Flexible thinking

FST had the pleasure of talking with Chris Hayward, Global CIO of Corporate Services at Deutsche Bank, about building horizontal platforms, outsourcing, and how IT needs to act as an internal consultancy.

Work in a large complex organization with multiple silos? Check. Need to constantly deliver innovative solutions? Check. Dealing with internal customers facing their own departmental budgetary constraints? Congratulations – you’re a senior IT decision-maker in the financial services.

Throughout the industry the set of challenges being faced is broadly the same – how do you offer more value through spending less – and this is no different for Chris Hayward, the Global CIO of Corporate Services at Deutsche Bank. Based in London, Hayward heads up a team of around 1700 staff, spread around the globe, and is responsible for corporate functions such as finance, risk and compliance, as well as rolling out utility platforms across the global business.

As a chartered accountant, and with a previous career in UK retail under his belt, it is perhaps no surprise that Hayward brings a refreshing perspective and insight to our conversation; he’s clearly a man who is focused on reaching the right destination, and seems relaxed about the exact path that will get him there.

We start by discussing innovation. Part of Deutsche Bank’s corporate mission is to deliver value through the delivery of innovative solutions and unique insights – so how crucial is Hayward’s role in technology to support this goal?

Hayward describes his most pressing challenges as “many and varied”. He cites the near exponential growth in volumes and the increased granularity of regulation as two standout issues, and related to both is the ability to cope with the volume of complex data that comes with it.

The other challenge taking up his time is his attempts to drive horizontal integration throughout Deutsche Bank’s operations. “Traditionally in most banks, business lines are very business unit/silo based, but when you’re as far away from the end client as me, it doesn’t make sense to have that difference,” he explains. “So we’ve been trying to build technology platforms that can be used throughout Deutsche Bank.”

Viral marketing
With an organisation as large and as globally dispersed as Deutsche Bank, this is no small task, and it’s easy to see how it might come off the rails. So what approach has Hayward taken to this challenge? It seems he has gone down what could be described as a viral marketing approach.

“In the past, when we’ve tried to build these kind of platforms, we’ve tended to build something and then tell everyone they must use it – and that doesn’t work very well with bankers,” Hayward comments. “Tell someone to do something, and they seem to think of a good reason why they shouldn’t.”

With this in mind, Hayward approached the challenge from a different perspective. “The idea was to build something and show people how great it was – so they’d ask to use it – rather than build something and tell people that they had to use it.”

To this end Hayward started with two initiatives: an interface platform and a reporting platform. “We started very small, with a limited level of investment, three years ago,” Hayward says. “We’re now at a point that we’re spending significantly more on these things. We have lots and lots of people throughout Deutsche Bank across all the different business units coming to us and saying ‘can we please use your service?’”

Why does he think the pick-up has been so strong? For Hayward it is both a case of showing people the quality of what can be delivered, but also developing it at a price point that better suits internal budgets.

“This approach allows you to prove that it’s cheaper to use a standard,” Hayward claims. The problem in the past, he explains, is that when you front load the investment to build the complete solution the “first few” users have to bear much of the development cost. So by starting small, you can take that ‘innovation penalty’ away from the early adopters.
“We tried to make it so that the early adopters did not pick up excessive costs, so we had to build small and grow; but that seems to have made for much better adoption all round.”

The bank is obviously pleased with the approach, as the methodology is one that will be repeated as it pushes out a set of BPM tools – and the service orientated architecture (SOA) that goes with it – to the business.

“As we move jobs around the globe, having a process modeling tool and a process modelling language is critical to us because it allows us to be fast, efficient and reuse things. So that’s probably the next platform we build during 2008.” As he explains, this will be partly an IT infrastructure but also aimed at the business. “Increasingly there are the businesspeople that know how to use it as well – as you move away even further from the traditional IT world, you allow users to start almost changing the systems.”

But how does this ambitious goal sit with the idea of starting small? As Hayward suggests, a true BPM methodology is about so much more than technology, so how can you embed the process changes if it’s optional – isn’t there a risk that it will just get lost?

