Measuring the Short Term Impact of Customer Retention

As you will recall from my previous blog, I plan to extend the implication of customer retention on market share from the long term solution to the short term solution. As I pointed out in my last blog, the long term solution is good for planning purposes but does not accurately describe the short term effects on market share and may not be the best way to assess the value of a specific retention improvement program.

As I have discussed this topic with friends and colleagues I have received a great deal of support. They have all acknowledged that there is an obvious relationship between customer retention and market share. My analysis indicates the relationship is non-linear and hence may have a very dramatic effect on market share under certain combinations of conditions. This insight is what has led to this series of blogs.

There are two very different perspectives that can benefit from this analysis. The first perspective is from the point of view of the manufacturer. If the manufacturer has a strategy to develop its service business, it becomes imperative that the manufacturer put programs in place to create and maintain a high level of customer retention. For more detail on building and maintaining high levels of customer retention refer to my four part series on customer retention in the previous months.
The second perspective is from the point of view of an independent service organization (ISO). The independent service organization has the same need as the manufacturer to create and maintain a high level of customer retention. Unlike the manufacturer, the ISO does not begin with 100% of the customer base for a product. To the contrary, the ISO begins with 0% of the customer base and must attract customers away from the manufacturer. At the same time, the ISO attracts customers from the manufacturer, it needs to protect against customer defections.
In the past, ISOs have used price as the lure to get customers away from the manufacturer. While this may still be true to some extent, more ISOs are using multi-vendor service as the lure today.

As an aside, the point of looking at these two perspectives is to note that ISOs have a more difficult time building and maintaining market share since they must protect their base from defection while simultaneously attracting new customers at the same time. The manufacturer generally needs only to concentrate on customer retention. The importance of building market share is that market leaders get to establish pricing levels and also benefit from economies of scale that generally go with volume operations. While history indicates that services pricing levels have been the purview of the manufacturer, with multi-vendor service the market share leader may be an ISO that may have the clout of setting pricing guidelines for the industry.

I will continue the scenario of the manufacturer, the large, national third party maintenance organization (TPM) and smaller third party service companies each competing for a share of a particular product market discussed in the previous blog. For this blog, it will be assumed that the manufacturer has recently announced a new product. All three service organizations (manufacturer, TPM, and the smaller service companies) have the general ability to service the new product but only the manufacturer has the parts and will perform service exclusively during the warranty period. If the manufacturer maintains the same 90% loyalty (retention rate) for the new product as it has for other products, the TPM maintains an 88% loyalty, and the other smaller service companies maintain 86% loyalty, the question is, “how will the market share for the service of this product change over time” (Note: these loyalty figures remain the same as the previous blog in order to provide continuity.)
The effect of market share erosion from the manufacturer is shown in the table below. Note the three levels of market share erosion for a manufacturer. They are labeled 98, 90 and 80 respectively to indicate the market share erosion for customer retention rates of 98, 90 and 80 percent.
Market share
98 90 80
1 1 1 1
2 0.980 0.900 0.800
3 0.962 0.815 0.651
4 0.944 0.744 0.540
5 0.929 0.684 0.457
6 0.914 0.633 0.396
7 0.900 0.590 0.350
8 0.888 0.553 0.315
9 0.876 0.522 0.290
10 0.865 0.496 0.271
11 0.856 0.474 0.257
12 0.846 0.456 0.246
13 0.838 0.440 0.238
14 0.830 0.427 0.232
15 0.823 0.415 0.228
16 0.816 0.406 0.224
17 0.810 0.398 0.222
18 0.804 0.391 0.220
19 0.799 0.385 0.219
20 0.794 0.380 0.218
21 0.789 0.376 0.217
22 0.785 0.373 0.216
23 0.781 0.370 0.216
24 0.777 0.367 0.216
25 0.774 0.365 0.215
26 0.771 0.363 0.215

The time scale is noted only as periods of time. For the hardware service industry this scale would most likely be in months since customers defect from one servicer to another at the end of the warranty period or end of a contract (if they have a contract). Thus, for the example described, manufacturers with retention rates of 90% and 80% will be near their steady state market share (noted in my blog last month as 35.3% and 21.5%, respectively) within about two years. On the other hand, when the retention rate is at 98%, the market share of the manufacturer is still about 5% above the steady state market share. This implies that the higher the customer retention rate for the manufacturer, the longer it takes to reach the steady state market share. In other words, as retention rate increases for the manufacturer, steady state market share conditions take longer to occur.

