Begin With the end in mind
It seems that the more things change, the more they stay the same. In the November 7, 2001, issue of DM News, I ran across a study conducted by Accenture that reported, "nearly three quarters of the marketing executives in the United States and the United Kingdom say that their companies are unable to measure a marketing campaign's return on investment." I clipped and saved the article, and have since used it dozens of speeches.
Now fast forward five years, to late 2006. Suspecting (based on observation of current practices in direct and database marketing) that the problem uncovered by Accenture in 2001 had yet to be resolved by many companies, I wrote a white paper on the subject of campaign tracking -- detailing how it should be designed and implemented -- and my company offered it via the DM News "White Paper of the Day." The response from the direct marketing community that the publication serves was remarkable. More than 250 requests for the white paper poured in within the first hour! Clearly, measurement of marketing is still a serious pain point for a lot of people in the business.
At about the same time, The Economist reported, in an article on transformation of the marketing function, that 66 percent of marketing executives believe that the pressure to quantify results is one of the driving forces of that change. But the same marketing executives said that only 2 percent of their time and effort is spent on measuring performance.
What the heck is going on here? Results measurement is Direct Marketing 101, and should fall in the "duh!" category. The ability to measure marketing efforts is at the heart of direct marketing. Anyone who has been in the business for longer than a couple of months should have already learned how to set up a test matrix -- what the statisticians call an "experimental design" -- and should have access to the tables that define statistical validity at various confidence levels. Yet, the problem clearly persists. Why is that?
My conclusion is that it is almost always a planning problem, rather than a technological or process challenge. At the risk of vastly over-simplifying a complex issue, I believe there is a sure-fire solution to the problem of campaign tracking. Here it is: Begin with the end in mind.
At the very beginning of planning every campaign, ask these questions:
- What do we want to learn from this initiative?
- What metrics will we need to answer our questions?
- Do we already have access to the data elements we will need?
- If not, where will we get the data and when will we have it?
- Who is responsible for making sure that it will be there when we need it?
- What will the reports actually look like? Be very specific about this -- design the finished reports complete with rows and columns before you start, sans the actual measures that will come from the data at the campaign's conclusion.
- How quickly can those reports be produced after the dust settles on the campaign?
- Who is going to be responsible for producing the reports, and how will they get it done?
Don't take "I don't know" for an answer, and don't pull the trigger on any campaign until every single one of those questions is answered with specificity. Do that, and you'll never again face an angry CFO or CMO who is demanding to know why you can't quantify the return on your marketing investments.