Is the RFM customer analysis model relevant for B2B marketing?

Continuing the discussion about customer metrics and analysis from previous posts; this post explores the relevance of the RFM (Recency, Frequency, Monetary) customer analysis model for B2B marketers and businesses.

The RFM customer analysis model has been around for over 40 years and is commonly used by Retail, Database Marketing, Direct Marketing, Non-profits and other primarily B2C businesses and marketing organizations. I have personally only encountered minimal use of RFM in B2B marketing but believe there is value in using this model in some aspects of B2B marketing depending on the specific circumstances of a business.

The premise of the RFM model is straightforward:

• Recency – when did a customer last buy? Research shows that Customers who purchased recently are more likely to respond to an offer than those who purchased some time ago.
• Frequency – how many times has a customer bought? More frequent buyers are more likely to buy again.
• Monetary value – what is the value of their lifetime actual spend? Big buyers are more likely to spend more than small buyers.

The RFM analysis ranks each customer for each RFM factor on a 1 to 5 scale (5 is highest). The 3 scores together are the RFM ‘cell’ for each customer ranking their historical propensity to buy with a 555 customer ranking being the best.

The RFM model has limitations and risks such as:

• Historical behavior does provide indicators for future behavior, but it’s not truly predictive.
• Continually targeting high-scoring customers could annoy or alienate them.
• Neglecting lower-scoring customers that should be nurtured.
• Just analyzing the numbers without relating the RFM score to specific business, product and marketing events and circumstances.
• The 125 cell (5x5x5) RFM model is too granular – rather group scores into clusters or bands to get a better picture of what the data are communicating.

The RFM model could be a valuable marketing analysis and segmentation tool to complement and qualify other analysis and segmentation tools used by B2B marketers:

• Relating customer RFM scores to lifetime customer value (LCV) can provide insights for developing and improving revenues from existing customers.
• In addition to the RFM score, the trend or migration between cells over time can provide further actionable information for marketing.
• The RFM score trend over time for major customers or segments of similar customers can provide insights into changing buying behavior and revenue performance.
• Relating RFM scores to results for various campaigns can provide insights into the effectiveness and appeal of particular campaigns for different RFM segments of customers.
• Relating RFM scores to products or product categories. For example, if a customer buys something in a product category do they usually buy more in that category or does it lead to cross-sell opportunities in other categories. Or if they buy something of low monetary value does that lead to buying something of higher monetary value or vice versa.

Do you use a RFM analysis in B2B marketing and if so, how has it worked for you? Your comments are always welcome.
Copyright © 2009 The Marketing Mélange and Ingistics LLC. http://marketing.infocat.com

Aligning marketing investment and campaigns with customer segments

Following on from the previous post ‘How many customers do you have? Really.’ which discussed basic customer count and group segmentation; this post explores some ideas for analyzing the segmentation for more effectively aligning marketing investment and campaigns. This diagram depicts the previously discussed basic customer count segments:

The fundamental customer objectives for any business are straightforward – acquire new customers, retain existing customers and grow revenues from existing customers. The challenge for marketing is how to effectively do this within budget and resource constraints.

Given this simplistic overall view, the next step is to categorize customers by value. One measure of customer value is how much revenue you have generated from a customer versus the total potential revenue for that customer. Let’s call this Realized Value – the percentage of the potential revenue already realized. We can now categorize customers by realized value:

• Most Valuable – customers with 75%* or greater Realized Value. These are the customers you most want to retain and keep active.
• Most Potential – customers with 25-75%* Realized Value. These are the customers you most want to grow, keep active and increase buying frequency.
• Marginal – customers with less than 25%* Realized Value. Although these customers may have lots of Realized Value upside, it’s a more difficult group to develop.
• Least Valuable – these are the customers from hell – the one’s that cause more problems, are never satisfied and cost more to manage than the revenue they produce. They could fall anywhere on Realized Value scale.

