BrainTrust Query: How should supermarkets manage the ‘long tail’ of in-store media?
By Herb Sorensen, Global
Scientific Director, Consumer and Shopper Insights, TNS North America
Supermarkets typically
have 30,000 to 40,000 distinct items on their shelves, of which less than
five percent contribute more than half the store’s sales. In fact,
the typical household only buys about 400 distinct items in an entire
year, many of those purchased over and over, month after month. The
aggregate of all these frequently purchased items constitutes the “big
head” of products in the store. The other 95 percent constitute the
“long tail,” that sells very little on a per item basis, but taken
all together, provides substantial sales.
This concept can be
applied to the various media in the store, too. Taking media to mean
whatever is used to communicate (mediate) from the merchant to the shopper,
the packaging of the products in the store is a major component of that
communication, although advertisers typically only think of media as flyers,
signs, digital displays, shelf-talkers, etc.
Turning from the media
to the shoppers themselves, there are a quadrillion “media” exposures
annually in stores around the world, of both the package label and other
signage/displays. This is based on mind/eye biology, and reflects
that the shopper is always seeing something. The eye sees
communication coming in from all sides – received more as colors, shapes
and iconic images, than as textual messages. Quantifying and analyzing
media consumption is quite complex, but not beyond the reach of scientific
metrics.Â
TNS, for example, uses
a discrete video camera hidden in an earpiece, called an EyeCam, to track
shoppers’ exposure to media. Shoppers wear the non-obtrusive device throughout
their normal shopping trip thus creating an accurate picture of exactly
what appears in their field of vision. Technicians study these videos,
noting how long designated media appears in the central portion of the
field of vision, the area most likely to impact the shopper. Introducing
time as a factor allows a workable GRP (gross rating points) equivalent
to be calculated for in-store media. Not considering packaging, here
is the distribution of other media in a store:
End aisle displays,
free standing product display racks and in-store flyers (weekly circulars)
dominate the big head. For about 25 percent of the shoppers’ time
in store either an end cap or free-standing display will be in view. There
is no doubt why 40 percent of purchases across all stores come from these
secondary displays – the non-gondola items. Although only one in
five shoppers carry weekly circulars to the store, they refer to them so
frequently during the trip that they also make it into the big head ranking.Â
The rapid fall off in
exposures for all other advertising is striking. Floor, shelf and
freezer door ads, display bins, couponing and signage all make up a portion
of the long tail media. Even though 80 percent of the shopper’s field
of vision in center-of-store aisles is packaging, when exposures are distributed
over those tens of thousands of items, each individual SKU (item) receives
very little exposure. The long tail is very long, indeed!
Therefore, retailers
need to keep in mind that all media is not created equal in terms of its
reach and effectiveness. The prices vendors pay for the medium should reflect
on its location in the big head or long tail. It is essential to match
the cost with the benefit, meaning effectively placed targeted long tail
media can be just as effective in terms of retail ROI as the big head.
Which method is right for you? Weigh the differences and make both the
big head and long tail mediums work together for your benefit.
Discussion Questions:
How would you suggest retailers apply TNS’ long tail theory of supermarket
in-store media? Should it simply a matter of investing more in the “big
head” media, or are there extenuating circumstances you can think
of?
Reproduced
with the permission of the Journal of Advertising Research
© Journal
of Advertising Research, Volume 48, Issue 3, pp. 329-38
I think this proves why retailers and manufacturers both need to be thinking holistically about in-store advertising, and not just cherry-picking the “big head” stuff. It’s not just about individual media channels, it’s also about the cumulative impact–a consistent message across multiple channels in a store is a stronger message than a disjointed one.
But it’s also about recency. Just because a floor graphic doesn’t get as much notice doesn’t mean it has less impact. If a floor graphic for, say a cake mix, catches a consumer right when they’re thinking about dessert options, then its impact will be far greater than when they are thinking about, for example, produce. You need to take that into account as well, when pricing in-store media options. It’s not just real estate–it’s context.
