Safe bets

4 September 2007

Richard Marshall's PhD in computer science has helped him build advanced mobile technologies - but turning them into a business spanning alternative theatre and gambling required very different skills

Richard Marshall is CEO of Rapid Mobile, which develops mobile applications and services. He started the company in February 2004 after selling the assets of a previous venture, and Rapid Mobile began what is now reckoned to be the largest turnover mobile service, Betfair Mobile, which was initially run as an own-branded venture before being sold back to Betfair. On the back of that success, the company won backing from Pentech Ventures and the Scottish Co-investment Fund, and has since fleshed out its technology base to include a mobile advertising platform, Ad360. The 18-strong company is now about to embark on a second round of fundraising, which it will use to bolster its commercial team and expand its interests abroad. Board advisors include Richard Eyre, deputy chair of GCap, and Andy Brannan, executive vice president of sales at Symbian.

You built a site for the recent Edinburgh Fringe Festival called Fringe Knowledge, providing details of shows, reviews and booking information. How was it received?

Well, as anyone who's been to the Fringe will appreciate, it's absolute mayhem! We were given the go-ahead to do it at the last minute, although we'd been planning it for a while because we knew it had to be live before the Festival started on August 5th. But it's a great reference and our Ad360 server enables us to put unobtrusive but unavoidable advertising on the service. That's a great way of paying for these sites because nobody likes paying for mobile services. But it's still a confused market and there are only a few early adopters, like the Fringe. The network operators themselves still think they can do things like running content portals themselves, even though they tend to lose money on them.

But the first application you developed wasn't around advertising. So how have your business plans evolved?

Our ideas have been pretty consistent since we started the company. Our aim was to develop highly-usable, portable, secure, transactional mobile applications - by which I mean things related to money - so people could take the things they do on the web onto the phone. And in particular, so the service providers could be safe in the knowledge that the person on the handset was indeed who they claimed to be.

We looked into what transactional areas people would want to develop and decided there were three. The first was looking at your bank account, but banks don't deal with small suppliers. Then there's share trading, but people don't trade shares unless they have a lot of information in front of them. And the third one was betting. We decided that we were going to create a framework that could do any of these things, but betting was the first market where people would want to keep maintaining the momentum they had built up on the online position.

Is betting a market you were familiar with?

Not at all. The closest I had come to it was when my grandfather, who was a baker, used to have a flutter on the horses on a Saturday. I didn't know anything about the betting industry, but I've become an expert. Betting firms want to put more of their business over platforms like web and mobile because they want to replace their shops, which, to put it simply, cost too much money. We put down a stake early on - and now it's coming back as companies look to refine their offerings. Just to give you some indication of the volumes involved, when we were running it, Betfair Mobile had £50m of bets put through in 12 months with 2,500 users. Now it's run by Betfair itself and it must have 50-60,000 users and 10,000 very active users. So there must be £250m going through that application.

You personally developed the core technologies for Rapid Mobile, Think Phone and your Active Provisioning platform. Tell us the thinking behind them.

We began with Think Phone, a design method that mobilises rather than miniaturises - which is one way of thinking about how you put an application onto a phone. There are a bunch of software components around it that let you build things really quickly. The principle is that if we can build prototypes quickly and show them to people, then they can say whether it's what they want.

The next thing I addressed through Active Provisioning was device diversity, where every phone is different. Even firmware versions in the same device will have subtle differences between them. The way that most people address this is through manual porting, so they have an army of developers porting to the new environment and testing it. That creates two problems - firstly, it's not economically viable, and secondly, it means the most recent handsets are the unsupported ones, which is not what you want.

I'd been writing multi-platform software since University, so I created an architecture that learns about handsets, recognises them as it comes across them and assembles the application, using a picklist approach for each particular device. There's a lot of implicit knowledge in there, there's a lot of heuristics - in other words, if it's one of these devices, it has this problem and this is what you need to do about it. But it's very much generalised. We've had that running since the beginning of 2005 and up to 3,000 device types are served.

But these technologies haven't been widely adopted by the mobile industry. Why was that - and how did you respond?

You'd think people would bite your arm off for something like that, but there are three types of people we've encountered - those who haven't come across the problem of device diversity and don't think it's an issue; those people who don't believe it can be automated; and those who get it. We've got a couple of deployments coming up that use our technology to deliver other people's applications. But software engineering is full of 'not invented here' syndrome, so we don't do a lot of that.

The thing that really appealed to us about mobile was its huge potential. There weren't any major players in there, and you knew the major players were going to come in and buy other companies. Of course at the time you don't know about all the downsides, and we learnt eventually that nobody buys tools. People will buy services and applications but they rarely buy tools, for several reasons. One is that people can get what they want for free off the internet and they're often better than the commercial ones. So now we sell vertical applications, and we use our platform for doing that.

You've now moved over to focus more on managing the company, letting your engineering team do the code crunching. How have you managed the transition?

I haven't touched code for some time, although I'm still heavily involved in the design of things - the user interface, whether it works and the functionality. I like to think of myself as a chef in an expensive restaurant - I write the recipes and sit at the chef's table, and then the other sous chefs bring up the dishes before they go to the customer. I get to pick at them and make sure they look right. My perpetual goal in life is to make myself redundant in what I'm doing so I can move on and do something else.

Well, if you create an exciting environment with techies running it, you can readily hire quality technical staff and attract people by your excitement. But getting on board quality sales staff is probably the hardest thing that we've been involved in. Any decent salesperson will by definition be able to sell themselves. This is an area where board advisor Andy Brannan has been helpful, as well as our investors. They have somebody they work with who's a sales consultant, who's helping us. I do a lot of the sales activity myself but I'm not a trained salesperson, I have a PhD in computer science. In the analogy of the people who make the tracks, the people who build the road, and then those who build a motorway on it, I'm the one who's cutting down the trees and making some muddy footprints.

Richard Marshall was talking to David Longworth of Webster Buchanan Research