The Q-Q matrix

Editor’s note: Wade G. Boudreaux is director of marketing Danos & Curole, a marine contracting company based in Larose, La.

The arrival of data warehousing has taken market segmentation analysis to greater, more desirable levels. The concept of “marketing to one” is now a reality for companies with thousands - even millions - of customers. But how do we apply segmentation analysis concepts to small-to-medium B2B companies? How can you really segment a market that brings, say, $70 million in revenue from only 10-20 major customers, and better yet, what would be the advantage to segment such a small customer base?

I used to argue that segmentation analysis was not a necessary practice, given my firm’s market situation, but recently I have discovered a few useful ways to use market segmentation analysis on my small-customer base, which is made up of roughly 70 companies. Of these firms, about 20 represent over 90 percent of company revenue. So, how does one group these customers in order to create segments that require a unique marketing strategy?

One simple solution is a Q-Q matrix (Q-Q stands for quantity-quality). The first step is to get annual revenue and profitability figures for the past two or three years for each customer. Keep in mind that revenue and profitability are not the only variables that can be plotted on the Q-Q matrix - they just so happen to make sense in the case of my firm. Other variables that can be plotted are units sold to revenues per unit or perhaps number of ads to customer calls per ad - whatever makes sense for your unique situation.

In this case, let’s use revenue by customer (Rc) to profitability by customer (Pc). First, build a chart in which the Y-axis represents annual revenue by customer in dollars and the X-axis represents profitability by customer as a percentage of annual revenue. I define profitability percentage as whatever is left after all direct costs, overhead, and general and administrative allocations have been taken out, divided by the revenue from that particular customer. Next, divide the grid into either four or nine sectors - again, whatever makes the most sense. Each sector represents a defined segmentation category. So far the graph should look something like the one on the below.

Chart 1

Note that the upper limit of the Y-axis scale should be the rounded-up dollar amount of sales revenue from the highest revenue-producing customer. The upper limit of the X-axis scale should be a rounded-up percentage from the profit percentage of the most profitable customer.

The next step is to plot customers on the matrix and let the chips fall where they may. The graph below is a sample of a plotted Q-Q matrix.

Chart 2

In this particular segmentation method, each customer is classified in one of the following classes and class sectors:

  • Class A — small customers. Sector 1: low volume-low profit. Sector 2: low volume-medium profit. Sector 3: low volume-high profit.
  • Class B — middle-of-the-road customers. Sector 1: medium volume-low profit. Sector 2: medium volume-medium profit. Sector 3: medium volume-high profit.
  • Class C — big customers. Sector 1: high volume-low profit. Sector 2: high volume-medium profit. Sector 3: high volume-high profit.

Now let’s look at the implications of the segmentation analysis. In this case, implications from the Q-Q matrix may lead a sales and marketing manager to the initiate the following strategies:

1) Have account representative meet with Customer C, D, and N and propose a rate increase on products and services used. If not granted, fire the customer and redistribute resources to customers in Class A, Sector 3 and attempt to sell additional quantities of service at a higher profit.

2) Have account representative meet with Customers I and R and attempt to increase rates on products and services used, moving them into Class B, Sector 2 for now. If not granted, have sales manager attempt to get a volume increase at same rates as bulk rate discount. If still not granted, fire Customer R and put additional resources on Customers L, F, and I to get a volume increase at higher profitability.

3) Do not attempt further growth with Customer P at current rates, have account representative inform Customer P that any additional product or service sales will be at higher rates. If not granted, keep Customer P, but transfer a reasonable quantity of Customer P’s dedicated resources to Customers J and Q to attempt to grow additional volume at higher profit levels.

Real world

Now, many will argue that all of this looks good on paper, but in the real world there are other issues involved that may not allow for the associated strategies. This is absolutely true. Many times there are other issues not seen in revenue and profitability tables that may dissuade the “logical” strategy from being implemented. For instance, in the case of Customer C, there may be a perfectly good reason to keep this customer even if a rate increase is not granted and revenues and profits remain low. There may be a big opportunity with Customer C in the near future that will involve a bid for a large volume of business, with a chance for higher profit margins. The customer may have hinted that if we help him out now, he will help us out later (wink, wink). In this case, the sales and marketing manager may make a strategic decision to keep Customer C, even though currently it would be a better move to let him go.

Not the last word

The Q-Q matrix method of segmentation should by no means be the last word for dictating marketing strategy and application of resources. However, it can serve as a useful tool to take a look at the current customer situation and get you thinking about what strategies will be used to deal with different customer groups. It is also useful for viewing the results of an implemented strategy by plotting the following year’s figures and then comparing them to those from the previous year. At least it will help discern which customers deserve a gift and a phone call during the holidays and which deserve a basket of rocks and a trip to the woodshed.