To be intelligent, an organisation needs data and analytics. That demands a combination of culture, processes and technology. Few companies have got that far, discovers David Reed, but there is more interest than ever in following the ones that have.
Do you believe that using data and insight to drive strategy and decision-making is the right thing to do? Almost certainly (otherwise you wouldn’t be reading DataIQ). Has your organisation taken significant steps to make that happen? Probably not. Even as data enjoys its highest profile year ever, the gap between acceptance of the goal and its delivery remains wide.
“The argument has been won, but not actioned,” says Colin Rickard, newly appointed as enterprise and channel sales director at Experian QAS, having left DataFlux earlier in the year. “The whole concept of creating a single view of the customer (SCV) is intrinsically right. So why has hardly anybody done it?”
Research carried out by Experian QAS into the presence of SCVs in business could not make the issue clearer - while 72% of UK companies see the competitive advantage in using data and insight to put the customer at the heart of their business, just 29% of the 400 companies surveyed had plans in place to take this step.
A straightforward explanation for this ongoing lack of investment is that it is just too hard. Mature organisations are likely to have complex IT infrastructures which hold data in multiple locations and formats, from legacy account systems to e-commerce platforms. Bringing those together is no easy task. Even pure-play digital companies can find that their data stores have multiplied and that insufficient rigour has been applied to data capture, making the process of matching and merger that much harder.
But that is still only part of the reason why so few truly intelligent businesses can be identified. The real obstacles are cultural and political. Business schools are only just starting to integrate data and analytics into their courses. In the meantime, the cadre of managers currently at the top table were educated into a business model which put sales and product first, with data an optional add-on (usually in the form of targeting lists for direct marketing).
Changing this is a hard task for the data management or customer insight function. “What you need is part magic bullet and part incentive,” says Andy Day, general manager, business intelligence, Telefónica UK. His organisation is on a journey towards being intelligent that started relatively recently, thanks in part to what Day describes as “a perfect storm moment”.
From events in the external world, his immediate boss (O2’s marketing director) and the chief executive started to gain an appreciation of how information could be used in a different way. Day helped to push this along by inviting Clive Humby in to talk to the marketing function about what he had been able to do for Tesco. The market advantage gained by that retailer in part through its use of Clubcard data makes for a compelling argument.
Internally, Day became a champion for information-driven decisions. “I went out on a bit of a limb and went to fight for it. It meant constantly picking away at people and saying there is a better way to do things like store locations, retailing, customer profitability. Eventually, some of those people started to say, ‘go on then’,” he recalls.
Winning buy-in might look like a significant hurdle has been overcome. But in reality it is just the start - the next step is to prove the point. In the early phases of the company’s journey, Day brought in an external consultant to look at key business processes. “I said, tell us what would happen if we were best in class in applying analytics to those,” says Day.
Based on 25 different processes that were considered, a total of between £200 million and £400 million in incremental revenue was identified as being on the table. Day says that the specifics were ultimately less important than the principle: “Any number that is positive matters to a business where revenues and margins are going backwards. There was a very real desire to re-engineer the cost base, to optimise and be more effective.”
Early projects delivered hard evidence that applying intelligence was the right way to go. In one case, it was identifying a small segment of customers who were actually costing the business money - up to £14,000 per year each in some cases.
Out of the list of 25 opportunities identified, some process owners were more willing than others to embrace data and analytics. As they did so, further proofs emerged of the benefits. The Pareto Principle does apply, with 80% of gains coming from 20% of the activity that has been undertaken.
Even so, it has been enough for Telefónica to make a sustained, multi-million pound investment into its Business Intelligence team, which Day now runs. This may have catapulted the business into being best in class, but he still says, “it is early days. It is satisfying, but we are on a journey that won’t be complete until we are turning in £10 million or £100 million in business benefit.”
Starting small and growing large is a typical journey for companies that have become intelligent businesses. Given the technical and organisational barriers that exist, it is unlikely that a “big bang” transformation would succeed.
Hotels.com is a good example. Despite being a pure-play digital business, it may have fewer obstacles to overcome, but it still needs to see the business benefits. Liz Shepherd, director of business intelligence and web analytics, joined the company to see through its deployment of the web analytics solution Omniture (now part of Adobe).
“When I first started, the infrastructure was not there. I did a lot of work across the group around data capture - what’s being done and why it matters,” recalls Shepherd. Through progressive development, she now heads a team and has ownership of the end-to-end process, from data capture through reporting to analytics. “Usually, you see separate people running data capture, business intelligence and the data warehouse. This way I know everything is robust,” she says.
