Chapter 8 – Segmentation and Contact Strategy

There’s more than one way to achieve loyalty.  I took you through the nuts and bolts of the promotional currency model in the previous chapter.  Segmentation and contact strategy is another option.

The segmentation and contact strategy approach doesn’t typically require the same commitment as a promotional currency program.  You’re not making a promise to customers.  You might be making a promise to yourself and to your company.

This approach assumes you have customers linked with transactions.  Companies that do not have the customer/transaction linkage might launch promotional currency programs to create the linkage.  If not a promotional currency program, they’ll create a discount program that requires customers to carry and use an identification mechanism.  The Readers’ Advantage program at Barnes & Noble falls into this category.  Many grocery store chain programs also fall into this category.

Segmentation

This is not a book about segmentation.  That’s a good thing.  An entire book on segmentation might be just a little boring.

I’m going to focus on some basic segmentation approaches.  If you’re not using segmentation to drive a contact strategy, chances are you should start fairly simply anyway.  Don’t underestimate simplicity and common sense.

Segmentation is a way to group customers.  I’m focusing on ways to group customers based on their past behavior, not on demographics or anything else.  If you haven’t done much of this in your company the best place to start is always with past transaction data.

Segmentation Dimensions

Segmentation dimensions are the elements you use to discriminate different groups of customers.  One of the most common sets of segmentation definitions is RFM.  RFM stands for Recency, Frequency and Monetary Value.  Here’s what they mean:

Recency.  How recent a customer’s last transaction was.  You might calculate this in absolute days and assign a score to each customer to indicate if they’re in the top third, middle third, or bottom third of the entire customer base.  I used thirds just as an example.  You can also do it in tenths (deciles) or quarters (quartiles).  You can do it any way you want.  It’s your data.  But this is my book, so I did it in thirds.  It’s a simple place to start.

Frequency.  How often a customer has a transaction.  There are at least two ways to look at this.  One is to calculate average number of transactions in a period of time (quarter, year).  The other is to calculate a mean time between transactions (MTBT).  Both approaches get at the same thing.  Typically you assign a score to this just like you did with recency.

Monetary value.  How much a customer spends in a period of time (quarter, year).  Calculate a score.

If you’ve done each dimension in thirds, you’ll have a three by three by three cube.  You have 27 different segments.  Take each segment and calculate some basic metrics such as:

·        Average number of transactions (in a given period) or MTBT

·        Average spending (in a period)

·        Average transaction size

This allows you to logically group customers and understand their relative value.  I’ve used some very simple metrics.  Ultimately it’s good to get into contribution margin, or gross profit metrics.  If you overlay demographic characteristics on customer records it allows you to get a better sense of who the customers are.  I find demographics are better for adding texture to your segments.  Actual transaction data is better for defining the segments.

RFM is a very common approach.  Let’s look at some variations.  These are all cousins of RFM, but they might yield better insight depending upon your business.  Here are some dimensions you might want to look at.

Tenure.  How long has a customer been with your brand?  We all hear about how 80% of our business comes from 20% of our customers (although it’s rarely really 80/20).  But are they really the same 20%?  Not in all businesses.  Do you know the extent to which it’s true in yours?  Long tenure says something about a customer.  You have a durable relationship with a customer who has been with you for a long period of time.

Breadth of transactions.  This depends on your business.  If you sell many lines of products or services, it indicates the extent to which a given customer uses many or few of your products.  Breadth says something about a customer.  It says the customer understands what you’re about as a brand.

Channel use.  This depends on whether or not you have multiple channels through which customers may buy.  The most common pair of channels is off-line and on-line.  Customers who only buy off-line are likely to be quite different from those who only buy on-line.  You probably have more single transaction customers in the on-line only channel.  There is typically more trial in on-line businesses.  Customers who fall into the category of both, are probably more engaged with your brand.

There may be other dimensions unique to your business.  But this provides a place to start.  I’ve suggested six different dimensions, with some inter-relationships between them.  I would not suggest using them all together when you start.  Focus on a few that seem to make sense for your business and add complexity as you go.

Developing a Contact Strategy

Go back to everything you discovered in the CLA.  Especially insights about why customers select and reject your brand.  Why they stay and why they defect.  This will help you develop ideas for your contact strategy. 

For example, suppose you’re a gourmet food retailer and you discover during the CLA that your customers think about your brand principally for themselves.  They buy your products for self-consumption.  You identify an opportunity to position your brand as a source for gifts.  This repositioning may be an overall theme for your contact strategy.

