We’re alums of P&G (Procter & Gamble) and both grateful and proud of it. P&G is where we cut our first marketing teeth. The company hired Mike despite his not having an MBA. (He has dual masters’ degrees in English and Human Resources.) They hired me despite an MBA with a concentration in finance, not marketing!
My offer was made before I started my final semester to graduation. At the time, I was slated to take directed studies in finance, more specifically in mergers and acquisitions—considered an area of future major importance. I asked P&G management if I should switch my directed studies to marketing. Their reply was for me to study what I wanted as they would teach me everything I needed to know about marketing.
Well, neither Mike nor I can say we learned everything we know about marketing from P&G. The learning never stops, nor should it. However, we learned what the vast majority of today’s marketers have not learned— what it takes to achieve authentic marketing excellence. We learned to practice and lead with evidence- versus eminence-based marketing.
Eminence-based marketing is mindlessly engaging in entrenched marketing practices. These practices include doing things the way the industry or category does or has done them. It also bows to authority—your boss’s and his senior management’s. It’s driven by unexamined experiences, intuition, and blind obedience to conventional wisdom. Eminence-based marketing is neither supported by proven principles, best practices, nor quality processes.
Evidence-based marketing separates the wheat from the chaff. It identifies causal relationships that lead to line-of-sight sales. I’m not talking about clicks. I’m talking about hard sales and changes in market share. Evidence-based marketing determines what works and how well something works under what conditions in both the absolute and relative to alternative actions.
In their book, THE POWER OF EXPERIMENTS: Decision-Making In A Data Driven World, authors Michael Luca & Max. H. Bazerman, tell the story of eBay changing their advertising strategy on Google to save millions of dollars (Chapter 6, eBay’s $50 Million Advertising Mistake). Namely, eBay had been annually spending $50 million advertising on Google because the marketing team believed it was a good investment—an eminence-based decision of intuitively doing what they had been doing.
However, at the urging of a tech economist, an experiment (i.e., testing) was conducted. The experiment revealed that much of eBay’s advertising budget on Google was a waste. Moreover, the marketing team learned under what situations the ad funding on Google boosted sales and profit. They used the learning from their experiment to change their strategy and improve their results. They made an evidence-based choice.
If you aren’t engaged in evidence-based marketing, you’re a long way off from marketing excellence!
Evidence-based marketing begins with a hypothesis. What we think we know is often nothing more than a hypothesis. Inclusion of KOLs, medical congresses, couponing, trade discounts, sampling, demonstrations, whatever we put into our marketing plans, and actions we take, we believe, will lead to growth. Well, maybe or maybe not. We don’t know until we test and measure for a relevant outcome. Even if it does stimulate growth, the question remains, will they lead to the achievement of our sales and marketing share goals and deliver an attractive ROI?
It is essential to identify the relevant outcome and test and measure for it. Again, the relevant outcomes are trial, adoption, switching, repeat purchasing—those things that contribute to and include sales, market share, and profit. Moreover, it is not enough to know that a potential action works in general or has worked for other brands. We need to determine if it will work for our brand, given its current or anticipated situation.
The Vacuum Folger’s Coffee case provides a good example. I was fortunate to manage P&G’s Vacuum Folger’s Coffee expansion into the geography east of the Mississippi River in the early 1970s. Based upon significant P&G experience, the hypothesis was that sampling was the most effective way to introduce to achieve our sales and market share goals. There were a few additional hypotheses to be tested. These included: sample size (6-ounce versus 1-pound); sampling versus couponing; and coupon values. The measures were on levels of consumer trial, conversion, and repeat purchasing behavior. These had to coincide with sales and market share goals for the brand.
The next step was market testing. Coffee purchasing households or promotion area markets matched for demographics received either a 6-ounce or 1-pound sample of Folger’s Coffee, or coupons of differing values ranging from 25-cents, 50-cents, or $1.00. We chose two sample sizes to test since we did not know how many cups of coffee, or consuming occasions, one needed to become acclimated to the taste of Folger’s versus their coffee brand. Moreover, the alternate coupon levels were to address the impact of the incentive. Coupon redemptions tend to be historically low. Their value is in stimulating trade merchandising, which was isolated out of the test.
