Dear Amazon.com Customer,

As someone who purchased an HD DVD player from us before February 23, 2008,* you might like to hear about a special offer available from Amazon.com.

New technologies don't always work out as planned. We at Amazon.com value our customer relationships more than anything and would like to support customers who purchased these players by offering a credit good for $50 off any products sold by Amazon.com.** Just use promotional code DHA8-HKVVYZ-????? when checking out. The code is valid through April 9, 2009, so you have plenty of time to use your credit. Purchases from third-party merchants on our site are not eligible.

In addition, we'd like to share some of our top offers on Blu-ray discs, HDTVs, and other high-def technology and remind you that the Amazon.com Marketplace is available to sell items you might not want anymore as you upgrade to new ones. Also be sure to check out our monthly Amazon.com Early Adopters Delivers e-mail to find out about the latest technology.

This may be a retailing first, viz. offering a customer a substantial credit on account simply because a technology they sold lost out in the end. 

Remember when AMAZON was chided for consistently failing to show a profit early on?  He who laughs last, laughs best.  This great company continues to shine.  My hat's off to them!

// lm 

Trying to determine precisely what each axis represented left me dazed and confused.  But a different exhibit, following our email exchange, made things much clearer…

In this example, roughly 40% of customers prove to be unprofitable following an ABC analysis.  This group erodes the overall profitability of the business due to the high cost of servicing their accounts.  The next step is to determine why these customers are unprofitable and to look for ways to reverse that trend — ultimately either by lowering costs or raising prices.  For example, some customers may be receiving service(s) at a higher cost to the company than the norm.  Special orders and rush deliveries, for instance, would increase the cost of fulfillment and consequently could make a given sale unprofitable.  A customer who atypically and routinely drains a companies resources beyond the norm might prove quite unprofitable in the final analysis.

Brian also provided some examples from his consulting practice that drive the point home even more ..

Finally, here's another representation of the hook curve which I think helps to bring the matter even more clearly into focus.

       I think there's considerable merit and food-for-thought to be found in this ABC costing approach to profitability analysis.

// lm 

What do you suppose would happen if you mailed a print catalog to 1x-buyers who placed their order on the web? Do you think they'd reorder at higher, lower or the same rate as the same web shoppers who did NOT receive a catalog? 

What about multi (2x+) web shoppers (again, catalog vs no catalog)? 

What about web shoppers who hadn't placed an order in well over a year? 

Internet buyers often behave quite a bit differently from their catalog counterparts.  So I would have guessed that mailing the catalog would have little to no effect on the web channel's reorder rate.  As it turns out (and this isn't in just one isolated test case, either) the catalog mailing significantly depressed the reorder rate under all three of the above scenarios!  Consistently and dramatically, too.  In short, he observed that web shoppers who received a catalog reordered at a substantially lower rate than those who did not.  Here's a few excerpts from his article: 

"In a recent test we conducted with a client, we created two panels of 50,000 customers each, both groups acquired on the marketer’s e-commerce site within the past three months — neither had been mailed a catalog.

One panel was mailed a 48-page catalog; the other was not mailed. Based on a later matchback to the mail files, the one-time buyers not mailed the catalog responded at 3.7 percent, while those mailed the catalog responded at 2.1 percent. Of the two-time-plus buyers, those not mailed a catalog responded at 8.5 percent; those mailed responded at 3.6 percent.

Panels also were created for older buyers. For buyers who previously had purchased more than 13 months ago, those not mailed responded at 3.4 percent; those mailed responded at 2.8 percent. This isn’t a one-time test. We’ve repeated this test with several clients and seen similar results."

This is fascinating and worthy of careful consideration given that one might actually be losing money - and not merely wasting it - by sending catalogs to internet-driven shoppers at all.  In an attempt to get a better handle on the implications of this finding, I wrote to John with a few questions … (more…)

1. They reordered only 1.2x on average compared to 3.4x for existing customers;
2. Their future sales and profit were on average 46% and 57% lower, respectively, than for ads run previously at the regular price;
3. Veteran customers who responded to the ad subsequently reordered at an 18% lower rate than their cohorts.