“I think that’s a very good question,” he muses. “I think the only way you can get away with building on a small scale and rolling out is if you have a clear vision for the future so that you know where you’re going and you’ve actually sold this to key stakeholders – you’re not going to ask for the full amount to get you there though!”

For Hayward this is the key: get the vision right so that people buy into it, and then start showing them examples. “If you’ve got the vision clear in your mind, you can go away and make it grow. It’s viral almost in that someone has a nice report on their BlackBerry, or they have a nice report on the screen. One of their colleagues sees the value and says “I’d quite like one of those as well.”

The result is that instead of asking for a multi million budget, you’re pitching for seed money only. This incremental approach has the added advantage that you can learn from the feedback and experience of the solution in action. “You might think you know the structure you’re building but it would be extremely arrogant to think you know what the future will be – so you might need to make some changes as you go,” Hayward muses. “As long as you’re going in the general direction though, you are building in the right way.”

Outsourcing
As we discuss this incremental approach to new platforms we come round to the subject of outsourcing. Given the nature of small-scale projects growing outwards, does this imply that Deutsche Bank is keeping lots of the development in-house, rather than outsourcing to the vendor market and buying in white label solutions?

Hayward’s approach is pragmatic and again based on iterations of previous experience. He explains how the approach is usually to pick an industry standard and define the strategy in-house, and then work with vendors to provide the actual infrastructure of the solution. “We do the design work, and then we’re outsourced after that.”

Indeed in terms of the staff in Hayward’s department, he estimates of 1700, about 550 are Deutsche Bank employees, while the balance work for third parties. This mixed strategy is obviously driven in part by cost considerations, but as Hayward argues, there needs to be some caution about divesting too much control of your projects.

“You have to be very careful that you understand what it is you’re outsourcing and that you don’t outsource knowledge, which is always a big risk,” he states. “We have an increasingly mixed strategy where we have Deutsche Bank sites in high cost locations, Deutsche Bank sites in low cost locations, and we’ll have vendor staff – we try and use a mix of those to provide the right level of service in the right place.”

He suggests that the success of any outsourcing project comes down to the quality of people that you retain onshore, as these are the staff who will architect the solution. “If we just pick up 200, 300 people and move them offshore it never quite seems to work perfectly,” he admits. “The projects that work better are where we analyse the roles and put the right roles in the right place – so of say 50 people, 10 should probably stay onshore in this high cost location, 10 should perhaps be offshore in Singapore or Manila, and the other 30 need to go to a vendor. Of course, the mix is different for every project; it depends upon the skills we have.”

This makes perfect sense in principle, but in practice it means project teams split across multiple countries and companies. Hayward gives the example of the bank’s General Ledger support to illustrate this in practice. “We used to do it all in-house with high-cost contractors working for us. We then had a try at outsourcing it to a single Indian vendor. What we have now is a completely different mix: we have a site in a lower cost economy in the Philippines, we also have people in each of our major hubs on the project, and we have vendor staff doing the simple program tasks that we can happily and faithfully move away from Deutsche Bank.”

The role of IT
With this kind of dispersed structure, how does Hayward think the role of IT is changing – is it really sensible, for example, to think of a monolithic technology department or rather say a General Ledger department with its own purpose and set up?

Hayward has firm views on the role IT should be playing in the future. “I think we’re really moving from that traditional IT shop into more of a service job,” he accepts. One of the main drivers of this change he says, is that these days a lot of the actual work can be done by the end user. “We’re almost acting as internal consultants, sometimes we’ll just provide the technology, and the end user actually makes the things happen, all the way through to cutting amazingly complex code.”

And it is when the technologists are acting as consultants and business partners that Hayward sees them being most effective and really adding value. “In the past, I think many IT organisations have actually been doers, rather than people that have helped create infrastructure and new ways of doing things.” As he argues, being told by the business “build this” is not the right way of doing things.

