Analysing and using the wealth of information that customers provide when they contact a company is proving difficult for the insurance industry. Keith Rodgers explains why

For years, telephone sales and service representatives in consumer markets have been judged on metrics that largely focus on efficiency. How long are customers being held in a queue before their calls get answered? How many calls has each agent taken? How long did the calls last, and if it's a service problem, was the query resolved first time round or did the calls need to be escalated to other specialists?

While all of these metrics have value, however, in recent years some of the underlying assumptions have been challenged. Is it right to focus on call volume, for example, when keeping a customer in conversation for a little longer might give you the chance to sell them a new product?

And how long is too long when it comes to keeping someone waiting on hold? As one senior executive points out: "We might have a service level target of answering 80% of calls in 20 seconds, but in practice, probably nothing happens to the overall sales performance until you get to 40 seconds or 50 seconds. By falsely attaching too much significance to the wrong tolerances, you can waste resources."

These kind of issues are now coming to the forefront across the financial services sector as organisations explore better ways to measure and manage the performance of their customer-facing operations. Earlier this year, Webster Buchanan Research carried out in-depth interviews among a selection of senior UK insurance executives. The research probed fundamental issues, such as how effectively they sell, how closely they can monitor the response from their marketing campaigns, and how well they handle customer inquiries in their call centres.

The findings demonstrate that at some levels, the industry is relatively advanced in terms of customer performance management. Yet at others, it's still struggling to tackle some fundamental issues.

For most insurance organisations, particularly those in direct selling, measuring sales, marketing and customer retention are top priorities, given the high cost of acquiring new customers and the thin margins on first product purchases.

Lead generation
Several respondents to the research have put fairly comprehensive measurement systems in place to track the effectiveness of their sales and marketing activities, from assessing the results of lead generation campaigns to measuring how many deals are closed.

These systems don't need to be that sophisticated: one respondent simply sets up individual phone lines for each individual marketing campaign so it can log what calls are generated and allocate the direct costs of each campaign against the return.

Others are taking steps to track the effectiveness of their web-based activities. Direct providers can learn a lot about how effective their website is as a marketing vehicle, for example, by analysing where visitors drop off.

If large numbers of people leave your site when they get to a particular page, it's worth further analysis - it could be that the navigation is badly designed, or it may simply be a pricing page and people are leaving the site to compare you to your rivals.

Other performance measurement techniques are more problematic, however. Managing customer retention, for example, remains a particular challenge: retention and attrition are driven by multiple factors, ranging from price to perceptions of service quality, and it's hard to link specific initiatives to retention outcomes. Is a percentage increase in retention down to the large sums invested in a customer service organisation, for example, or is it because targeted customers were those about to lapse with a cut-price promotion?

In addition, retention itself has only limited value as a metric, after all, the most important thing isn't how many customers are retained, but how many high-value customers are retained. Figuring out who's who requires organisations to profile and segment their customer base according to spend and preferences. But so far, several respondents said they had struggled to carry out this kind of segmentation in any great detail, even in niche markets.

Similarly, sales performance management requires a relatively high degree of sophistication. One respondent at a multinational insurance organisation acknowledged that while it's good at measuring results, it's not so good at interpreting them. It's a long way from knowing, for example, what variables make some agents better than others at

certain types of sales call. Is it to do with the type of product? How much does experience count? To what extent does the number of calls that an agent handles matter, and how much does success depend on whether the agent is upselling an existing client or dealing with a brand new customer?

Analyse performance
As they grapple with these more sophisticated issues, many organisations are also trying to come to grips with some more fundamental challenges. Most insurers still struggle to get access to the customer information they need, which makes it hard to sell and market in the first place and harder still to analyse performance after the event.

It's an age-old problem, caused in part by a lack of communication between departments and divisions and their tendency to operate as 'silos' - a problem that's exacerbated for insurers selling through intermediaries, given that much customer-related data will be held by third parties.

These problems also tend to be reflected in a fragmented IT infrastructure. As organisations evolve, particularly through acquisition, they develop and acquire a range of different systems, from ageing mainframe applications to personal laptops. These systems often don't talk to each other, so again, it's hard to share and synchronise data.

IT vendors now offer a number of approaches to help tackle these kinds of issues.

Organisations can layer a customer management application across the top of their different systems, for example, which provides a single framework for capturing and subsequently analysing operational data and relating it to particular customer groups.

Similarly, a number of organisations have built 'data warehouses' to collate information from multiple incompatible systems. While these can be sizeable IT initiatives, they do provide a rich pool of data to analyse. One multinational insurance group now uses sophisticated analytical techniques within its warehouses to track past performance and work out which products are statistically likely to be the best to sell, and to whom.

With or without a warehouse, the leading IT vendors offer a range of applications and tools to support company-wide performance management initiatives.

Ultimately, the quality of customer performance management will also benefit indirectly from other IT initiatives that may take more immediate priority. In particular, most insurers are focusing hard today on customer-related process management to increase efficiency and quality of service. One important by-product of this kind of automation is that it generates a lot of cross-functional data that can feed analytical systems.