Developing Objectives

We want to develop objectives for each segment.  If you have 27 segments (as our first example did) you might think about collapsing some of them together for the sake of establishing objectives.  You can look at all 27 segments when you’re doing the analysis, but you don’t necessarily need to define objectives for each one (because it’s likely that many objectives will look similar).  If you’re starting with the RFM segmentation, just look at the F and the M (frequency and monetary value).  That makes it a three by three matrix (nine segments) for the sake of defining objectives.

Now let’s set some objectives.  We’re going to set objectives for a year, and measure our results, by segment, over the course of the year.  We’ll review some of the measurement approaches in a later chapter.  There are several categories of objectives:

Revenue growth.  By getting customers to consolidate more of their business with you (share of customer growth), you’ll see increased revenue.  You’ll be able to determine whether that growth is from increased frequency or increased transaction size.  Set objectives as a percentage.  For instance, a revenue growth objective might be to increase revenue by 15% in all segments.

Improved retention.  By communicating with customers in a methodical, meaningful and relevant way, you should see an improvement in retention.  Set objectives as an improvement in the retention rate in each segment.  For instance, a retention objective might be to improve retention rate in segment one by five percentage points.

Changes in perceptions about the brand.  These are objectives that will be measured with research, not with customer transaction data.  Things such as increases in unaided awareness.  Likelihood to choose your brand first in the category.  Likelihood to complain rather than quietly defect.  These are important objectives that point to the future.  They help indicate changes in the long-term impression the customer has about the brand.  These impressions may lead to long-term loyalty.

Developing Budgets

The budgets you develop for contact strategy can be linked directly with the business improvements you’ll create.  However, the business improvements may not completely materialize in the short run (the first year) because you’ll need to do some testing.  When you’re testing, you often have fixed costs that are spread across relatively small groups of customers.  The cost per customer is high and the resultant return on investment (ROI) doesn’t look good.  However, things improve when you spread the costs over the larger quantities of customers when you roll out your efforts.

Good contact strategies can yield a 10%-25% lift in sales.  You must spend money to get that lift.  I’ve seen effective contact strategies that generate 25% to 100% ROI.  Using this information, you can develop a proposed budget.

Let’s assume the following for the sake of this example:

·        The customers you’re going to include in your contact strategy represent $25 million in annual sales.

·        Your gross profit is 40%.

·        You can consistently generate a 50% ROI on your contact strategy spending.

·        You can generate a 15% lift with your contact strategy.

Here’s the way it works out:

You spend one million dollars.  You generate $1.5 million in incremental gross profit. That’s based on a number of assumptions.  This is an approach for developing a starting point.  You’ll need to factor in your own assumptions and other unique circumstances about your business.

If one million dollars is your rollout budget, you’ll probably need 25% to 50% of that during the testing phase.  You won’t be achieving the 50% ROI or the 15% lift across the board.  You’ll probably need to go through six months to a year of testing to get to optimal levels.  Then you’ll need to do periodic re-testing to ensure that you’re always improving your contact strategy efforts.  It’s a journey, not a destination.

Developing Programs

You have segments, objectives and budgets.  Now we need to do some marketing.  Remember that we’re not talking about defined or announced programs.  We’re not talking about hard benefits programs.  We’re talking about communication and, potentially, unexpected benefits.  The media may be traditional direct mail and e-mail.  E-mail is obviously much less expensive, but it’s not necessarily more effective.  You’ll need to test with your customers in your situation.

Your ideas will be stimulated by your segment dimensions, your discoveries from the CLA, and your experience and intuition.  Let me suggest some ideas.

·        Recognition for highest spending customers.  A thank you message.  An exclusive offer.  I worked with a retailer that offered a “no strings attached” free gift.  Customers picked up the free gift at point of sale and almost half of them made a purchase along with it.

·        Stimulation programs for customers who have high average transactions.  These are very often your most profitable customers.  Even small incremental response rates generate large ROI.

·        Education programs for customers who only buy from one of your lines of business.  Teach customers about all you offer.  Leverage the passion they have for your brand into other areas of your business.

·        Referral programs.  Ask your good customers to introduce their friends and relatives to your brand.

·        Reactivate dormant customers.  Test algorithms for defining dormancy.  Test offers for reactivating.

·        New customer development programs.  If you can engage with a customer early in the process, you may improve the size and duration of the relationship.

Those are some jump-start ideas.  Combine those with your thoughts and you’ll be on your way to defining the first version of your contact strategy.  Start by allocating your budgets evenly across your segments.  I used three by three segmentation examples because it’s easy to grasp and manage.  Start by allocating one third of your budget to the lowest monetary band, one third to the middle and one third to the top.  After some testing you’ll begin to get a sense of where your biggest opportunities for ROI are.