ROI, or what?

Nowadays you won’t find anymore a single marketer who is not worried about the particular ROI of their marketing programs. This is definitely the reaction by getting constantly asked by their CMO, CFO, and even their CEO about the ROI of a company’s marketing spend. A lot is written about this mounting accountability pressure for all marketers. Interestingly enough, most marketers respond by trying to understand, measure, and monitor the individual ROI of particular programs without too much consideration for the overall value of their marketing investments. They get pretty upset if one program ROI is negative, and try to find the fault either at their In-house creative, their agency partner, or within the executional challenges of that program.

I think this increasing focus on ROI is needed and can generate better programs but it misses one more strategic and long-term more relevant issue: How to understand the overall correlation between marketing spend and related impact and how to identify at which point on this correlation curve is the company’s marketing spend. Generally this correlation curve follows one simple pattern; the more you spend the less efficient your impact becomes (Curve of diminishing return).

I plead for a stronger scientific focus on understanding and computing this spend/impact correlation curve than on investing an increasing amount of time to fix the ROI of one particular program. I am aware of the fact that the overall marketing activities are a sum of all programs, but it is more helpful to understand the impact of all programs in aggregate than trying to optimize one small element. It’s just more helpful to understand first the structure of a forest than trying to grow one particular tree.

Additionally the first step toward understanding the spend/impact correlation does not have to equal impact with sales numbers but can expand the definition of impact. It can be a number of enrolled registrants on a website, or an improved number of awareness. Naturally a second analytical task will be to build a proxy describing the connection between the impact number and incremental company revenue but both analytical challenges (Spend/Impact Correlation and Impact Proxy to Revenue) can be treated and solved separately.

This recommended approach has the advantage that a marketer does not have to focus on all channels, all segments, or all programs but he can select one particular data rich segment and/or channel that facilitates the computing of one preliminary correlation curve. It’s much more productive to start with an easier understandable spend/impact relationship (e.g. DRTV or Telesales) than with the more difficult ones (e.g. Brand TV spend and retention impact). This is still a long and difficult process but it’s a smarter first step than trying to optimize singular marketing programs.