My expertise is as a long-ago stock clerk and current shopper. I buy from end caps IF they are sale items. It’s a good way to ensure adequate stock of sales items and a way to quickly shop the sales circulars. It works for the customer, but it might keep me from going down the paper aisle for the sale items where I might buy something else. So it might not work best for the grocer.
I also like the concept of stocking the ketchup next to the buns as well as on the condiments aisle. How would you get that extra space? Demand that suppliers provide a six pack case instead of the standard 12 so the shelf is not filled with three rows of long tail items. Before you were born, I always wondered why we had 3 rows of capers when we only sold one jar a month.
The attempts to influence shoppers in store has been a never-ending search for the “silver bullet.”
“The last 500 things haven’t worked too well, but the next 500 fast-talking salesmen that show up in the lobby just may have the next best silver bullet.”
The customer’s mind, while shopping, is working on a different set of principles and needs than searching out new products and information. If retailers could actually walk in their customer’s shoes, they would stop wasting money on gimmicks and telling the “What” and give the customer more of the reasons “Why”.
Stores fail most at delivering good old customer benefits and selling points when needed to improve their performance in some way. Complementing a product launch with in-store POP may have some reason for existing as a reminder.
Until the customer stops coming to the store thinking about the million and one things she has to do… get something for dinner, what the kids need tomorrow, pick the kids up from soccer, dance, basketball, jazz or tomorrow’s presentation…retailers should invest in getting her in and out of the store faster and easier.
This is an interesting piece of analysis and provocative on its face. It raises several questions in my mind that I believe deserve to be aired.
First, is a “long tail” distribution a valid way of looking at a list of just 14 types of communications vehicles? True, the gross shape of the curve is coincidentally similar to the product sales curve used to compare performance of 30,000 in-store SKUs. But here we are looking at the rate of traffic passing a brief menu of promotion installations. The differential between the most frequently viewed and the least frequently viewed is about 1-2 orders of magnitude. In contrast, the differential between the rate of sales of top-200 SKUs versus the slowest sellers is more like 4-5 orders of magnitude.
My second concern is more fundamental and goes to the assumption that shopper media channels should be measured in GRP equivalents in the first place. I’ve gone on about this topic in this forum in the past, so I won’t give the whole long lecture here. Suffice to say, shopper media deliver a blend of opportunities to see, confirmed views, actions taken and feelings felt. Rating points may seem convenient to media buyers steeped in the TV mindset, but the benefits may be better expressed in terms of profitable sales lift, not limited to audience metrics.
On the positive side, there are many reasons to agree that end aisle displays, free-standing displays and circulars are highly effective at driving sales lifts. But so are digital media at the shelf (or within the end caps); floor media; coupon machines; and other vehicles listed in this chart. Individually they may not be seen as often because they are not located in main traffic aisles, but their positioning close to the final product decision point is still highly consequential.
I’m not sure what can be done to generate sales of items that historically don’t respond to causal stimuli. Each brand has its own factors that need to be considered when creating an expectation of incremental sales. I think you will find that most of these items fall into that category of convenience for the shopper and strengthen a retailers position as a provider of variety to their consumers.
I completely agree with Nikki regarding the need for context when weighing long tail vs. short tail metrics. A closer look at long tail movie renters (those who routinely rented off-beat/niche movies), for example, revealed that they just have bigger appetites. Far from having exclusively fringe tastes, it turns out that they had rented all they could from the “fat head” and were forced into the long tail through their heavy consumption.
Back to the store…there is a reason why retailers are talking more about consumer “missions,” “life stages,” and “need states” these days; they’ve figured out that the context that consumers are placing around their shopping trip, the “why” vs. the “what,” makes all the difference. Unfortunately, it also continues to make the in-store marketing ROI equation more complex.
I am trying to understand the full impact of the “long tail” distribution. If the 5% “big head” contributes half of gross store revenue, how much does it contribute to net margin on sales? In other words, is it more profitable to sell long tail items than big head items? Has anyone done that curve?
