by Emily Chan
Data
Expect significant data collection, analysis, and iteration. To determine needs, characteristics, competitive landscape, and profitability, you will need large volumes of data and substantial analysis. Some of the tools are similar to the ones described in Part III. Others are more specific to marketing, such as customer questionnaires, focus groups, and conjoint analysis.3 HBS does not go into technical details of these mechanisms. In my experience, it is usually more time- and cost-effective to hire market research companies to apply these tools, as they have access to teams of part-time questionnaire conductors, large databases of potential focus group participants, experienced focus group leaders and data analysis experts, and so on. But market research companies should be seen as the arms and legs, not the brains of the project. The choice of data to gather and possible segmentation of the customers should be led by company personnel or by consultants who know the company, its business, its strategy, and its operations well.
To conclude, segmentation may be easy in theory but it tends to be tough in practice. There are many ways to segment a market. Significant resources and experience are often needed to identify an effective segmentation. It takes strategic and creative thinking. It also takes a lot of patience as iteration and rework (such as another round of questionnaires or focus groups) may be necessary.
3M: New Brand
In FedEx’s case, both premium and standard services fit well with the FedEx brand. But sometimes a company may need to introduce a new brand to fight the war. This is especially true if the original brand has a premium image. An example is 3M. Back in the early 1990s, competitor Kao Corporation introduced a low-priced diskette. Despite the need to answer, 3M was reluctant to drop the prices on the 3M brand as it knew very well that a customer segment was perfectly willing to pay the higher price for the quality that 3M represented. As a result, 3M introduced a new brand called Highland to fight Kao.
Consumer Products: Kill the Competition Before It Kills You
The Harvard Business Review article “How to Fight a Price War” describes a very impressive episode.4 The authors did not name the company, and though I have heard of the same case from various business school friends, no one has been able to tell me the real identity of the company or the product.
As the story goes, a major consumer products company was faced with an aggressive price-cutting competitor. The company defended itself by dropping the price of its economy-size product with a “buy one, get one free” offer. Because each unit of the economy-size product would last six months, the high-volume, price-sensitive customers that stocked up with the offer were off the market for almost a year. The competitor was put in a very difficult situation. It was not possible for the competitor to respond quickly given the stock levels, logistics, and so on. Even if it managed to respond with a similar offer, its margins would be severely affected. It also would not change the fact that the high-volume, price-sensitive customers would stock up and be off the market for a long time, hence severely affecting sales in the following months. This aggressive counter convinced the competitor to withdraw from the price-cutting adventure. In some similar situations for other markets, smaller competitors could be totally driven out of business.
Intel: Strategic Retreat
The four cases discussed thus far describe some of the ways to fight a price war. However, just as in real war, sometimes there is just no way to win a price war. It is lose-lose no matter how the war is fought. This is especially true when a product becomes generic with very limited differentiation. In such cases, some companies will choose to withdraw from the market altogether. Naturally, this is only possible if the company has other product lines or is able to constantly innovate. A key example is Intel. Intel stopped making DRAM chips and focused on other products when price pressure from Asian manufacturers on DRAM intensified. Another example is 3M, which exited from the videotape business when videotapes became a generic product offered by many low-priced manufacturers.
PROMOTION: WHERE IS THE MONEY?
Promotion includes above-the-line and below-the-line marketing. Above-the-line marketing uses mass media such as television, radio, newspapers, Internet banners, and the like. Below-the-line usually focuses on more direct means, such as direct paper mail or e-mail, flyers, coupons, gift with purchase, and lucky draws.
I am a naturally cost-conscious person. Hence, though I believe promotion is an essential part of marketing, I want to be sure that the money is spent wisely. When examining effectiveness of promotional spending, my education and experience tell me to adhere to three fundamental principles:
Brands cannot live on advertising alone
Find the “monopoly window”
Understand promotion profitability
Brands Cannot Live on Advertising Alone
Advertising is important. It is the fastest way to reach a large audience. But advertising is not omnipotent. It is expensive and risky to rely on advertising alone to build a brand. As Professor Clayton M. Christensen of HBS and his coauthors clearly stated in the Harvard Business Review article “Marketing Malpractice”: “advertising alone cannot build a brand, but it can tell people about an existing branded product’s ability to do a job well (fulfill a need).”5 Here are two examples to help illustrate this point:
Unilever Soupy Snax—4:00. Through market research, Unilever discovered a need in the office worker segment for an afternoon snack. Many office workers who got tired around 4 PM were using caffeinated beverages, junk food, or short breaks to reenergize. So Unilever introduced Soupy Snax—4:00, a nutritious soup that was easy to heat up in the office pantry. To market this product, Unilever launched an advertisement featuring drained office workers perking up after drinking Soupy Snax—4:00.6
Miller Lite Catfight. As discussed in the section on segmentation, through market research, Miller Lite identified the segment of young male that would enjoy advertising featuring mud-wrestling supermodels. Hence, they launched the “Catfight” advertising campaign.
