Monday, August 07, 2006

100 Measurement Stories Don’t Add Up to Any Great Insight

Most large marketing organizations have made significant strides in the development of sophisticated methods to improve marketing measurement. Ph.D. mathematicians are commonly on staff, stewarding elaborate survey research, media-mix models, and analytical models for assessing the return from proposed initiatives.

But step back from the complexity and one can’t help but wonder if all that measurement is being approached in too tactical a way to credibly tell the overall story of marketing effectiveness and efficiency. With few exceptions, marketing departments appear to be measuring payback in a disjointed series of technically sound but ad-hoc ways in four distinct measurement silos: customer metrics, unit metrics, cashflow metrics, and brand metrics.

The customer metrics silo often looks at how prospects become customers. From awareness to preference to trial to repeat purchase, many companies track progression through a “hierarchy of effects” model to track evolution of broad market potential to specific revenue opportunities. Satisfaction with the customer experience is measured by surveys and reported by channel and touchpoint, although rarely in correlation to specific customer behaviors.

In some companies, the customer metrics silo includes robust attitudinal data on customer segments — why they want what they want or buy what they buy — which is often correlated with actual customer transactional data to create a robust segmentation model. The segments are then monitored for “mobility” (the directional progression of prospects/customers from one segment to a presumably more valuable one) and velocity (the speed with which customers are moving between segments). In many B2B organizations, this customer pathway can go all the way to developing a customer-specific P&L.

The unit metrics silo is the one likely to be at an advanced state of maturity in most companies, owing to the underlying IT systems ability to tell what was sold, where, and at what price. Most marketers have fairly good information on how many redheaded, left-handed, overweight men in their 40s have purchased a minimum number of units in the prior six months with an “r” in them. (Yet surprisingly few know the identity of the individual they actually sold it to.) With some quick math, they can figure out the marketing cost per unit as a gross method of measuring efficiency. Some further mathematical gymnastics can get to pricing optimization analysis, which in turn can provide some insight (albeit a bit oblique) into the value of branding.

The cash-flow metrics silo focuses on efficiency of marketing expenditures in achieving short-term returns. Program and campaign ROI models measure the immediate impact or net present value of profits expected to be derived from a given investment initiative. Media-mix models use statistical regression techniques to identify which combinations of media placements, integrated media elements, and even copy executions generate the most profitable response from customers. And armed with those insights, the marketing department can demonstrate how it is optimizing resource allocations toward the activities and executions with the greatest forecast return in a sort of “portfolio management” exercise.

The brand metrics silo often tracks the development of the longer-term impact of marketing through brand health. Survey-based tracking studies gauge customer and prospective customer perspectives on the brand — its functionality, personality, accessibility, and value propositions. Brand scorecards monitor the evolution of these perspectives over time within market segments and across multiple constituencies like employees, regulators, and community influencers to get a full view of brand equity drivers. And many have taken the successful leap to develop financial models for estimating the financial value of the brand as a means of determining the aggregation of assets on the balance sheet as an outcome of marketing investments.


Yet despite the implementation of effective measurement systems within one or more of the silos, most marketing departments still struggle to synthesize insights gained across silos in a manner that helps one silo explain another or clarifies the predictive drivers of the business on a broader level.

For most companies, it’s not possible to do this scientifically since it’s not an econometric modeling problem solvable by equations and computers. Each silo measures very different components of marketing effectiveness in very different ways. Some are shorter term and some longer term. Linking them algorithmically forces you to make some very large assumptions that may be unreliable in the face of actual marketplace dynamics.

Even if you can solve the problem algebraically, you will likely have to employ statistical techniques of such sophistication that few people in either marketing or finance will understand sufficiently to embrace and defend the method.

The net effect of all this uncoordinated measurement is that marketing gets lost trying to divine the true story of effectiveness of resource allocation from 100 data points on a three-dimensional scatter plot with no clear picture emerging. And while it may have been accepted practice in the past to throw this measurement spaghetti at the wall when asked about the payback on spend, today’s CEOs and CFOs have little patience for the fog of complexity.

To tell the complete story of effectiveness and efficiency of marketing investments, consider developing a marketing dashboard. A dashboard can structure many disparate sources of information in a comprehensive, organized manner and present the insights derived from each silo in a graphically related view that facilitates the human brain’s incredible power to find subtle, contextual links. A well-designed dashboard suggests to the user that the many individual component metrics are actually all part of one single story, not a jumble of dozens.

The debate on the “art” or “science” nature of marketing is over. It’s both.
The science is reflected in the mathematical, cognitive, and behavioral tools we employ to identify opportunities and gauge our success at exploiting them. Our repertoires are expanding with every passing year as more researchers develop better tools and techniques.

The art has historically been defined as the creative spark of imagination behind our execution of marketing messages in words, pictures, and forms used to engage the customer.

Today, the art is increasingly needed to help make sense of the science. The best scientific measurement techniques are lost on the audience that suffers through dry and uninspired soliloquies of interpretation, or, worse yet, death by 100 pages of charts and tables.

As true marketers, we should be able to paint a picture to tell a better story. The dashboard can be a powerful canvas.

1 comment:

MineThatData said...

All too often, geeky folks like me fail to tell a compelling story with metrics. We math folks need to humanize the information, so that it is trusted. When information across silos is trusted, then change can happen.