Q&A with Donald R. Lehmann (Former Executive Director - MSI)

 

Former executive director of Marketing Science Institute (2001–2003) and advocate for the translation — literally of overworked jargon and figuratively of concepts useful to marketing practitioners — of academic research for everyday marketers, Don Lehmann practically needs no introduction.

Recently he shared some of his thinking on the relevance of rigorous marketing measurement with MarketingNPV Journal.

MarketingNPV (MNPV): Why have marketers and academics traditionally been on such different places on the marketing spectrum?

Don Lehmann: I don't think they all are. I think of there being two normal bell curves, representing the two groups, that intersect, and there are people in that intersection. Of course, for academics, thinking is their job and they tend to be more interested in theory. Similarly, a lot of practitioners have a "show me the money" attitude, but that is their job.

There is also a huge variance in managers. Some are very much numbers people and/or are in numbers companies that appreciate research. And there are others who disdain academics as too slow or too out of touch. Research I've long wanted to do would look at whether you could type management as to willingness to integrate academic research and general knowledge vs. just being focused on short-term gains and see if this typing is related to performance.

MNPV: You have written that best practice studies are not a best practice. Why not? How should marketers incorporate intelligence about their competitors and use their own data to set goals and objectives?

Lehmann: The problem with best practices is that they're not controlled studies. Just because companies are doing well doesn't mean everything they do is the cause of doing well.

We have to systematically look at lots of situations and deduce statistically what matters, control for lots of variables. We can't do what the people in medicine do. We can't give someone the placebo and someone Marketing Dashboard X. I use something called meta-analysis as a tool for empirical generalization. Instead of "GE does well, and GE does X, so X must be good" — a singularly bad way to assess causality and a neglect of complexity — meta-analysis looks across both well and poorly performing companies for factors that might be the cause or might work better or offer advantage under specific conditions.

MNPV: How can judgments of causality be improved? How can redundancy in metrics be identified and properly eliminated to yield a stronger, simpler tool?

Lehmann: I don't think there's an easy answer. You could gather tons of data and analyze it. You also could observe market changes, either planned or unplanned experiments in which there's a major change in what's going on, before and after. Most metrics are correlated with other measures so a reduced set captures most of the information. You also need links between measures. I'd have four levels in a causal chain:

The fourth question only applies if you're a public, for-profit company. If you're a church, your net present value may be converts. This is what I call product or people results. The fourth question is almost a constraint for non-profits. They have to find a way to get money, but that isn't their business objective.

MNPV: How do you get marketers to stop thinking of metrics as the problem and start thinking of them as a symptom of organic growth, if not a solution?

Lehmann: You might say that growth is the problem and metrics don't matter if you have the growth.

Given two scenarios, which would you choose? In one, your ROI is 30%, you have great customer service, and your growth rate is 3%. Or in the other, your ROI is 10%, you have many unhappy customers, but your growth rate is 48%.

Most people I know would choose the second. Now that's a short-term best choice. A business can't sustain growth with unhappy customers unless it's a monopoly, and then the question is, how long can you maintain the monopoly? In general that's not healthy, but the U.S. Postal Service was able to do that for a long time.

MNPV: What is your point of view on marketing dashboards? What's important to include in them to achieve organic growth?

Lehmann: You want to have both the diagnostic and criterion metrics. You need to include both in a chain of metrics, distinguish between the two, and find out how they are connected.

For instance, customer satisfaction likely is connected in some way to future profits. But you don't want everyone completely satisfied. The only way to get everybody satisfied is to drag in the last customer who didn't want the product you're selling anyway and get him or her to buy it, probably by dropping the price so low, he or she can't refuse. But you lose money doing that.

Diagnostic metrics can be important foretellers of what's coming in general, but not all of them. Therefore maximizing diagnostic metrics rarely, if ever, will help you meet your goals. Thus a lot of simple objectives — "always maximize quality," "always maximize customer satisfaction" — are pretty much always wrong.

Part of the problem marketing does have is that it doesn't have as common a model as finance does. Finance has an orthodoxy. Finance's orthodoxy isn't perfect, but at least it has a common model. If marketing moved in that direction, it would be in better shape. Making marketing more productive and generating organic growth are key to its mission.

A lot of what used to be marketing has migrated out to other parts of the company. That's great for the concept of marketing but terrible for the marketing function. Either arketers are really broad and consider themselves as business people who think they're really good at marketing, or they're creative marketers who can't see the rest of the business.

Companies will have different dashboards, especially on the diagnostic side, but I hope there is a core common to just about all of them. I hope marketing develops its own orthodoxy with maybe 10 basic measures: five basic criterion metrics — net revenue or gross revenue or sales share, stock price or market cap — and five basic diagnostics. (Frederick) Reichheld (of Bain & Co.) is recommending "Would you recommend this product to a friend?" as the new question to focus on. That could be one of the core diagnostic metrics.

MNPV: Could marketing dashboards become as prevalent and standardized in a core set of metrics as Six Sigma has become?

Lehmann: They could, but Six Sigma is like a mantra. It has almost nothing to do with the operations research procedures out of which it was developed. I coach football. Like in football, it doesn't matter as much what system you run as long as you do it well. So it is more important to have a dashboard that is widely used than to obsess over getting the optimal one.

MNPV: What questions are marketers not asking that they should be asking in order to accomplish organic growth?

Lehmann: It seems to me that they aren't thinking of how to integrate new product thinking and brand thinking and customer thinking or CRM (what the term meant before it became associated with technology). Managers are impatient. They want to bump sales right now, and that's why they tend to use promotions. Growth takes foresight and patience.

Of course, marketers often expect to rotate through a position in two years. They won't be there to pick up the pieces of poor, short-term focused decisions nor to be recognized for contributions to long-run growth. This presents no small challenge when trying to generate a long-term orientation.

To read Don Lehmann's article "The Relevance of Rigor," click here.