Mining data: Helping deliver what customers really crave
Building business through customer relationship metrics
Published: April 26, 2006
By Fawn Fitter
Learn how to mine your databases for customer relationship metrics that will result in better services and offerings for your customers.
In Summary:
| • | Business drivers should direct the focus of a customer relationship metrics and analysis project. |
| • | You are probably only using a fraction of the customer relationship metrics and analyzing functions built into your existing customer relationship management (CRM) solution. Desktop applications offer help if you do not have CRM yet. |
| • | Outsourcing and Web-based solutions can deliver the marketing insight you need while keeping your information technology (IT) staff available to handle other tasks. |
An electronics manufacturer is losing sales to the competition. Why?
A local hardware store learns a giant chain is opening a store nearby. How can it keep customers loyal?
A health club wants to double enrollment in two years. Should it buy more elliptical trainers or hire more personal trainers?
These companies' sales records probably hold the answers, but they will remain hidden without a system for turning customer relationship metrics data into useful business intelligence.
Where customer analytics was once the exclusive realm of expensive consulting firms, today your company will find no shortage of efficient and affordable tools to turn consumer data into competitive advantage. A recent study by AMI-Partners, a consulting firm specializing in global small and midsize business (SMB) trends, showed that more than half of midsize companies have purchased customer relationship management (CRM) solutions or plan to purchase them in the next 12 months.
The following steps will help your business stakeholders evaluate options based on your organization's analytical needs. In fact, you may discover you need not make any further investments at all.
1. | Determine your business drivers. What do your business executives hope to learn from your analysis, and how will the results influence the products or services you offer? Those questions will pinpoint what data to gather and what patterns to look for, according to Eric V. Siegel, a former professor of computer science at Columbia University and now a senior consultant at Prediction Impact, a San Francisco, California, consulting firm specializing in data mining and analytics. For example, your salespeople may wonder why your online customers never buy your most expensive product. Is it because A) the product includes more features than your average customer needs; B) a less expensive product appears first on your Web site; or C) someone else is selling it at a lower price? Ask! Suggest generating a survey asking those three questions of everyone who looks at the product but does not buy it. When you push the resulting data into your CRM solution, you may discover people who answered the survey similarly share other traits, like geographic area. How can you use that information? If most people who answered C live in a particular city, you might want to program your Web site to offer shoppers with Internet Protocol (IP) addresses from that city a shipping discount or a bundled package that makes the product a better bargain. |
2. | Maximize existing applications. Even with no CRM system at all, remember that you can still set up manual methods of sorting through your customer information using Microsoft Excel spreadsheets, the filters and search functions in Microsoft Outlook, and desktop search tools, says Martin Schneider, enterprise software analyst for the global consulting firm The 451 Group and a former editor of CRM Magazine. This approach is time-consuming, but it can be a convenient way to answer such basic questions as: How many people sent us an e-mail this month to find out if we sell a product that does XYZ? If you already use a CRM package, investigate its built-in data storage, sorting and manipulation tools before spending more money buying and integrating another layer of software. For example, Microsoft Dynamics CRM has customizable forms, data fields and information relationships that make it easier to conducted detailed analyses. Most organizations that purchase a CRM solution use only 20 percent of its capabilities, says Laurie McCabe, vice president of SMB insights and solutions at AMI-Partners. She suggests working briefly with a competent value-added reseller (VAR) to learn how to get the most from your CRM deployment. Here's one idea: Link your CRM database to Caller ID to find out whether callers have done business with you before. Then suggest that your customer service representatives ask those repeat clients what new products or services they would like the company to offer. |
3. | Evaluate new software with advanced analytics capabilities. If your existing databases are in multiple formats and locations, you may choose to deploy a CRM product that integrates case histories, contact information, account information and other data. While your information technology (IT) staff will have to scrub and repopulate older data, gathering and analyzing new data will be significantly easier and more accurate, Schneider says. Also, many of these packages are now hosted online, eliminating the cost of purchasing a server. |
4. | Find a niche application that will work with your existing CRM solution to add only the functionality you need. Say you have noticed that specific offers on your Web site generate a significant amount of e-mail from customers. Software packages that integrate e-mail data and Web statistics will allow sales and marketing managers to track and predict customer reaction to online offers and then generate event-driven e-mail to customers whose online behavior meets particular rules, Schneider says. |
5. | Consider outsourcing. If you have a lot of data and a small IT staff, dedicating in-house resources to customer analytics may not be cost-effective, McCabe says. Online services that can tap into your databases to analyze data and deliver actionable reports are becoming increasingly common and affordable. For more information, visit the Web sites of DeepMetrix and Visual Mining, two such service providers. |
6. | Turn analysis to action. Once you have spotted trends in the data, you can help business stakeholders improve your company's sales or services processes. If your CEO can now see that the online checkout process is so complex that a significant number of customers abandon their shopping carts, or that customers will gladly pay higher prices for home delivery, or that the company lost sales because the product line is too small, she will know precisely what kind of new or enhanced products or services to consider. This may, for instance, include IT-related solutions, such as an ecommerce site redesign to simplify customer navigation and ordering. |
Beware of "analysis paralysis"
The greatest danger in launching a customer analytics program is trying to pursue multiple goals at once and becoming overwhelmed. In other words, you may be collecting and archiving gigabytes of customer data, but you should not—and cannot—use it all at once. Instead, choose one market channel to improve or one sales problem to solve. Analyze the relevant customer data. Create an action plan, put it into effect, and measure the results. Once you achieve a positive return on investment, apply what you learned to the next project. One targeted mailing, customer satisfaction campaign or salvaged deal at a time, your company can turn its customer information into gold.
Fawn Fitter is a freelance writer specializing in business and technology. Based in San Francisco, California, she has written for publications including Fortune Small Business, Worthwhile and Knowledge Management.