Buy-in to the value of data and analytics has steadily grown, helped by the background of some senior managers. “My boss is old-school operational research. I recently presented at the Institute of Operational Research and Engineering on the concept of testing and multivariate testing which takes that one step further,” she says.
Even so, she notes that, “the hardest thing is culture, so we have invested a lot of time educating the group. We have a ‘submit a test idea’ approach. People send their ideas in and we pick them up or not. That makes for inclusiveness. For me, that is one of the big things that I have achieved over the last few years, removing some of the myths and opaqueness around data.”
Testing also delivers proofs of the value of ideas. In doing so, it removes the “HIPPO” problem - highest paid person’s opinion. Shepherd has ensured that the business intelligence team is independent of any function, which further guarantees the objectivity of its findings.
Its influence is now extending to embed the insight culture yet more deeply. “It is already in our DNA - most people come from some sort of analytics background. But when new people come on board, we have an analytics person give them an interview to see if they have got the basics,” says Shepherd.
Business opportunities are one of the drivers of intelligence. Another is regulatory pressure - in regulated industries like financial services, there is a constant flood of demands for organisations to identify, manage and report risks and exposures. Each of these assumes the presence of relevant data which can be assembled to ensure compliance. In reality, each new regulation often demands further investment.
TABB Group recently published a study, “Data Agility: Turning a Liability into an Asset”, which considered how these compliance investments could be leveraged to return more revenue. While organisation and culture may present problems, the report found that technology costs were also an concern.
“Data management is a business issue that has traditionally masqueraded as a technology problem, and it needs a top-down view from the business. Those that can harness distribution and complexity and make intelligence work in their favour, rather than become a problem to solve, will fix their roots rather than repair their errors,” said the report.
Miranda Mizen, author of the report, says: “The problem everybody has is that issues with data can be overwhelming. It is so vast, moving at high speed and with so many complications. Access to the data you need is a critical problem.”
In her research, it became obvious that the data necessary for regulatory compliance exists in organisations, but accessing it and then distributing it where it is needed is hard. “Just because you have it, you can’t always see it,” says Mizen.
When this gets addressed as part of a compliance project, it is rarely with a view to delivering benefit more broadly to the organisation. That is why data management projects are more often siloed than enterprise-wide, which limits the degree of intelligence which the business gains.
“It is like travel planning compared to town planning. For your own journey, you look at how to get from A to B quickly. But if everybody does that, you end up with Spaghetti Junction,” says Mizen. “Data in companies is the same. Getting it from A to B works in a small environment, but it does not benefit everybody.”
Standalone projects also tend to multiply the total cost of ownership because each adds an extra layer of maintenance and support. Based on its study, TABB Group argues that a new model needs to be adopted which extracts data from its source, transforms it for each new purpose and delivers it to the end user where they can manipulate it as necessary, without changing the original data itself.
That concept of the dynamic data warehouse has been around for a while and looks appealling, but has tended to suffer from a lack of viable solutions. Pneuron Corporation has recognised the problem and written a new solution from scratch (earning a name check from TABB Group in the process).
“We did research into the issues and asked why it is that, no matter how big your IT budget, 80 cents in the dollar goes on maintenance. The total cost of ownership for compliance projects is the same now as it was in the 1990s. And given that your applications need to be replaced every five to seven years, you are reliant on the roadmap of your software developers,” says Simon Moss, CEO at Pneuron.
Having sold a previous software business (Mantas) to Oracle for $127 million cash, he has a track record in identifying and solving this type of problem. The hypothesis of the new solution is to extract data from existing databases, transform it and load it into high performance analytical packages for use however the organisation needs. “Users can build reports and models without having to move any data around,” he says. This new middleware runs on just 90,000 lines of code and creates real-time business intelligence using existing infrastructure.
Gaining commercial insights from compliance data management investments may sound too good to be true. But it is typical of the new thinking which is emerging to help create the intelligent business. Senior management may have the vision - now they no longer need to have the fear. They can just do it anyway.
The Single Minded Solution:
Competitive advantage can be won by building a single view of the customer. So say nearly three quarters of businesses surveyed by Experian QAS. But of the 400 organisations it asked, only 29% have plans to introduce a Single Customr View (SCV). That leaves open the question of how many have actually built one, given that 57% admit they have taken no further action beyond developing a SCV strategy.
It is not that the benefits are not recognised - having a SCV could have helped more than eight out of ten avoid specific business problems over the last year, while nearly six out of ten could have driven out inefficiences such as costs.
So what are the barriers? People is the most commonly-cited problem and data itself can get in the way. Half of companies said they could not get buy-in from stakeholders and the board. With so much opportunity and yet so little action, it is no wonder that industry leaders who have taken the plunge are reaping substantial benefits.