Promotions pour vast amounts of money down the drain. How can they be made more efficient? Continual evaluation is one answer, as Helen Feary of IRI reports More than 60% of some grocery lines are now sold on promotion, with millions of pounds spent annually by fmcg retailers and manufacturers on special offers and price discounts. The Institute of Grocery Distribution estimates that in 1997, retail promotions for food alone cost £7 billion. The sum spent on promotions is often second only to the cost of goods on the balance sheet, and for many companies may be greater than the total profit made by the company. At a conservative estimate, there are 35,000 separate retail based promotions per year, with hundreds of separate promotional events within a product category. This can result in ridiculous situations. John Bowman, general manager, Coca-Cola, said: "I found more than 40 promotions in soft drinks in one store visit, but on average consumers spend only two minutes at the shop fixture." With only three seconds to evaluate each promotion on offer, most consumers probably wouldn't even try, and so all but the most alluring promotions would fail. With all the time and money put into promotions, it is staggering that a macro IRI study conducted in the USA proved 80% of promotions did not break even. Work in the UK suggests the same is true here. Clearly, the promotional process is crying out for objective and scientific evaluation, and yet very little is actually carried out. If questioned, most marketers would admit they don't know whether the promotions they run actually achieve their objectives, let alone whether they are the most efficient method of doing so. However, as ECR principles penetrate into most major fmcg organisations, there is a new focus on making promotions work better. That is, ensuring the right mechanics are selected to optimise consumer response; that the sales generated are indeed incremental and do not just compromise future sales; and that expected sales can be forecast accurately, enabling supply chain efficiencies to be achieved. Historically, fmcg firms have analysed their promotions on a sporadic ad hoc basis. But continual evaluation is now readily available, and many manufacturers are starting to invest in ongoing tracking of promotional activity alongside their regular IRI or Nielsen sales tracking services. Promotional tracking records the presence of different promotional activities such as a price cut on its own, in-store materials, or secondary displays such as gondola ends, and places these in context with the sales achieved. From the underlying level of sales before and after the promotion, a base line' of sales that would have been expected during this period without the promotion can be calculated, taking into account any seasonality or market trends, and the sales directly attributable to the promotion identified. The use of continuous techniques within the fmcg community is currently far from universal. But manufacturers such as Lever and Spillers Petfoods have found that when they start using the database to analyse certain deals, they quickly move on to comparing the results from different promotions over time, monitoring competitive activities in comparison to their own, and discovering the underlying trends in their business. For these companies, access to continuous promotional data has proved valuable in developing and evaluating their own below-the-line strategies. If the response of the consumer to different promotional mechanics is understood, then the most appropriate promotion can be selected for each product, every time. Continuous evaluation techniques allow the impact of different activities to be isolated and the relative merits to be calculated. Charts 1 and 2 show the response of Dairy Crest's Frijj, the brand leader in fresh flavoured milk drinks, to various promotional activities throughout 1997 and the uplift achieved through different promotional techniques. Incremental sales are achieved by temporary price reductions alone (+32%) and by highlighting these with in-store materials (+104%). But by far the biggest uplift in sales is achieved when a secondary display is used (+258%). These figures are typical for a product that is less frequently purchased than a known value item such as margarine. A known value item would be expected to show a stronger response to price reduction alone ­ perhaps to the extent where it would not be worth the cost of a secondary site. Of course, whenever a promotion is implemented, a quantity of product which would have been sold to the product's regular consumers, even without the promotion, is sold at a discount. In order to calculate the true cost of a promotion, the cost of this unnecessary promotion must be taken into account and added to the more obvious costs of PoS materials, retailer support, and so on. Choosing the most appropriate promotion type also means getting it right from the retailer's point of view. In the past, the impact of promotions on the category as a whole may not have always been top of the mind, but as retailers move towards category management, they will require more emphasis on the category. At a recent IRI seminar, Andrew Glover, Sainsbury senior manager, marketing planning, emphasised the need for suppliers to recognise the role of promotions within the retailer and to focus on category building activities, particularly within destination or core categories, which are where the retailer will appreciate support towards his objectives. However appropriate for the product and the category, a promotion can only be effective if it is actually implemented as planned. Needless to say, this often doesn't happen. Compliance can be affected by many variables: special packs or PoS materials that don't arrive on time, delayed communication to store management, local activity in one area, or a whole variety of reasons. The availability of named account volumetric data has made store compliance much easier to identify and to analyse. Chart 3 ­ a theoretical example ­ shows the level of promotional support given to a certain product across several major multiple grocers between February 1997 and January 1998. The chart clearly shows a high level of promotional activity in the November to January period, and in this instance revealed Safeway did not participate. This information is very easily extracted from named account data. Using a delivery system such as DataServer Analyzer, the process only takes about 20 minutes. One of the most far-reaching consequences of the huge number of promotions now taking place is their impact on the supply chain. A recent Coopers & Lybrand study commissioned by ECR Europe estimated that inefficiencies within the supply chain for promoted food goods in Europe cost $33 billion. This is not surprising considering how difficult it has been historically to forecast promotional stock requirements. Until continuous sales and continuous promotion tracking became available, vast under or over-estimates of volume requirements were commonplace. These either led to out-of-stocks, lost sales and frustrated consumers, or to warehouses packed with promotional stock which was never needed and ultimately had to be repacked or destroyed ­ incurring further costs. The latest techniques have greatly improved the ability to accurately forecast promotional demand, either by continuous promotional monitoring or through more in-depth analyses of store-by-store named account data which can isolate the impact of even the smallest door drop or local promotion. The potential savings in the supply chain are probably the single most important factor which will change the way retailers and manufacturers treat promotions in the future.This is already happening within manufacturers such as Kellogg which is in the process of consolidating its information technology infrastructure worldwide to make its global operations, and specifically its supply chain, more efficient. This company-wide approach treats promotions as part of the process of bringing products to market as efficiently as possible, and makes improved promotional evaluation and forecasting techniques essential. Predictions of promotional performance can now be highly accurate thanks to advanced statistical modelling techniques. The results for forward thinking manufacturers like Kellogg should be considerable cost savings for the company and also increased retailer confidence as a result of its ability to create the right promotions to fulfil both brand and retailer objectives. n {{FEATURES }}