In short, the new ad attracted lower quality customers who — on average — ordered much less frequently and spent far less money than existing ones. The result? The CLV of this new group of discount-acquired customers was dramatically lower than previous ones acquired at the normal/higher price point.  Just as bad, existing customers who responded to the promotion were "tainted" by the experience and became less valuable buyers afterwards.  This seemingly positive promotion turns out to have negatively impacted the health of the business.  Unfortunately, such long term effects are (i) frequently overlooked, and (ii) often lead to poor strategic decision making resulting in lower sales and profit in the future.  Difficult though these factors may be to measure, it's essential to make your assessments based upon the long view, not the short one.  So, when you're evaluating promotions be sure to think tail  — not head. // lm

Calculating future LTV – which is what really matters – is neither simple nor straight forward, however. LTV is driven essentially by customer lifetime duration (LTD) and future purchases. How much a customer is likely to spend , how often, and for how long will together determine their LTV. To the extent that these may have been poorly estimated up front, the accuracy of any subsequent analyses will be compromised.   That huge challenge aside, what’s even more difficult is to qualify and quantify the relationship between discreet business decisions (primarily strategic but sometimes tactical) and their back-end results. In other words, assessing which business drivers had what direct and specific impact on LTV.

For example, even if we can reasonably estimate that LTV has increased, say, 15% overall, how do we tie this increase back to a particular driver when many may be at work? Was it our redesigned website, expansion of our product line, more liberal returns policies, or perhaps the increased production from our new factory that is primarily responsible? Whether a particular strategic driver had a positive or negative affect is difficult enough to discern. Assigning its quantitative score is typically next to impossible. Therefore, while it’s perfectly valid to assert that it’s the impact of a given business driver on customer lifetime value that’s most important, it’s just as important to recognize that leveraging this principle is exceedingly difficult due to the sheer complexity of the numerous interactions taking place – especially internal, but also external as well (e.g. the actions of competitors and dynamic changes within the marketplace itself).

One approach (which we call “Deltalytics”) is to periodically estimate the average customer lifetime duration and average customer spend (among others) and subsequently track their change over time to expose trends that will ultimately govern future business performance. If we know that both are increasing, for example, it’s safe to say that the business is trending upwards. The rate at which this is occurring can, of course, be used to make specific predictions about future growth (or decline, as the case may be). But these two metrics are just the tip of the iceberg. Others that can, and should, be used to gauge business performance include:

1. Rate of new customer acquisition (and, conversely, attrition)

2. Customer distribution by recency (the greater the proportion of recent buyers, the better the business will perform)

3. Average latency (the sooner customers place subsequent orders the better)

4. Customer distribution by frequency (the higher the better, although not nearly as predictive as recency)

5. Multi-buyer conversion rate (the percentage of 1X buyers who become multi-buyers)

6. Customer re-order rate by recency (the ratio of repeat buyers as a function of their recency segment, e.g. <30 days, 30-60 days, etc.)

7. Customer reactivation rate (customers flagged as having lapsed but eventually reordered)

8. RF Delta (the change in population density over time at the intersection of recency and frequency)

Although quite useful in themselves, the greater utility in each of the above metrics lies in evaluating and forecasting their deltas over time. Change, and the rate thereof, is far more meaningful and insightful in this context than the more common “static” approaches to predictive analytics. Getting back to business drivers (and measuring their impact on the bottom line, viz. LTV) one must concede that no single solution or approach can effectively gauge them all.

At some point, a seat-of-the-pants determination must be made based upon relevant, albeit inherently incomplete, data. Tests can be conducted to measure the impact of, say, introducing a new product line or instituting wide scale changes in pricing. But even then, other contributing factors that cannot be controlled for, and are likely to cloud the results, must be acknowledged (such as a new website, outsourcing the call center, and so on).

In a perfect world, strategic changes would be implemented in a linear and mostly piecemeal fashion to ensure that consistent and reliable analyses can be made. Because this is so rarely possible, however, some compromises must be made in terms of measuring and forecasting the impact of such changes.  It is our position that an optimal way to approach the problem is to analyze trends amongst the aforementioned business performance measures – more specifically, their change (and rate thereof) over time, and subsequently tying these back, as best we can, to their underlying business drivers. This, unfortunately, is much easier said than done.

NOTE:  Originally posted (although edited since) on David Raab's Customer Experience Matrix blog.  I didn't want anyone to think I had stolen it from myself ;-.

// lm