Integrating customer data into a single view has long been an objective for companies. But as Michael Green, director of Transactis explains to David Reed, with the pace of change in data, you need to avoid creating a solution that ends up becoming a legacy.
Are you running your business to a model from 2006? Of course not - e-commerce, mobile and social networks mean things have changed dramatically in the way you interact and transact with customers. Despite this, there is a risk that you could be running a data model that was created five years ago (or more).
Single customer view (SCV) is an important objective in order to provide lines of business with the key information they need to make decisions. Creating it requires integration of multiple data sets from operating systems. That process may well have been architected as part of an IT migration in the last decade. Since then, many new streams of data have come online, often with very different structures and technical demands.
Adapting to a flexible customer view requires data modelling and integration skills that reflect this newly dynamic environment. Winning investment from the business to back an updated SCV requires proof that functions will benefit significantly from a more up-to-date view, rather than just cashing in on their legacy.
Michael Green, director at Transactis, explains that SCV should not be seen as a once-for-all fix. “CRM systems and single customer views built in the past were never intended to run themselves. Companies that don’t consistently monitor and update them run a risk in that they will not fully optimise intelligence in support of the business and the opportunity to improve the service that the customer receives.”
The dynamics which drive the need to rebuild a SCV are fairly obvious. Source data feeds change and degrade with some variables no longer being populated, like fax number, and others altered by demographic movements, like marriage or divorce. New variables may also become business-critical, such as mobile phone number or even Twitter hash tags. An existing customer database is likely to have been built around assumptions on the availability or dynamism of these feeds.
As a consequence, the operating costs of a SCV may have risen. “Analysts may have to import specific data feeds into their system to work with resorting to workarounds,” says Green. Time spent preparing data reduces the time available for value-adding activity, such as model building and insight generation.
“The result is that a company may be paying for system, software maintenance and licenses that are either out-of-date or underutilised. A road map and commercial plan must be developed that is both dynamic and constantly reviewed and updated,” he says.
In the last wave of SCV building during the Noughties, the focus was on growth and maximising revenue from customers. Investment was relatively easy to win because incremental income flowed steadily every time data management improved. In the current climate, there is a different emphasis - short-term gains have to be in the foreground as companies strive to maintain income streams.
That does not mean the mid-term should be overlooked - data will be key to commercial objectives and business change based on customer development. In the long-term, the new generation SCV should be able to support the full spectrum of customer management and also drive business intelligence further into the organisation.
Betfair experienced the benefits of a new SCV through greater efficiency in its marketing process. The online betting exchange has over one million customers and handles more than five million bets every day. Its previous data warehouse operation required three to five weeks to set up campaign selections and exclusions, with the data then sent to an external bureau for cleansing. Overheads for each campaign were as high as £3,000.
A new SCV was created which maintained data at a recency of 24 hours and gave marketing teams daily access to the previous day’s activity. Cleansing was automated and a managed environment allowed for significant gains in activity rate - Betfair went from running five campaigns per month to 500 with turnaround times of 48 hours. Integration of multiple accounts into a single customer view also increased customer satisfaction and reduced call volumes to customer service teams. The project achieved payback in under a year.
Green notes that changes in data volumes and types are another reason for rebuilding the SCV, from increased numbers of products, higher volumes of customers and data components required to manage them, through to components that support new activities. These include the introduction of data governance, which will fuel a demand for suppression flags, and unstructured data, like text captured from email or social networks.
“Technology and the availability to track an individual’s behaviour, engagement and touch-points will naturally continue to grow at a faster pace than ever before. Unstructured data can now be managed and provide significant value in modelled solutions,” he says. “The danger is information overload. Therefore flexibility in terms of storage and architecture development is vital to maintain a competitive advantage.”
Data feeds which have stretched the boundaries of the classic SCV are not the only consideration. Business functions are increasingly demanding access to both customer data and insight resources. More users mean a system needs to be in place that can cope with enterprise-wide data flows, rather than supporting a single function.
Green notes: “Traditionally, customer data was primarily mined and analysed by marketing in support of communication development. However, the knowledge that this drives in terms of understanding a customer’s behaviour now means that this data can be used in a plethora of ways across the corporate enterprise.”
