July 12, 2006

An old marketing joke has it that half of all advertising spending is wasted. The problem is, knowing which half. (The line is attributed to John Wanamaker of the eponymous Philadelphia department store.)

Measuring the success of marketing programs and promotions is notoriously difficult to do well. This is particularly true in service firms with multiple channels like banks. Determining successful return on marketing dollars goes beyond tracking "purchases of new accounts," "saving on a regular basis," or "migration to cheaper channels for routine transactions."

In our own experience with clients, we frequently see promotions tracked either too narrowly to tell the client anything useful about the promotion being tested or too broadly to learn useful information about customer dynamics. As a result, clients have continued with "tried and true" marketing programs that, according to lore and legend, are thought to be part of the general growth the firm may be experiencing but are never truly known to be so.

However, over time, marketing promotions cost more than money. They also cost--and ideally benefit--the customer relationship that is at the heart of any service business. Though it may seem time-consuming, and like an unnecessary expense with lower-budget promotions, only by learning about how a promotion affects the whole customer relationship over time, in both short-term and long-term metrics, will you be able to judge its value. Only by identifying the specific strengths and weaknesses of a promotion will you be able to leverage that information to target the right potential customers, in the most effective way, for short- and long-term results.

The following example from a banking client is illustrative:
A U.S. bank sought to promote a new type of account in a local community with many immigrants. The account features an inexpensive wire transfer capability to a major bank in their country of origin. The planned promotion included three elements: in-branch merchandising, advertising in local freesheets, and some neighborhood presence (a car wrapped in a Mylar banner advertising the account). The objective was to drive traffic to the branches to open the new accounts. So what did the client need to measure?

Measuring foot traffic (new prospects). Since the objective of all the marketing elements was to drive foot traffic to the branches, we had to measure foot traffic.If you're planning your own test, you'll want to compare foot traffic both to the prior period, the same period a year earlier, and the same period at control branches with similar characteristics but whose customers (and prospects) were likely not exposed to the marketing.

You'll also want to confirm that any measured increase was driven by the marketing program, and by which element(s) of program. Again, primary research is required. In this case an exit survey was carefully constructed to provide the needed information with strict attention paid to potentially leading questions.We wanted to learn if visitors were aware of the merchandising elements, if their reason for visiting the branch was motivated by that awareness, and if their visit resulted in a purchase of the account or likelihood to make a future purchase.Measuring new account openings and wire transfers. The easiest results to track, and critical measures of success, were the number of new account openings and wire transfers from branches in the target community to the home country. But even with these measures how could we be sure that marketing efforts had anything to do with the success? (A traffic study offers some indication of motivation, but not a definitive answer.)

Perhaps it was created by the high-quality salesmanship of the branch staff who were helping customers who entered the branch for other reasons?

Could it have been achieved with only one or two of the marketing elements, without the others?

Measuring the profitability of the relationship over time. Assuming that the marketing promotion would be successful in its initial objective, driving in customers who can be sold the new account, we also wanted to measure whether these customers comprised a good, profitable business? We recommend you measure how these accounts grow over time.You might track how many services a customer is using six months or a year (or more) later, how many accounts they hold, their average balances and borrowings, and/or their revenues. Other metrics might include what volume of transactions these customers are performing, of what type and size, and through which channels.

You'll want to compare this information with metrics from other relevant customer sub-segments.Is this detailed tracking worth it for a simple marketing program? Yes. The cost of tracking results may be equal to a small marketing test, but if you don't learn more about the effectiveness of your promotion, you won't be learning more about your customers. You'll have no idea whether the program worked: in the short term to drive in new customers for the advertised product, and, in the longer term, to develop a profitable strategy for acquiring new business. Without answers to these questions you'll be condemned to never knowing which half of your budget is effective - and that's no joke.