"Pre-CRM" Testing Techniques - 
Determining the Potential for
Marketing ROI Before You Buy

First Published 2/28/01

If you're looking to measure ROI on an existing CRM project, see this article.

The following article is from the advanced topics section; you might want to take the tutorial Comparing the Potential Value of Customer Groups before reading it.  If you would rather see a general description of the Drilling Down method and specific benefits first, go to the home page.

CRM can generate increased profitability for your business in two ways:

1. Reducing costs, usually in the call center or distribution system (operational CRM).  The analysis of ROI benefits here is usually pretty simple if you understand current operating costs in detail.

It's possible to determine in advance whether the "customer marketing side" of a CRM initiative has the potential to increase the value of your customers.  Doing this exercise will also help identify any special needs you may have to consider when choosing a CRM package, and serve to educate participants in some core issues on customer valuation.  

You can use the following methods "pre-CRM" to estimate the potential for the marketing side ROI of CRM investment, or use them to prove plain old database marketing will be as effective as the marketing / analytical side of CRM when trying to increase customer value.  

If you have already implemented CRM and are looking for a way to measure the ROI of the implementation from the customer marketing side, see the "how to" details in this article.

Background

There's been a lot of CRM commentary on having a "relationship" with your customers.  Personally, I think this idea is a bit of a stretch.  How many companies do you want to have a "relationship" with?  CRM shouldn't mean playing "buddy-buddy" with customers.  You're in business to make money, not to fawn over customers.  You have to provide good service, of course - that's a minimum requirement.  

The fact is, many high-end CRM analytical packages rely on existing database marketing techniques in use for decades, but the techniques themselves have been lost in the CRM shuffle.  You can easily use these same proven techniques to model your likely success with analytical CRM before you buy. 

Relationship Marketing, the foundation strategy behind One-to-One and CRM, is not about having a relationship with the customer, it's about having a dialog (sound familiar?). 

This dialog is not a speaking conversation, it is about understanding the customer LifeCycle and responding proactively to it.  The LifeCycle is defined by the customer's behavior from the first day of business with you until they stop doing business with you.  It's the LifeCycle of the customer that determines LifeTime Value.  The concept of LifeTime Value without the LifeCycle is meaningless; if you don't know what the LifeCycle is, how do you determine the "LifeTime" to calculate a "Value" for?

If you understand the customer LifeCycle, you can react to changes in customer behavior, and these reactions to their actions are your dialog with the customer.  It's this action-reaction, back and forth, give and take dance that is Relationship Marketing.

You can increase LifeTime Value by affecting the LifeCycle, in one of two ways:

1.  Extend the LifeCycle, creating more time for the customer to increase in value.  This is usually accomplished by concentrating acquisition efforts on known long LifeCycle, high future value customers. 

2.  Increase average value within the existing LifeCycle.  This is usually accomplished by knowing when customers are most likely to engage in revenue-generating activities with you, and taking action to get the business.  

The implementation of  "anti-defection" campaigns can be used to address both of the above issues.  The following tests should give you a good idea if you have the kind of customer base that will respond in a profitable way to these techniques for increasing customer value.

Requirements of Tests

You need a customer database, with all possible transactions with the customer included in it.  "All transactions" is not a requirement; it's nice to have.  If your shop does not have this capability, focus on the highest value transactions, usually sales (perhaps page views or visits). 

If you're multi-channel and have data in different databases, it would be best to combine the data at the customer level.  If this isn't practical, do the following tests on each database separately.

You need an analytical tool, a software program you can use to query the database and produce reports (and someone who knows how to use it).  And you need a little bit of marketing money to do a test e-mail or direct mail promotion.

The Tests

The first test, customer acquisition, addresses your ability to extend the LifeCycle.  The second, Customer Retention, addresses your capability to increase average customer value during the LifeCycle and the potential effectiveness of anti-defection campaigns.  Combined, they should give you a good idea if CRM (or database marketing without CRM) will work in a profitable way for you.

Test 1, Customer Acquisition - Pick a start date, say one year ago, and calculate the current value of each customer who started with you in this month one year ago (you could use gross sales or total visits as a proxy for customer value in this case).  Sort customers by this current value.

Look at the top 10% best (most profitable) customers, and the bottom 10% worst (least profitable) customers.  Other than the revenue / profit variable, are they different?  Are they from different places, were they acquired by different methods using different offers, do they tend to buy certain products?  If you can find any significant differences between your best and worst customers, you have a shot at improving customer retention (increasing LifeTime Value) by fine-tuning customer acquisition methods.

The idea here is twofold:  If you can identify high and low value customers by some data element, you can allocate a larger share of budget to acquiring high value customers, and automatically target follow-up programs to "fix" low value customers.  If you have the data, look closely at elements of the first transaction (product, media, price, offer); the first experience of a customer heavily influences the LifeCycle / LifeTime Value. 

Test 2, Customer Retention - Pick the highest value generating activity in the database.  It's probably product sales, but it could be page views or other ideas.  Sort your customers by their last date of this activity.  Label the top 20% (those who engaged in the activity most Recently) of customers 5, the next lowest 20% 4, and so on, so the bottom (least Recent) 20% is labeled 1.  

Take an equal sized random sample of people from each 20% group, say 10%.  Send them a promotion with broad appeal; don't be overly restrictive in your offer.  You don't have to give away the store, just make sure the offer isn't "targeted" to any particular group.  For commerce, this offer might be "10% off anything in the store."  For content, it might be a free day of access to normally paid content as a "trial period."

When you look at response rates to the promotion by the "scores" 5 down to 1, they should appear similar to the following:

You won't see these numbers specifically, but the pattern should be evident.  The customers who are "5" should have a response rate anywhere from 5 to 40 times higher than the customers labeled "1," and response should noticeably decrease at each level.  

If you see this pattern, CRM marketing techniques, or what we used to call Relationship Marketing (using the customer LifeCycle to determine marketing strategy), will be able to increase the average value of a customer within the existing LifeCycle.  This kind of response pattern also indicates anti-defection campaigns will be a profitable way to both extend the LifeCycle and increase customer value within the existing LifeCycle.

The next test would be to try and manage customer value - to make money by creating very high ROI customer marketing campaigns and site designs.  If you can make money with simple database marketing techniques, you will have "proof of CRM ROI" and will be able to estimate your return on the CRM investment.

The Drilling Down book describes step by step how to do this.  You will learn how to create future value and likelihood to respond scores for each customer, and how to use these scores to continuously increase the value of your customer base.  Or I can do it all for you.