Understanding how customers buy across a portfolio of products and services is essential for new product development. Patterns like seasonality or product combinations, through to packaging preferences and price elasticity are all business-critical. Location planning also has a major impact on the business and can be driven by mapping customer location, distribution network and catchment potential.
Service operations have been using customer data from CRM systems for decades, but now face an enhanced set of touchpoints they need to support and understand. Finance departments can budget better if they understand the customer funnel. And legal is increasingly required to report or audit customer data as part of its compliance work.
Green says that heightened data usage will highlight any deficiencies in existing solutions. “As any system ages, its performance needs to be monitored to prevent deterioration - poor operating performance can result from poor indexing. Delayed build times will likewise impact customer service, as will running a number of redundant or obsolete routines. So periodic review is vital,” he says.
With a broader set of business functions to support, the business case for a new generation SCV should be relatively easy to create. Cost savings from avoiding workarounds and production wastage sit on one side of the balance sheet. On the other side are uplifts in targeting, response rate and income. Faster and more accurate decision making also makes the whole business more efficient.
So should companies be looking at a new, integrated data warehouse with every instance of customer data pulled together in a single system? Green argues that this is not always necessary. “This may be desirable for optimised performance dependent upon the application. Our SCV builds processes that are modular and can then be adjusted dynamically without impacting existing code,” he says.
For IT departments struggling to keep pace with growing infrastructure and wary of any new introductions that might knock over an existing solution, this can only be good news. Integrating data dynamically to support specific views can now be achieved without the extensive timelines that have been typical of previous SCV iterations.
Working with an external outsourcing partner like Transactis also avoids the perils of the SCV becoming an IT project, rather than being driven by the line of business. “To ensure that the data is effective in all parts of an organisation, it must be developed as an enterprise-wide solution and approach,” he argues.
Green adds: “Ideally, with board or senior-level sponsorship, a steering committee or similar should be developed providing representation from all relevant parts of the business including IT, operations, marketing, legal and finance. This ensures that each area understands their role at the outset and the value that each brings.”
An external third party can also help to avoid some of the political issues that customer data integration often gives rise to. At its simplest, marketing and IT rarely speak the same language and can struggle to understand the other party’s challenges. Says Green: “This may be the reason why a system needs updating in the first place. If this is the case, then external support can help and ensure that benefits are evidenced in the short term in order to maintain momentum.”
Choosing the right external partner is essential if the organisation does have its problems, either in terms of existing data feeds and technology or in getting stakeholders to work together. A third party should be chosen which has experience of the right type, scale and character of SCV required. Equally, it needs to be offering the speed, flexibility and a dynamic approach to ensure the new solution delivers the scalability and value required.
While many companies are looking to build a new customer data mart in response to their data flood, their are others with the reverse problem. Data famine can be a real issue for businesses that rely on prospecting to keep their market fresh and achieve sales targets. For these end-users, Transactis has another proposition - the pre-integrated customer and prospect universe, Vision. Providing a 360-degree view of all activity and a suite of selection options, combined with sophisticated analytical tools and support, it helps to fill a gap where the infrastructure or data feeds are simply not available.
Green points out that working with an independent provider to create a prospect universe avoids conflicts of interest that can otherwise arise with prospect pools. “We can act in an unbiased way for clients evaluating data that is available from numerous sources, both here and across Europe,” he says.
Damart adopted this approach to support its two million active customers and drive its business goal of 10 per cent annual growth. Vision integrated account and transaction history with campaign codes for all customers, as well as streamlining its third-party data buying by identifying overlaps and differences in performance.
Within 12 months, the system had paid for itself through cost-savings on campaign processing and data purchases, while cutting turnaround times for campaigns by 50 per cent. Cost per acquisition fell by an average of 30 per cent and response rates from customer reactivation campaigns rose by between 30 and 50 per cent.
Damart’s marketing director John Bottomley says: “Vision has given us more control and better results from our direct marketing, over-delivering in almost every respect. Campaign timings have been slashed and costs have significantly decreased because we now buy data more effectively. Vision has already paid for itself and the improved response it is able to generate means our campaign ROI is consistently on the up.”
Single customer view remains an important concept. But the way it is delivered has changed with a move towards more flexible, dynamic solutions that can work alongside operating systems and bring external prospect data into the mix. As a result, both marketing and business performance can improve. Now that is a legacy worth having.