“I think that the effective part, increasingly in my world, is that level of inherent consultancy-type of understanding and how you interact – how do you hear what someone says they want and then deliver properly what they need?” When this is happening effectively, Hayward argues, the technologists will be providing creative solutions, rather than just be acting as “order takers”.

“The business needs to tell us where it wants to go, and let us come back with some options for how it might get there,” he remarks. “Don’t tell us you want a reporting package based on a particular suppliers’ offering that does this and that because you’ve met the CEO. Tell us what it is you’re trying to achieve and then let us come back with some options for you.” This isn’t just more satisfying from IT’s point of view – for Deutsche Bank there is more value in a creative solution that does what it’s meant to and will drive reuse. “If we have a particular reporting tool, we will continue to use that one, and if people just come along and say, “I need it to look like this,” we can use that tool to make it look the way they want whilst reusing 90 percent of the previous investment,” he says.

Increased volumes
It has been a fascinating conversation, and with half an eye on the clock, we ask about the challenges Deutsche Bank faces going forward in 2008. Especially with MiFID coming into force, what are the key challenges facing Deutsche Bank on the trading side of technology?

“I think the big thing for us has been volume and stability,” Hayward replies. On the volume side of the coin he explains how the traditional batch systems that were fine for 30,000 trades a day just won’t cut it when you’re making nearer three million. “We’re all dealing with volumes now that people couldn’t imagine when the systems were built.”

The implication of these increased volumes is that the bank needs to start thinking more like a commodity provider rather than a “specialist” investment bank. “When you look at the sort of volumes we deal with, we’re now getting up to utility company or telecom company types of volume – you don’t make very much money on each transaction; you just do a hell of a lot of them.” This shift of course brings a whole new set of challenges.

And foremost of these challenges is the other side of the coin – stability. “At the end of the day, the most important thing our front office colleagues will ever say is “it didn’t break’,” he insists. “When the markets are really volatile, when the screens are all red, what they want is the certainty that they can continue to trade. Everything else is really interesting, but it’s that absolute, inherent dial tone-like stability that’s the biggest challenge for us.”

And how is this challenge met – is it a case of investing in lots of backup capacity for example? “The natural thing is you throw money or hardware at it,” he confides. “But it doesn’t help at the end of the day – the thing you have to do is you have to re-architect the process properly.” In order to do this Deutsche Bank is moving from a batch-based to real-time processing platform. “It’s something we’re certainly spending a lot of time and energy on,” he reveals.

Our last question is to quiz Hayward on what lessons he thinks the industry can draw from the chaos on the markets that has followed the summer’s credit crunch. Does the industry’s technology need to evolve to avoid some of the large write-offs that have bedeviled large players this year?

Hayward is quick to point out that Deutsche Bank has remained relatively unscathed this year, and he puts this down to the fact that it knew exactly what its positions were. “If the markets move against what you think is the norm, it can hurt anybody,” he accepts. “But we always knew what our position was.”

And he thinks this is the key lesson to be learnt from the crisis. “The challenge is really understanding the data that’s flowing through the organisation,” he says. “It’s understanding at what point trades occur, whatever the trades are, and making sure that you have good capture routines at that point in time.”

This is important because once the data is captured, you’re in a better state because you know about it. “You can then feed it into risk or finance, whatever systems you like, as long as it’s captured. So the big control point is not an IT control point; it’s a process con trol point. If you’re going to trade a new product or you’ve created a new product, you’ve got to tell someone about it before you start selling it. If you do that, we can figure out quite quickly how to link it up and make sure that it appears in our numbers.”

He thinks this is particularly important because of the increased volumes that are coming through trading floors. “The other point is that data quality is critical. I’m sure this will be a focus for all of us in the next few years; we’ve got to absolutely make sure that when you first capture the transaction, the data is absolutely right.” This will avoid lots of adjustments later down the line, and make flowing data through control systems relatively straightforward.

“Everyone has to understand their products, understand what they’re selling, and understand your infrastructure to know where that data is,” he concludes. “Then you can go and pick it up and do whatever you need to do with it. Then you can play.”