Gary Flood talks to David Jack about the challenges facing his organisation, why a move to the cloud would be a good thing and why it’s important for a CIO to be surrounded by people who can be a pain in the backside…
David Jack has been CIO at UK rail travel eCommerce engine Thetrainline.com since 2009 after a career in both the vendor (Ashton-Tate, Citrix, Novell) and commercial/end-user ends of ICT leadership (he was previously CIO at BetFair, where his last role was setting up both that firm’s internal VC arm and then a separate financial trading platform).
The company he works for is owned by a private equity firm called Exponent, being spun out of Virgin some ten years ago. It provides access to timetables, fares, reservations, and tickets through its Internet site and contact centre operations to both consumers and businesses, and is also the underlying platform for many TOC (train operating companies) individual booking sites. There are rumours Exponent may sell the firm next year for a potential asking price of at least £400m.
Many of us will have used your company’s web site; but is there anything new and different about it we should know, to get us started?
A way to think about us is that we are to rail what a Sabre or an Amadeus are to air. And the latest thing we’ve done is launch our iPhone app, I suppose, which is actually tremendously exciting. After we added a retail option to it it went from zero to 1% of our total business within two weeks.
I also understand there’s been some re-architecting work going on?
You could say that! When the system started it was run on an outsourced basis with CapGemini as tech partner, but we have both taken that back in-house but also totally rebuilt it, soup to nuts. As part of the re-write, it’s been made into a truly service- and virtualised operation, built on a true SOA (service-oriented architecture) basis, with everything working off one full API set.
Why is all this a ‘good thing’?
Because we run a highly spiky business. It’s not a few peaks and troughs of internal usage – our site had 40% more traffic less than a minute after the announcement of the Icelandic volcano problem with air travel, for instance. So it’s a very high-volume, low margin business, all about lots of tiny commissions, so cost is a very key part of our profitability and I am delighted to be able to say that the new architecture trend-broke our average cost per transaction.
Sounds impressive. Who are your main technology partners on all this?
We have Oracle at the back end but Microsoft at the middle and front end.
So you are a Microsoft Azure customer?
Potentially. At the moment we are internally virtualised and are looking at a possible private cloud move and maybe if it makes sense an eventual move to a public one, too.
But the great thing about the system is that it is constant evolution. We have a very complex core data structure because of our heritage but it may make sense for us to switch from Oracle to MySQL or maybe even flat file if that offered us performance advantages, for example. After all, we are a high transaction site but we don’t care so much about latency as a financial site – it’s OK if we can close in 500 milliseconds as opposed to the sub-200 I had to do at BetFair.
The point is we are always looking to see how we can drive the system to deliver more. It’s simple, it’s not that sexy – but it brings huge value to a lot of people. With cloud, for example, I just challenged my team and asked them what it would take to be able to build and test the entire system in a minute. First they laughed, then they said they could do it in 10 on 500 environments. Well, with cloud, one day you could set up 1,000 environments that lasted ten minutes. That’s very interesting to me – the cloud is not about hosting enterprise apps for me but more about totally changing the way I can deliver what I need to deliver.
Was that what attracted you away from your previous job – the chance to keep a system this big evolving?
Partly, though I loved BetFair. It was a very stimulating job and I still say one of the coolest technologies being worked on in the UK – at one time, we were running the fifth hottest Oracle database on the planet. But I am at heart a technologist and I am not credible as the CEO of a multi-billion dollar finance firm, which is the direction I was headed there.
Though I note you are on the main board at Thetrainline.com.
Yes – that’s incredibly important to me, actually. I came to this post as there were big challenges, a job to be done and it is by no means a BAU (business as usual) sort of job.
But a lot of CIOs do do ‘BAU’; there’s this whole idea that the role is about "keeping the lights on," right?
I hate that. My dog can do that – keep the lights on. That totally misses the point.
Because if you aren’t adding value, what are you there for? I also hate this whole IT thing about "the business" versus "us". When one of my guys moans "the business" can’t make up its mind about something, I say back, "Do you mean Steve over there? Well, go over and talk to him and help him make up his mind!" Using the language of division helps maintain that false division and I won’t have it in my shop.
Which makes me wonder what sort of a CIO, an ICT leader, you are yourself, David?
I’ve worked with some of the best software engineers in the world, people who turned down jobs at Microsoft because they thought they were too good for the place. It’s fantastically important for me personally to work with people like that. I need that sort of individual around – even though they can be total pains in the arse.
[Laughs]: Indeed! So how do you get the most out of such people?
Because what I can do, I think, and what all good CIOs should be able to do, is to synthesise genius. I’m not a genius. But what I can do is take the edge off a genius team and make them deliver what the business needs, I think.
Great stuff. So let’s conclude with some tips from the David Jack book on how to be a good CIO?
Oh I don’t know. If I have anything to say it’s that, if you want to use that lights talk, you need to say to the business the lights will go out occasionally. Business is about risk and why should IT be the one part of the CEO’s business that doesn’t talk to him on that basis?
I’ve always said to my bosses, I’ll do three things and two will fail. They won’t kill the company, but they will cost money. But the one thing that works will be right and will add value. I think that’s what the job of the CIO should be. Think like a VC and the CEO will be up for it, I think.