We’re told that businesses can find tremendous benefits from analyzing the vast amounts of data they collect. And it’s true. In fact, you can find hidden nuggets of information among your data, once you know how to go through it systematically.
To give you an example, consider an international trade-show and convention organizer, which represents the epitome of silo culture. Operating through a number of discrete industry-specific events, publications and services, the company collected a huge amount of data that it brought into a single repository to be exploited in enhancing old services and creating new services for customers—and increasing revenue.
The business explored the resultant knowledge or “exploration” universe for the following purposes:
- To find important features regarding companies that visit, exhibit and advertise—or combinations of these.
- To identify key performers and appraise their relationship with the company across all activities.
- To assess the cross-sales activity of companies within one given year.
What it came up with was not only key knowledge with regard to lost advertising sales and lead times for reserving exhibition space but also extraordinary exhibition visitor behavior. The data analysis revealed the fact that members of the exhibitors’ staff were posing as visitors to gain additional free entrance for booth duty. It also discovered a sub-community of visitors—small mom-and-pop operations—that, it seemed, used events for networking with other very small businesses in the aisles, rather than visiting the exhibitors’ booths. The data showed that the members of the community pre-registered and regularly attended the events but never attended any of the free seminar sessions or used the appointment-setting service for visits to booths.
By sponsoring a small business forum at the show the following year, the event organizer capitalized on the knowledge. The sponsorship fee from the forum went straight to the bottom line.
Up-selling in retail
The knowledge to be derived out of customer data can also be used outside the realm of direct selling.
A major U.K.-based chain of specialist apparel stores used data it collected through promotional activity and prospecting surrounding their mainstream stores, their airport duty-free stores and special seasonal sale stores to identify the different types of shoppers. Using an incentivized market research questionnaire (by telephone and mail), the company could identify recency and frequency of purchase and how much shoppers intended to purchase in the future.
By adding questions regarding demographics and lifestyle, the company developed “profiled communities.” It began by looking at those core customers who had made a purchase in a mainstream store in the previous six months and who said they planned to shop there in the future and developing groups based on age and gender, number of children, type of car and value of watch.
By applying the segmentation models and looking at the key measurement process—recency, frequency and value (RFV)—the company was able to identify customers’ preferences with regard to type of store, location and products purchased, make sure offers and communication were relevant and target the customers most susceptible to up-selling.
The company went further, though. It created data profile “clones”—shoppers with the same attributes who had not made a recent purchase or plan to purchase—so it could target its offers to them.
The natural corollary to this is the tracking of combinations of product sales, often referred to as market-basket analysis. Knowing what products are purchased in combination can not only help a business understand its customers better but also has implications back into supply chain and merchandising.
Another retailer, with stores throughout the United Kingdom, in continental Europe and some cities in North America, applies this process, combined with exit-poll market research to understand the implications of products being out of stock. If product A is not available, will the customer still purchase product B (the substitute) and product C (the product most usually bought alongside product A); or has the total multi-item sale been lost?
Building profiles of products, matched to purchase-behavior of customers, provides direction to the store’s buyers and logistics managers, in terms of which products they reorder, the lead-time for reorders and which products are loaded onto the truck to ensure swiftest route to the sales floor. This improves the bottom line at the same time that it ensures customer satisfaction and supports the management of the customer relationship.
The increasing accessibility of data in real time and the ability to manage the resultant dynamics within the database will be key to the success of true one-to-one marketing. As these examples demonstrate, the complementary operation of tools for data acquisition, integration, manipulation, analysis and campaign management can provide a chain of triggers and actions to help marketers predict the potential outcome of their initiatives—and provide greater confidence in their execution and profitability.
© 2005 Michael Collins. All rights reserved.