*suggested percentage – use appropriate measures relative to your business specifics.

These four categories should provide a good indication which marketing approaches would be most appropriate for each within the context of your business and market.

Now overlay these four Realized Value categories with the customer count segmentation discussed in the previous post and you’ll have an interesting matrix of customer insights to make objective marketing decisions:

For each intersection in the above matrix you would define specific marketing objectives, engagements, campaigns and execution programs. That should provide targeted alignment to most effectively align your marketing investment to produce better results from your existing customer base.

The concept of Realized Value is related to Lifetime Customer Value (LCV) which was previously covered in several posts; How to determine Lifetime Customer Value, Strategic Insights from Calculating Lifetime Customer Value and Impact of Customer Retention on Lifetime Customer Value.

I have more ideas to share on the customer analysis topic in upcoming posts. How does this approach relate to what you’re currently doing? Do you think this approach could improve your marketing results? Do you use a Relative Value type of analysis? Your comments are always welcome.
Copyright © 2009 The Marketing Mélange and Ingistics LLC. http://marketing.infocat.com

How many customers do you have? Really.

For many B2B and Information Technology companies the number of customers is a common and often cited measure of success and implied market share. We’ve all heard claims of “we have 50,000 customers” or “we have 300,000 customers” or whatever the number. While this gross number may provide some intuitive measure of market presence and size to the casual outside observer it does raise questions. More astute observers such as Industry Analysts will want further clarification and breakdown of the number. Marketing needs to segment this number to align investment and campaigns with the reality of various segments in the customer base.

Although most B2B companies are generally reticent to disclose these segmentation numbers to outsiders, they are critical internally for targeted and relevant marketing to existing customers. However, many B2B companies struggle to produce accurate and dependable customer segmentation numbers. The data may be in different databases, variable record-keeping over the years, acquisitions, divestitures, product life cycles, time, staff turnover, etc. all contribute to the difficulty for producing more granular customer counts. Here’s a basic segmentation breakout that all B2B marketing groups should know:

• Customers who bought – this is the most frequently quoted external number of customers who ever bought a product/service/solution from the company. A mostly irrelevant number as actionable marketing data.
• Customers don’t exist – the customer company no longer exists for various reasons. Identify and flag these records accordingly in your database. Never delete any customer records, just use appropriate status flags.
• Customers not using – those customers no longer using the product/service/solution bought from the company. These are ex-customers and should be flagged and counted as such.
• Customers currently using – those customers actively using the product/service/solution they bought. Although this may seem like an obvious number to know, it requires a continuing customer contact program to keep track of active customers.
• Customers in continuity relationship – these are customers that send you money on a regular basis for license/maintenance/service/support/hosting/etc. It should be straightforward to identify this subset from billing records.
• Customers who bought recently – there are two subsets to track in this group; those who recently bought for the first time and those who recently bought again. The qualification of ‘recent’ depends on the cost and scope of the product/service/solution – anything from 12-24 months.
• Customers who bought multiple times recently – these are customers that have made multiple independent purchases during the ‘recent’ period. Although this group could be a third subset of the previous group of customers who bought recently, the frequency attribute is important and should be of particular interest for marketing.
• Customers with unknown status – take all the customers who ever bought and subtract all the other subsets leaving a group of customers with unknown status (the remaining yellow area in the diagram). For companies that are large, or have diverse products lines, or done acquisitions, or been around a long time, this could be a sizable group of customers.

Customer records must contain product/service/solution line item data to provide more relevant analysis. For example, a customer may no longer use one product line but they could have purchased another product line within the past year.

The obvious observation from this relatively straightforward list of customer segments is that we should be marketing to each group differently to be most effective.

I’ll continue discussing some ideas for analyzing and using this information in next week’s post. In the meantime, it would be interesting to hear how you approach this customer count challenge or other comments on this topic.
Copyright © 2009 The Marketing Mélange and Ingistics LLC. http://marketing.infocat.com