The Soupy Snax campaign was a major success resulting in high sales. The “Catfight” got a lot of attention from media and consumers but failed to increase sales. One key difference: in the first one, advertising was used to highlight an unsatisfied need and how the product could fulfill that need. The second one was creative and entertaining but did nothing to convince consumers to switch.
I do not have the research data to say that no brand can succeed on advertising alone, and it’s hard to prove a negative anyway. Probably some have. But being of a cost-conscious turn of mind, I would prefer to maximize the chance of success by ensuring that advertising is used in a supporting role, not the lead role.
Find the “Monopoly Window”
The U.S. motor industry has always been highly competitive, with a handful of major players. The industry provides a lot of interesting case examples in many areas, including below-the-line promotions. Two well-known ones:
In 2005, GM implemented a high-profile “you pay what we (employees) pay” promotion. In this promotion, customers were offered the “staff price,” which was a deep discount from the regular retail price.
In the same year, the Pontiac division of GM sponsored an episode of the TV reality show The Apprentice where competing teams were given the task of producing a brochure for the limited edition 2006 Pontiac Solstice. The episode showed how the teams designed the brochure and, of course, the distinguishing features of Solstice were highlighted. Viewers were pointed to Yahoo for additional information and a “chance” to order a unit of the limited edition.
The results of the two promotions were vastly different. For the employee-discount offer, competitors retaliated with major price cuts. The resulting bloodbath cost GM and its competitors billions of dollars. Meanwhile, the TV show was a major success. The target was to sell 1,000 cars in 10 days. All 1,000 were sold in less than an hour.
The key differences between these two efforts highlight a number of best practices in design
ing promotions:
Avoid promotions that just mask price reductions. Saying “you pay what we pay” is a price reduction. It is not even well-disguised. Some other schemes such as double loyalty points are price reductions in a slightly better disguise. However, competitors and customers both see through such ploys quickly. This can result in tit-for-tat reductions from competitors, that is, a price war—which is lose-lose. Such tactics can also inspire customers who would otherwise have paid full price to load up on extra inventory for future use (called “forward buying”) so thatnot only do they not pay full price now, they don’t pay it for a long time to come either.
Invest creative and organizational resources to maximize the “monopoly window.” This is a term originated in the MIT Sloan Management Review article “A Strategic Perspective on Sales Promotions.”7 In the article, Professor Betsy Gelb and coauthors Demetra Andrews and Son K. Lam defined “monopoly window” as the time between the point where customers start making purchases driven by the promotion and the time competitors launch their defenses. Figure 8.1 illustrates the concept.
In other words, the period between the time customers start to respond to the promotion and competitors can react will be the window where the company’s promotion can “monopolize” the market with the offer available to the customers. The bigger the monopoly window, the greater the chance of the promotion being a success. The window can be maximized by speeding up customer response or delaying competitors’ response. In the case of the Pontiac Solstice, customer response was sped up by leveraging a popular TV program, immediate Internet access for information, and the “ticking clock” of a limited edition. Competitors simply could not respond—it would take a new product ready to launch and a lot of time to figure out a countermove.
Figure 8.1 The Monopoly Window
Source: Betsy Gelb, Demetra Andrews, and Son K. Lam, “A Strategic Perspective on Sales Promotions,” MIT Sloan Management Review 48, no. 4 (Summer 2007): 3.
A long monopoly window does not happen by chance. It takes ingenuity, creativity, competitive savvy, marketing expertise, discipline, stealth negotiations, and well-coordinated execution.
Understand Promotion Profitability
It is surprising how many companies do not rigorously analyze the profitability of their promotions. I once had a liquor client that measured promotion profitability by the total number of cases sold during the promotion. The fact that each case was sold at a discount price and that a lot of the sales reflected forward buying (customers stocking up for future use rather than really increasing consumption) was ignored. And I worked for a property investment firm that owned a high-end shopping mall. The mall had a very active promotion calendar with promotions such as free parking, gift with purchase, and other maneuvers. But the profitability of all these promotions was never measured because it was, according to the head of sales and marketing, “too difficult to quantify.”
