It’s really difficult. That is I guess the other main strand of what I do, it’s really just kind of like keeping your ear to the ground and seeing what’s out there. We use a web application called GetSatisfaction which is a community support/discussion forum application. We were one of the first people to use that, we started a couple of years ago, and that has been incredibly useful for us just to gauge what people want, what they like, what they don’t like, and then they can start discussions with other users and things like that. In the early days, it was really good because we could be incredibly responsive to people. They could say “what about this feature?” and we could say “no problem, we’ll have that implemented in 3 week’s time” or something, and now it just isn’t like that. So as an application matures, as it gets much more complicated and you have to worry about scaling and things like that like we do now, the management process of customer feedback and the roadmap is much more…delicate, shall we say.

Ryan: And being responsible to your users as well, the more people who are using it.

Roan: That’s right. There’s a much bigger support burden on us nowadays because we’ve got a lot more people using the application, so it’s about managing people’s expectations, but at the same time making sure you are totally engaged with your users because at the end of the day you’re making it for them, and they do know best, so using GetSatisfaction, Twitter’s another thing we use a lot, we’re always talking with our users on Twitter as well. So there’s a lot of these different tools out there but the key thing is just to make sure you listen to people, just make sure you listen to them. Even if you’re not saying you’ll put it in next week or next monthm, put it to the back of your mind and then if somebody else mentions it and then somebody else mentions it after them you can start to say “actually guys, this needs to be a priority for us” and we can start bumping it up the roadmap.

Ryan: So what are you most excited about at the minute, in the industry?

Roan: I think there are some really good things coming along. As a front-end guy, the fairly rapid adoption of HTML5 and CSS3 is quite an exciting thing, and people are starting to do some really crazy things just using CSS3, I saw some Andy Clarke stuff recently which was pretty mental. I think that stuff, as a designer, being able do do those things easily without having to resort to stuff like Flash is really exciting. We’re really looking forward to rebuilding FreeAgent as an HTML5 app and some of the stuff that’s functionally going to bring us as well, offline storage and stuff like that. I think that really excites me as well.

Ryan: OK so, I’m digressing because you’re an application man and I’m helping build a web app, but what do you do about browser support? Do you have graded browser support for FreeAgent and things like that?

Roan: A little bit, we’re not quite at the Yahoo! standard or anything like that. Because of our customer base, because of our user base, we don’t really have to worry about people using IE6 or anything like that, or we can say “no, we’re not going to support that” and that’s a decision we’ve made. So we’ve said we will support IE7 and above. You do need JavaScript to run FreeAgent. I think that as you get more and more involved with web apps, it becomes really really difficult to support a non-JavaScript version of the app for closed applications. We tried, in first year of running FreeAgent, we really tried to make sure we always had graceful degradation, but eventually it’s just become too much of a development overhead to maintain, so we had to make the decision that it was going to be a JavaScript-only application. And for our user base, which is really technically-able people by and large – it is web designers, developers, IT consultants, stuff like thatm, it’s not a problem. As for the visual stuff like using CSS3, I think it’s fine that browsers like Internet Explorer aren’t getting the full polished package, that’s absolutely fine as long as the app looks good in those browsers too, I think it’s fine.

Ryan: I agree. It’s one of those things when you get to a web app it’s a controlled environment isn’t it, and you can make what’s going to happen once you get into the application completely clear before they sign up, so I think it’s to be expected that you’re going to get JavaScript. Where do you see yourself in the future? What are your plans?

Roan: It’s a difficult one, because if you asked me that 2 years ago, I would have had fairly modest expectations for what we were doing with FreeAgent, and I think we’ve surpassed those expectations so…I mean we’ve got our own plans for where we see our business in a year or two’s time in terms of how many people are going to be in the company and how many users we hope we’ll have and things like that, but in terms of where we’ll end up, I don’t think any of us really know. We’re so focused on the next…for example the next soft release, which is next weekm, so all of our concentration is about how we can make sure the next release goes out and it’s as great as it possibly can be, so I think a couple of years down the line is different. For me, having come from a client/agency background, and now coming into application design, I absolutely love application design. It’s a much more challenging problem, rather than a lot of the time when you’re doing websites you’re kind of churning out websites that are not of a similar look, but certainly of a similar level of functionality, so another content management site, or another event booking site or something like that, and there’s a lot of repetition. When you’re doing a web app, you’ve quite often got some really difficult challenges and interesting problems, so as a designer they’re fascinating to get your teeth into, and working with the community, working with your users, getting their feedback and stuff, it’s a really good thing. So, I’ll definitely be involved in application UI in some way.

Ryan: And being able to go back and fix things that you’re not happy with, which you often can’t do with websites.

Roan: Yeah, that’s an amazing thing, I mean there was a little while back where I spent an entire day just designing our login page, and you would never have that luxury if you were doing client work because you’ve got deadlines and budgets and things like that. Actually we’ve got a new designer come on board, and I’ve said I really want him to have the ability to just pick apart bits of the app you don’t like, and actually spend a bit of time redesigning that and see if we can make it better, and don’t worry about deadlines and things like that, just have their freedom to say can we approach this whole thing from a different angle, and that’s a really nice thing to be able to do.

Ryan: OK, well thank you very much for taking the time to talk to me Roan, that’s really interesting, and I’ll speak to you soon.

Roan: Great, speak to you soon.

Thanks goes to David Goss for transcribing this interview.