Experts Denise Hopkins, David King, Pete Mehr, and Arthur Middleton Hughes weigh in on improving customer data for improved marketing strategies. From modeling to data enhancement, it helps to consider your ideal target’s buying behavior.
Denise Hopkins, VP, marketing and product development, Experian Marketing Information Services
Increasingly fragmented media channels that give consumers more control over marketing messages they receive make it more difficult than ever for marketers to deliver ROI demanded of them. To increase response rates and grow customer base, marketers must establish targeted, one-to-one customer relationshipsû and that means CRM.
CRM strategies are implemented with one thing in mind: results. To achieve strong results through CRM, ensure that every campaign and offer is the most relevant possible for the end customer. CRM strategies need be founded on the best data possible to deliver relevant offers.
The Takeaway:á Enhance your promotions with buyers’ attitudinal and behavioral data
A complete view of a targeted customer can bring relevance to campaigns. Marketer databases hold information that can help define best customers. Data can be enhanced with information from outside resources. This external data can provide more information to drive models or update information that has not been updated by recent transactions.
Marketers also need to look past the traditional demographic and geo-demographic data sets. Incorporating behavioral and attitudinal data will help marketers better understand what drives their customers and, in turn, to deliver a more relevant message.
Once marketers better understand customers by enhancing data, they need to apply that intelligence to their larger prospecting pool. Building custom models based on best customers and applying that to a prospecting environment, marketers can drive better campaign response through targeted offers.
A good CRM strategy is delivering the most relevant offer every time a campaign mails. This means marketers should continuously learn how to make their offers even more relevant based on customer response. By enhancing, mining and constantly learning from customer data, marketers will be able to implement CRM strategies that drive stronger ROI.
******
David King, CEO, Fulcrum
Building statistical models can be expensive, especially when multiple models are developed around the same problem. But newer techniques offer more value for the money.
Let’s say that we want to build a better response model for marketing our service or product to existing customers. One thing an analyst would consider is splitting the population into more than one group, and building maybe two or three different regression models.
The Takeaway: Don’t overlook the value of latent class regression models
Instead of a standard regression model, consider building a latent class regression. Latent class analysis has been used for a long time in many areas, including marketing research, but it is powerful for database marketing as well. In a latent class regression, the software creates segments that strengthen the prediction.
You’ll get a “trifecta” of benefits: a better response model for your customers, a mini-segmentation and demographic profiles of the segments that can be used as prospecting models.
A latent class model will improve prediction over a standard regression model built on the same population. Each segment will be tied to the likelihood to respond, and the difference in segments can be used to tailor messages and offers. Because the segments are linked directly to response to your offers, we can borrow that information to identify prospects more reliably than a traditional “look-alike” model.
Traditionalists, who may be concerned that this is yet another modeling black box, need not worry. You have the same visibility into the workings of the model as with a standard regression model. There are a number of commercial packages available, but the open-source R statistical program can be used for those who want to try it for free.
******
Pete Mehr, VP and client leader, Merkle
Typically, database marketing enables quantitative direct campaigns, while media planning determines cross-channel spend using largely qualitative methods.
Technology and new ways of thinking, have made it possible to combine both elements and use the content within a database to optimize media mix allocation strategy and improve results.
Media measurement determines the impact of marketing efforts by analyzing spend across channels. When done right, it enables marketers to determine the cause and effect connection between marketing actions and results. Marketers can then optimize spend across all channels — and allocate media dollars effectively.
Imagine using your database to dynamically adjust your media spend across specific channels, based on results from your latest campaign. You could increase the number of radio spots and direct mail in one market while simultaneously decreasing print ads û and improving results. Media mix allocation makes this possible.
Using your database infrastructure to perform media mix allocation involves integrating three main elements: competitor data, response data and historical media spend data.
The Takeaway:á Integrate competitor data, response data and historical data
Track what and where your competitors are spending. It is a major component of the media optimization process. Regarding response data, think on a granular level û down to the creative on your last e-mail sent. Typically, historical media spend data is housed in another database or system altogether. Include it in your database infrastructure.
Once the data are collected, statistical and optimization models can be developed to quantify value of each channel and to determine optimal spend level for each one. Marketers are in the best position further allocate media spend when all this information is tracked and measured. Potential savings that result from media mix allocation can be as high as 20%.
******
Arthur Middleton Hughes, VP /solutions architect, KnowledgeBase Marketing
By classifying your customers into the following four segments, you can focus your marketing efforts on the one segment that is really profitable: relationship buyers.
Bargain hunters, such as Wal-Mart, have tremendous market power. They demand – and get – the absolutely lowest prices in the market. At the same time, they demand — and get – a very high level of service from suppliers. For these customers, the supplier has to meet specified requirements in order to sell massive amounts at very low prices.
The Takeaway:á Don’t Miss opportunities to market to relationship buyers
Program buyers don’t have the time or economic incentive to shop for the best deal. They have the worst of both worlds: they pay the highest prices and get the lowest level of service. Their purchases are usually small. They’re generally unprofitable and not worth cultivating.
Transaction buyers’ purchases are large enough to make it worthwhile for them to engage in comparison shopping for every transaction. They have absolutely no loyalty. They will shift suppliers for a few pennies. Service is not important — price is everything. Database marketing will not work — only discounting. They are not profitable customers, even though they buy a lot of product.
Relationship buyers are people for which database marketing was invented. They’re looking for: someone who cares about their needs, and who looks out for them; someone who remembers what they bought in the past, and gives them special services as a reward; someone who takes an interest in their business and in them as individuals. They know that they could save by shopping around, but they would lose the relationship that they have built with a dependable supplier.
Your database is used to record the purchases of these buyers, and to give them personal recognition and special services. You set up Gold card categories for them. You communicate with them. You partner with them. Partner with them.