It is critical to measure profitability and to measure it correctly. Between 1982 and 1990, Professor Leonard M. Lodish and Magid M. Abraham studied promotions of all brands in 65 product categories in the United States and found that only 16 percent were profitable!8 I helped my liquor and shopping mall clients make some rough assessments of their promotion profitability. The analysis found that the former was losing money from many of its promotions and more than half of the promotions done by the latter were unprofitable.
The key in measuring promotion profitability is to focus on incremental profit. Incremental profit is what the company makes above and beyond what would have been made without the promotion. There are two categories of profit that would have been made without promotion:
Baseline profit: Profits from purchases that customers would have made at full price without any promotion. With the promotion, this same volume is purchased at a discount price.
Forward buying profit: Profits from purchases the customers would have made at full price in a later period without the promotion. With the promotion, customers buy the same volume at a discount price earlier and stock up for future use.
The reduction from what that would have been made if baseline and forward buying purchases had been at full price must be taken into account when assessing promotion profitability.
Table 8.1 is an example of a framework to estimate incremental profit using disguised data from my liquor client, which sells to retailers. Assume the liquor costs $40 a case to produce (cost of goods), sells at $100 a case without promotion, and is promoted at a 10 percent discount for a month. Assume $5,000 for the cost of planning and execution of the promotion, including advertising and printing of promotion materials, human resources, allocated overhead, and other details.
This framework can be used to assess a promotion after it is finished, and it can be applied in the planning stage with “target volume to be sold” replacing actual “volume sold” during the promotion period. But whether it is for post-promotion assessment or for planning, the challenge is in estimating baseline sales and forward buying. There are two main ways to estimate these figures: you can build a systematic approach for it, or you can take your best estimate based on the data you have.
Table 8.1 Estimating Incremental Profit
1. Total volume sold during promotion period 2,000 cases
2. Estimated baseline volume 900 cases
3. Estimated forward buying volume 300 cases
4. Estimated incremental volume 2,000 – 900–300 = 800 cases
5. Revenues from incremental volume 800 cases × $90 = $72,000
6. Manufacturing cost (cost of goods sold) of incremental volume 800 cases × $40 = $32,000
7. Loss of revenues due to discount (900 + 300) cases × $10 = $12,000
8. Other costs of promotion $5,000
9. Total cost of the promotion $32,000 + $12,000 + $5,000 = $49,000
10. Profit from the promotion $72,000 – $49,000 = $23,000
System
The most accurate way of course is to develop and apply statistical analytical algorithms. This is possible if you have accurate and timely sales data such as supermarket scanning data for consumer goods. In the study mentioned earlier, Lodish and Abraham estimated baseline sales of the 65 product categories using scanning data from supermarkets that showed purchases by consumers before, during, and after the promotion period. A detailed algorithm was developed to project baseline sales for the promotion period itself. However, even with this amount of data and analytical power, the system could not estimate the consumers’ forward buying.
Best Estimate
As discussed earlier, hard data is often unavailable. Also, systems take a lot of time and resources to build and are often imperfect even if built. In such cases, best-effort estimates should be made based on historical data, market trends, sales force field input, and management judgment. In the liquor client example, we identified some smaller trade customers who were not forward buying (because they had policies against it or couldn’t afford it). Trends of orders from these clients helped to estimate the baseline. For those that were forward buying, salespeople held discussions with their management to understand volume and pattern of forward buying. Sales to these clients before and after the promotion were also studied to estimate the volume of forward buying. In the end, the team of marketing and sales personnel agreed on a best estimate based on information collected. This process was difficult at first, but it became more structured and easier once people got familiar with the process. Since the numbers produced by such a process are estimates, they should be used with caution.
To help provide estimates, some mechanisms can also be implemented, either permanently or as part of each promotion. While most of these mechanisms do not provide enough accurate and detailed data for systemtype analysis, they do provide invaluable data to help with getting the best estimates.
The property investment client started to collect turnover rent, which requires retail tenants to provide them with each store’s total monthly sales dollars (please
see Appendix A for detailed explanation). Although not as timely and detailed as actual scanning data, monthly data are very useful in helping to analyze the effectiveness of promotions.
The shopping mall client also started a discipline of conducting shopper questionnaires during each promotion period with such questions as “Why did you shop here today?” and “Do you find this promotion attractive?” These questionnaires were simple (with three to five questions) and conducted randomly at the mall or at gift redemption desks.
The liquor client started including lucky draw tickets or gift redemption forms whereby it could capture consumer purchase data when the consumers sent in their tickets or forms.
If data is simply not sufficient to project baseline or forward buying figures, you can calculate them by a shortcut. The formula I use is:
G´ = Gross margin $ per unit after discount (in other words, discounted price minus cost of goods sold)
V´ = Total volume sold during promotion (including incremental, baseline and, forward buying)