Paul Farnell is co-founder and CEO of Litmus.
So what’s your background?
I went to uni at Manchester Met and throughout that, as well as a bit before, I was freelancing in web design.
As part of doing web design for clients I would need to test them between browsers. This was back when there were much greater discrepancies between browsers. The tools that were out there to automate, there was one called BrowserCam, were not well designed and expensive for a student freelancer. I’d been very inspired by the 37 signals people who had started Basecamp two or three years before, alongside running a web design consultancy.
My dream was to have something like that and establish a recurring income. They ran a workshop in Copenhagen about how they built it which was super helpful. So that was the inspiration. The need was there, and so I set about building it with my friends Matthew Brindley and David Smalley.
How did you get your first customers?
Being a web developer in the North of England, we went to a lot of meetups and made a lot of friends through the industry so I’d post the app to lists and because people knew me they would pay attention to it. We would just give logins away at first to get some feedback and then those that tried it out we would send them an email when we did launch with a discounted offer.
When did you start to take Litmus seriously?
It was hard to get dedicated time to build Litmus as we were still doing client work which was paying the bills. Client work can be very time sensitive, very demanding but they were paying reasonably well so it was difficult to turn it down to work on Litmus.
We’d seen a bit of traction but it had stalled and wasn’t growing as fast as we wanted. We were at about 60 customers and it was only $19 a month at the time so it wasn’t a lot of money. We had lots of ideas for how we could improve it but no time to do it.
I was living in Montreal at the time and one day I can remember really clearly coming up with this new way of laying out the UI. I sketched it out really quickly and said to my girlfriend “I’ve got it, this is it, I’m going to cancel all my client work and just do this”.
Since Litmus wasn’t making enough money to sustain us we took out a loan for £10,000. That was supposed to be enough to allow us to survive for four months and focus on it full time. That made a huge difference. In retrospect we probably could have done that a few months earlier but it was scary to email clients and say we’re not doing any more work for you and pinning our hopes on Litmus.
We aimed to break even after four months but we we didn’t. We ended up borrowing a bit of money from parents and putting things on credit cards and really living frugally until we did.
I understand your first server in Leeds was in a cupboard?
That was actually the upgraded server room. The very first server was in a dorm room at university on the university network. At the peak there were three PC’s and an old iMac under my desk. You couldn’t turn the screen off on the old iMac and it would wake my girlfriend up because someone in the US would run a test and the screen would come on and the fans would kick in full blast.
For the upgrade, we had advertised down time while David drove over to Manchester, put all the computers in the back of his car and drove to Leeds and then installed them in a cupboard in his house he’d just bought. I think we had to do it because I was moving out of halls. After that we went to a proper data centre but that setup lasted us quite a while.
What changed after you went full time?
It was around this time (2005-2006) we added email testing, obviously now it is all about email testing but at the time that was new. Email was technically harder to test but people seemed to value it more so we got more customers as a result of that.
Something worth noting around this time is this. We started off pricing in dollars, but we decided to change pricing to Euro’s because we thought, well everyone knows the Euro, right? It sounds so strange saying it now, but we changed all the pricing to Euros.
That really affected things, growth totally levelled off. We eventually got some feedback from the US, which was our biggest customer base. People just did not know what Euros were, they were seeing the Euro sign and thinking “Oh this isn’t for me”. So we switched back to dollars and immediately growth ticked back up again.
The next phase was in 2009, we had about 500 users and I moved to Boston from Leeds and we set up an office there, we hired some incredible engineers and support people. We were very lucky with the early hires and we were in an incubator like space which had an incredible energy about it. That opened our eyes a bit to what Litmus could be and how we could do this on a bigger scale.
The reason for opening this office was that we were selling the software to Email Service Providers (ESPs) and they were based in the US. So when this little company from Britain was bidding on these contracts we felt a bit like we weren’t taken as seriously and we felt an office in the US would help us get taken more seriously. We won a bunch of ESP partners through that strategy which exposed the brand to lots more people.
What was the next phase?
So this was about 2010/11 and things were starting to take off. But things went wrong technically, we had a lot of trouble scaling up the infrastructure that took the screenshots, enormous trouble. When someone used the service it was so intense on the backend and that was falling over as we were adding customers at the rate we were.
This was getting very stressful as we were getting negative reviews, bigger customers were signing up and results were just too slow. So thankfully we had some really talented new developers in the US who then spent many months rebuilding the whole thing from the ground up while we tried to fight fires as best we could. This meant growth levelled off as the perception began to grow that Litmus was slow and less reliable.
