Smart Use Of Data Can Improve Bank Insurance Marketing
Customer data may be one of the most underused assets banks have in marketing insurance and other financial products and services.
Many banks–and insurers as well–have focused on applying the data to meet operating requirements such as workflow management, risk management, claims adjudication and processing, and customer care. Few, however, have leveraged the power of the data to find new customers, align the appropriate agent with a pool of prospects, cross-sell existing relationships or manage customer retention.
The good news is that methods now available can make it easier to exploit the marketing value of the data. And many banks have an exceptionally rich base of customer data to develop.
The obvious question is, “If its so easy, why arent banks routinely doing it?”
The problem is that the best approaches, while effective and productive, are not well understood or easy to implement. No data-mining tool or customer relationship management system will magically convert the data into marketing results.
Banks–and, for that matter, all financial services–need a systematic program with a series of steps that can be applied over and over again to get effective results.
The core purpose of the most successful programs is to grow the value of the customer portfolio by systematically targeting financial services and product offers to the most potentially profitable customers. These approaches work well across the full range of insurance and financial products.
The following highlights seven of the steps that represent best practices in exploiting customer data for profitable growth.
Start with the customer data that are available. The conventional wisdom was to focus initially on building a comprehensive customer database that captures all potentially useful data. While data systems integration is commendable, it is both costly and time consuming, and it can fail to meet todays marketing objectives.
Virtually all banks are well served to begin with the data thats available: basic information on products held by customers, purchase dates, transaction frequencies and transaction size. Over time, the results can be improved further with additional and more accurate data. This learn-by-doing approach can help prioritize the value of new data and build the business case for making a richer investment in data mining.
Calculate the value of individual customers. A simple marketing measure of customer value can provide extraordinary returns. The calculation can be as basic as the revenue accruing from each product held by the customer. The measure can be enhanced by capturing variable marketing costs and behavior-based servicing costs, such as costs associated with customers visits to a branch or calls to customer service.
The calculation does not need to be viewed as an accounting activity. Its purpose is to measure the impact of marketing programs on customer value. The purpose of the measure is to target product or service offers to customers who will be the highest value users.
Target the offers on an individual customer level. The technology is now available to make it cost effective to use advanced predictive models that will clearly identify the best customers to receive an offer and to determine if the offer should be made directly to the consumer through the mail, by phone, by e-mail, through a branch office or through an agent lead-generation program.
Despite their popularity, segmentation approaches that cluster customers into large groups with common attributes have been found to be a weak targeting tool. Recent advances in decision-support systems have proved much more effective in identifying customers individually for each product or service offer, thus greatly improving marketing ROI. The accompanying chart illustrates the increased accuracy provided by the new modeling methods relative to conventional models or to segmentation.
Maintain complete customer demographic data. Inevitably, some customers have incomplete demographic information, much of which can be accurately imputed using advanced methods. In addition, extensive demographic data can be easily and cost-effectively appended to virtually all the prospective or existing customers through a number of syndicated sources. The demographic data is not as good at predicting as transactional and behavioral data, but it can add to the accuracy of marketing programs.
Measure the ROI of every marketing program. It is critical to measure the return on variable marketing costs from each campaign. A focus on ROI provides a systematic method for evaluating the relative success of each program, and it removes much of the typical guesswork involved in allocating marketing dollars.
Update the predictive models regularly. A hallmark of the best practices in managing customer data is a focus on continuous refinement, based on a systematic test-and-refine process for capturing results and updating the predictive models to reflect the latest results. This ensures that improvements will continue indefinitely.
Many practitioners let models decay well beyond their useful life. As a result, the model can fail to capture shifts in customer buying behavior or changes in the customer mix over time. With new methods, it is now efficient to update the models regularly to reflect continuing shifts in customer data and behavior patterns.
Manage the full customer life cycle. Many banks find that a small proportion of customers represent a large share of overall customer value. For some insurance products, 10% to 15% of the customers can represent 90% of the value.
The methods described above, when applied throughout the customer life cycle, are very effective in developing customers to their full value potential and retaining the highest value customers.
There are proven and accessible methods for accomplishing all of the steps described here. However, any bank planning to strengthen its approach may find the options difficult to sort through. The approach is likely to disappoint if it claims to encompass all that is needed in a software package, requires the bank to make extensive infrastructure changes or requires large investments and many months to install and develop.
Existing customer data can be applied systematically to generate strong, measured marketing returns. The returns will be realized in the form of increased sales and a greater return on marketing dollars, and the gains will continue to improve over time.
Stephen K. Pinto is chairman and a founder of Fortelligent, Inc., a Boston-based business analytics firm. He can be reached at email@example.com.
Reproduced from National Underwriter Life & Health/Financial Services Edition, July 22, 2002. Copyright 2002 by The National Underwriter Company in the serial publication. All rights reserved.Copyright in this article as an independent work may be held by the author.