Applying for a mortgage can be a frustrating experience. In many cases, customers spend hours filling in an application form, only to wait around for days, even weeks, to find out if their application has been accepted. UK bank Abbey National has whittled this process down to around 10 minutes. Instead of forcing customers to fill out a 17-page application form, the banks asks them a set of questions over the phone, or allows independent financial advisers to enter clients’ responses into its mortgage application, ‘Introducer Internet’, over the web.
Once this data has been submitted, Introducer Internet combines these responses with historical data from Abbey National’s customer database and data from an external credit reference agency. The application then performs an in-depth analysis on this combined data and provides the customer with a ‘decision in principle’ in 60 seconds.
Since it went live in October 2000, Introducer Internet has processed £16 billion worth of mortgage applications. Abbey National’s nearest competitor, meanwhile, can only boast a 48-hour turnaround for the same process. While this gives Abbey National a distinct competitive advantage over rival mortgage providers, the data Abbey National requires to make a mortgage decision already exists. It is simply the ability to consolidate data from disparate systems and analyse it that enables Abbey National to speed up and improve customer interaction and boost satisfaction levels.
It is no secret that many customer relationship management (CRM) projects have failed to generate projected returns. According to analyst company IDC, UK companies will invest £17 billion in customer relationship management (CRM) projects in 2003. But according to some CRM consultants, as much of 70% of this investment will be wasted.
One reason is that most CRM projects to date have focused on operational CRM systems; that is, software that automates many of the tasks involved in processing interactions with customers, whether that be registering a complaint in a call centre, answering a customer query, or closing a sale. This has clear advantages in terms of reducing costs and streamlining processes.
However, the problem is that the data generated in these transactions is often left to languish in application-specific databases dotted around the organisation, and not used to fine-tune marketing campaigns or improve customer service.
The key to gaining value from CRM investments, say consultants, is quite simple: better analysis of the data that CRM systems create. With this in mind, IT decision-makers are beginning to dig deeper into sources of customer data. Analyst group Gartner has identified a growing trend towards ‘customer relationship optimisation’, where organisations apply business intelligence and analytics technologies to the customer data across these operational systems and act more proactively on what they discover.
Some organisations are going one step further, says Gartner analyst John Radcliffe. “The classic scenario is to take all the data out of your operational system, put it into a data warehouse, clean it, analyse it and decide what to do. Now organisations are closing the loop, getting the analysis back into the operational systems,” he explains.
But in many organisations, that ‘blood supply’ is neither single nor consistent; they tend to collect customer information in a fragmented way and analyse it only intermittently.
This was the case at Scandinavian financial services company Union Bank of Norway (UBN), for example. Because, like many organisations, it uses multiple computer platforms and systems, and has disparate data scattered across a network of branches, the bank found it difficult to gather and store customer information in a uniform manner. As a result, analysing the success of its marketing initiatives proved complex. UBN has since brought all these data sources together into a single data warehouse and its customer contact and marketing systems sit on top.
“Before we installed the data warehouse, it could take weeks to get the information and formulate it into one view of the customer,” says Kari Opdal, head of CRM at UBN. “Now, we analyse historical data and combine that with transactional ‘events’, which we monitor daily.” One campaign triggered by analysis of the information in the data warehouse, based around customers’ lifestyles, generated a response rate of 70%.
One of the challenges of creating a loop between processing customer data, analysing it and acting upon it, however, is simply moving that data around. A recent survey of Siebel users by US analyst company Nucleus Research found that one of the most commonly cited difficulties cited by customers was extracting and inputting data to and from the system. Radcliffe at Gartner advises organisations to use application middleware to link up disparate customer databases. This creates the perception of a ‘single view of the customer’ regardless of who accesses it and the type of queries they place.
US technology giant Hewlett-Packard has followed this route. The company extracts data from its multi-thousand-user Siebel implementation, along with data from reseller and corporate customer sources, and places it in a ‘virtual data warehouse’ for analysis. It then feeds the results of this analysis back into the Siebel system as the basis for marketing campaigns or sales activities in its call centres.
“The key benefits have been around marketing planning and execution,” says Gary Allen, manager of customer knowledge at HP. “The other thing is the ability to be more coherent with customers. If you have a better idea of who they are, and what products they own, you can make offers that are more relevant.”
Some technologies take this process to still more sophisticated levels. In November 2002, analytics software vendor SAS Institute acquired a customer behaviour tracking engine from a company called Verbind. This will enable companies to combine historical analysis of customer data with real-time monitoring of customer ‘events’- a significant deposit into a bank account, for example – to further personalise customer interactions. Another company, US-based Unica, includes a feature in its marketing automation suite that ensures an organisation has sufficient resources available to respond to these events, such as checking there is enough inventory available to satisfy an event-driven customer offer.
Kevin Scott, an analyst at IT market research company AMR Research, believes these newer elements of customer analysis and marketing automation will become increasingly important in helping organisations gain more value from their customer databases. “This coupling of analytics and behaviour tracking will allow companies to truly personalise all customer interactions, not just marketing campaigns,” he explains.
But bringing together customer data to formulate marketing campaigns and event-driven responses can have a downside. At Union Bank of Norway, for example, Kari Opdal found that there were a lot of internal arguments because sales and marketing people no longer had as many sales leads to work with – even though the leads they were given had greater potential. “The scoring and predictive modelling we performed on our data showed that only around 20% of the people you target with a campaign are actually likely to respond. You have to change your marketing strategy to be more event-driven,” she says.
Closing the customer interaction and analysis loop also requires organisations to adopt a more holistic view of their sales and marketing strategies and how they relate to their core business activities as a whole. As Godfrey Sullivan, president of analytics software vendor Hyperion, points out, identifying customer needs is important, “but if you can’t actually fulfil that demand, it defeats the purpose of having an analysis system at all.” The companies that can achieve this balance will be the ones that meet with most success.