It took about a year to completely solve that problem. We also made a key hire in Justine Jordan to really beef up our marketing as before it was just me writing blog posts. She had a really excellent background in marketing and design and had worked in the email industry and was already a thought leader.
So two years after we moved to Boston, the system was fast and reliable, we had Justine on board doing the marketing and we were in a position where we could really focus on growing again. We had good strong growth for the next three years and gradually added to the team. We built out the product adding new features.
One of the important things that happened during that time was mobile email became pervasive, once the iPhone and Android were out people began to read full html emails on their devices in a way they’d never done before.
We had an analytics tool which would report what percentage of your mailing list were reading on mobile and when people started to see that mobile figure exploding they started to be able to make the case that they needed to redesign their emails for mobile which required tools like Litmus to test them so that movement was a big factor in growth. Since we had done all that work growing the team we were in a great position to be able to capitalise on that opportunity.
More recently, we took a late stage growth round of investment from Spectrum Equity of $49 million, and since that we have really doubled down on growth. We’ve doubled the team in the last 12 months.
Going back to those periods where you’ve had to rewrite the codebase, was it difficult to step away from the code and become more strategic and managerial?
By that time I wasn’t coding anymore but Matt was right in the thick of it with the two other engineers we had at the time, incredibly hands on. We ran into the same problem four years later though when that rewrite started to fall over as we got to even greater scale. We anticipated it much better that time so I don’t know if users would have noticed as much. That time, Matt was much more hands off. He would do little R&D projects to figure out if you could do something with Outlook, could you make it do this in 1 second etc. He’d make a scrappy little prototype and hand it to the team and they’d use the idea but not necessarily the code. He was much more of a guiding hand. I can’t speak for him but I didn’t get the impression that was difficult for him to let that go.
What is the current strategy?
When we first launched the product, our focus was just on selling to end users and what started to happen was we got a lot of enquiries from ESPs asking about an API so they could integrate the tool. As we didn’t have an API at the time we just put those people on a list and said, when we do we’ll get back in touch with you. We were getting so many of these, it was up to about 125. So we thought maybe this is something we should build.
It was quite a bit of work to build an API so what we did was we pre-sold it. We planned out how it would work and sent out an email to these 125 people which said we are launching an API. We then set out how it would work and the pricing structure. We also explained that there was an upfront fee of $5,000 and we were limiting access to the first 5 customers. We said we would actively work together with their development team to make sure the API was what they needed.
So we did the whole ‘launch formula’ thing where we sent out emails building up to the launch and saying when the page was going live. We really tried to build up some excitement about it.
There was a great moment where a guy who worked at an ESP in Australia, he was the first person to buy, sent us an email about how he had an alarm set because it was the middle of the night for him and he had jumped out of bed and loaded the page and he wrote to us saying how much he hoped he’d made it.
We didn’t actually know whether anyone was going to pay of course but sure enough, within an hour, we had all 5 of our slots filled which was incredibly exciting and a huge amount of money for us at the time. It was only after that we worked with them to build the API.
Our first hire was a sales guy, who is still with us, who we brought in to manage this process as it is quite a long process to get them set up.
Today, almost every major ESP is integrated with us but it represents a relatively small amount of revenue. It changes year to year but about 15-18% of our revenue is from that channel. It is a great lead source for us as it is branded within the app, but most of our revenue is direct customers signing up.
What is your proudest Litmus moment?
Probably about three years ago. We always do a trip away ever since we had our first employee and since we are remote it’s a chance for everyone to get together in person. I give a little talk about how things are going. I remember looking around that room and thinking “Wow this is a lot of people. They’re all spending their day working Litmus, something that I really believe in” that was incredibly exciting.
Does that come with a bit of pressure?
Yes, I would say there is a lot of pressure and I do feel it. That definitely grew as we got bigger. For instance, someone would bring their kids to the office and I would think “God, I hope this keeps going well”.
That was really getting to me before when it was just me in the senior leadership role, I would lie awake at night as a lot of people were relying on us.
We have investors and a board now and some great, very experienced people on the board and in leadership positions. Now it is harder for me to screw it up, there are a lot more checks and balances around so I am more relaxed about it.
In the same way that the demand for browser testing withered away do you see the demand for email testing withering away in the future? If so what’s next for Litmus?
We actually have a partnership with Outlook where Litmus users can report bugs to Outlook. We submit these to Microsoft and funnily enough the first fixes of bugs reported by Litmus users are now in updates to Outlook. We are hoping to build partnerships with other providers too of course to fix those problems.
Litmus will be successful in the future by making email marketers as successful as possible, at the moment there is a core offering that loads email previews but I’d love those users to be able to spend no time on email previews and have Litmus be a tool that lets them develop more effective, more exciting email campaigns.