The data in the table above are appropriate for new product introductions but are not appropriate for service of existing products. The first question asked in my last blog was: “given my current market position (market share), what will happen in the next year to my market share if I can improve my customer retention rate x percentage points?”

Obviously, since there are many scenarios (different levels of market share and different increases in customer retention rate); only two scenarios will be considered to demonstrate the concept. The first is from the manufacturer’s perspective. For this case, assume the same scenario noted above; namely, 90% retention for the manufacturer, 88% retention for the TPM and 86% for all other service organizations. Further, consider the market in steady state (market shares of 35.3% for the manufacturer, 33.1% for the national TPM and 31.5% for all other service organizations). Now, the manufacturer institutes a customer retention program that increases customer retention rates from the current level of 90 to 92%. The following table indicates the change in market share for the manufacturer for the next year following implementation of the program.
Months Manufacturer
1 0.353
2 0.360
3 0.366
4 0.372
5 0.376
6 0.380
7 0.384
8 0.387
9 0.389
10 0.391
11 0.393
12 0.395
13 0.396

Note that the change in market share during the ensuing twelve month period is about four percent. Thus, the payoff for the improved customer retention program of two percent, from 90% to 92% is four market share points. When the size of the market is known, this can be translated directly into revenue dollars and with further refinement into profit dollars.

As an example, if the service market for a given product is known to be ten million dollars (based on the number of pieces of equipment in the market), a four percent improvement in market share is worth $400,000 in revenue. If the service organization is working at a 25% margin level, this translates into $100,000 in increased profits once the new level of market share is reached. In fact, additional profits will accrue as the market share increases. So long as the cost of the customer retention improvement program is less than the increase in profit (over some specified time interval - such as two years), it makes economic sense to implement the program. The subtle implication to this situation is that the increase in customer retention, which leads to increased market share, also leads to increased activity for the organization. Thus, the concern is that the organization must be prepared to handle the increased workload and staff accordingly to maintain the increased level of customer retention.

The second scenario is from the TPM’s perspective. For this case assume the same scenario noted above; namely, 90% retention for the manufacturer, 88% retention for the TPM and 86% for all other service organizations. Further, consider the market in steady state (market shares of 35.3% for the manufacturer, 33.1% for the national TPM and 31.5% for all other service organizations). Now, the TPM institutes a customer retention program that increases customer retention rates from the current level of 88% to 90%. The following table indicates the change in market share for the TPM for the next year following implementation of the program.
Months TPM
1 0.332
2 0.338
3 0.344
4 0.348
5 0.352
6 0.356
7 0.358
8 0.360
9 0.362
10 0.364
11 0.365
12 0.367
13 0.367

As you can see, the change in market share during the ensuing twelve month period is about three and one half percent. Thus, the payoff for the improved customer retention program of two percent from 88% to 90% is three and one half market share points. The gain is slightly less than the manufacturer since the TPM had a smaller market share initially. When the size of the market is known, this can be translated directly into revenue dollars and with further refinement into profit dollars in the same manner as the analysis for the manufacturer.

Once again it must be clearly understood that these results are based on a change in only one of competitors in the market. When the other competitors see the change occurring, it is likely they will institute some kind of program to reduce the market share growth. The key point to get from this blog is that IMPROVEMENTS IN CUSTOMER RETENTION BEGIN PAYING OFF IMMEDIATELY. It does not take long for customer retention improvement programs to pay for themselves since they begin increasing market share almost immediately...as soon as the first customer is preserved that would have otherwise defected.
I will continue this analysis for the other two questions noted in last blog next.