Stepping back, in addition to market testing, we also used ballot testing at P&G. Suppose we wanted to employ a consumer near-pack display promotion. A near-pack was a give-away to the consumer when s/he purchased our product at retail. Through our marketing research arm, we’d test several potential near pack items ranging from books to toys to cooking utensils such as spatulas. Even within a winning category of near-packs, such as books, we’d test alternate titles. Those winning near-pack items would then proceed to a market test before being used on a broadscale basis.
We conducted many analyses with the guidance of our marketing research managers. One analysis included reading sales and market share data. Sales and market share data was available when we tested our hypotheses in-market using different, highly defined promotion areas. When we tested in promotion areas, we also included markets as no-effort controls and markets. We also included markets where we reintroduced a past action with a proven track record of producing highly favorable sales outcomes. Thus we could determine the absolute and relative impact of the action under review.
Otherwise, we needed marketing research, most notably something such as an AA&U (i.e., Attitude, Awareness, and Usage) Study, which measured attitudes toward our brand, sales, and repeat purchasing pre- and post-action. Sometimes, as in the case of Vacuum Folger’s Coffee, we conducted both types of analyses. In practice, we conducted an AA&U every two-weeks in Folger test markets and received Nielsen market share data every month (versus the standard two months). We examined sales weekly.
We marketers were tasked with analyzing the marketing research and making recommendations based upon our learning. However, our analyses were not forwarded to senior management until they were reviewed and approved by our MRD (i.e., marketing research department) managers. Each report included a statement that went something like this: “This analysis has been reviewed and approved by MRD (Ms. or Mr. “So and So”) who agrees that the conclusions are technically correct and consistent with the findings.”
The next stage was gaining senior management’s approval of our recommended action(s). Their support was based upon the evidence and nothing but the evidence. Did the testing prove the hypothesis or hypotheses? If so, we received their go-ahead to proceed. The successful ballot testing led to in-market testing. Successful in-market testing led to regional or nationwide execution. Moreover, our findings were added to the P&G database, so other marketers from other brands, both within and outside of our division, could learn from our testing experience.
Every single one of P&G’s marketers on every single one of its brands worked the same quality process—an evidence-based best practice to demonstrate and achieve line-of-sight to sales. You could not get by with eminence-based marketing as marketing funding is considered an investment in the business. Investments are intended and expected to bolster sales and deliver a highly favorable ROI.
Senior management demanded that we conduct marketing testing to support evidence-based marketing. Our marketing plans ALWAYS included a testing plan consisting of several or more tests that we would undertake to address critical issues and test hypotheses.
Back to the Future
The database of all tests conducted by P&G brands was available for marketers to review. However, if an action proved highly impactful for a given brand, we were not given carte blanch to use it on our brand. We had to test it to ensure it generated the desired outcome for our specific brand’s sales and market share. The impact of any actions is both brand and context specific.
Return to the Past
While we hypothesized that sampling would work better than couponing in generating trial and repeat purchasing, we learned that it wasn’t effective in achieving our Year I sales and market share goals. Yes, it certainly generated higher levels of trial, but the purchasing and repeat purchasing levels were not sustainable without continued high levels of support—particularly with highly aggressive competitors. We simply could not afford the frequency of support needed to acclimate consumers to the taste of Vacuum Folger’s Coffee, given the inordinately high cost of sampling. Accordingly, we switched from sampling to couponing to promote the brand when rolling-out into new markets.
The hallmark of genuine marketing excellence is evidenced-based marketing. It establishes the causal relationship of actions to line-of-sight sales. Evidence-based marketing is not a sometimes thing, but a every time practice.
What testing plan do you have in your marketing plan? If you don’t have one, you are practicing eminence-based marketing. You are not investing your marketing funding. You are gambling it away.
Achieve genuine marketing excellence to make your marketing matter more. Create high-impact advertising. Learn more here: http://bdn-intl.com/avoiding-critical-marketing-errors
Stay safe and be well,
Richard Czerniawski and Mike Maloney