If there is research about “baskets” or what items are purchased with what other items, that could be very useful. Then, either groups of items could be promoted or when there is an in-aisle promotion, the complementary items can be placed in the aisle. Maybe the complementary items should always be grouped together. Thinking about “items” is a mistake. Think about usage occasions or meals or events or analyze what consumers purchase together, then promote or merchandise the “bundles.”
As the economy enters an apparent recession, manufacturers of all types and size will be analyzing and looking more closely at how their retail promotion dollars are being invested. Yes, ideally, the prices vendors pay for the medium should reflect on its location in the big head or long tail to match the cost with the benefit. However, to anticipate that retailers will absorb the time and the pain to lower the cost to suppliers for the less effective media, well, it’s not going to happen.
That said, as a consultant to both retailers and suppliers alike, I strongly recommend that both parties make certain that each understand how, where, and to what extent the product will be exposed, and to agree on this before contracts are signed and money is committed.
The same theme is being communicated in all the above comments. The fact is suppliers and retailers need to work together on all strategies that touch the consumer whether in-store or otherwise. As a consultant this is our primary objective.
However, to me to look at this as clean and dry as a big head and long tail misses the target. Just the same as managing assortments, media/promotional strategies have to be looked at from the perspective of segmentation. What is the role of the media/promotional strategy? This needs to align with what the role is for the category and the product segment.
It is not just about sales either. Profit productivity needs to be a factor as well including inventory management. I like to look at this from four quadrants/roles based across two parameters–sales and GMROI. Dependent on where the analytics fall out then the strategies come online.
What we find is when the strategies are developed from this perspective, both suppliers and retailers are more conducive.
I agree with James Tenser here – I don’t think this is a good example of a Long Tail profile. The items in the “big head” are where the bulk of products — presumably the reason one goes to the store in the first place — are located. The remainder are sales and promotional tools. So of course endcaps and aisles are going to see much more store traffic and get much more attention. People go to the store to buy stuff, not to be advertised to.
Excluding the two items items at the head of the curve illustrates the real problem with marketing at-retail: there is no one, single big winner. A number of ad formats and marketing strategies compete for the shopper’s attention, but depending on the store profile, shopper psycho/demographics, etc., any one could wind up performing better than the others.
The paper being discussed here is a part of my longer term goal of monetizing every square foot of the store, and quantitating all of the activity in it. In this sense, the particular distribution doesn’t matter, it is a matter of paying for what you are getting; and getting what you pay for. There is a great deal more long tail than big head, so properly monetized, there is no reason the long tail shouldn’t carry its own weight economically in the store.
But rather than focus on the monetizing, I would prefer that everyone recognize that shoppers vote every day, in every store, first with where their eyes look, and secondly with the dollars they leave at the checkout. Clearly, retailers and their brand suppliers tend to stir the big head and long tail together, in the hopes of generating action in the long tail. This SERIOUSLY suppresses big head sales by denying shoppers reasonable access to what they really want to buy. My thoughts on the long tail of media are inextricably linked to the management of the long tail of products. BOTH need to be distinctly and appropriately managed.
James Tenser’s observation on the greater orders of magnitude in the differential on the product distribution and the media distribution is probably a direct consequence of the additional orders of magnitude of products compared to distribution. I hope I understood his first point correctly, and that this addresses it.
I share his aversion to GRP media metrics, and prefer absolute exposures as computed by the formula specified in the PDF of the article. The most important value of expressing this in GRPs is to create a bridge to the media world from the serious science of shopping world. The article does comment on the solid relation of the two. I have commented elsewhere on the difference between “opportunity” metrics and “direct” metrics. The fact that today we can directly measure exposures means that failing to do that is retrograde.
James Tenser also mentions a variety of media, including “digital media at the shelf (or within the end caps)”. We have observed that embedded media does seem to be particularly effective. When we are talking about media metrics, per se, we are talking about the Reach of the media. The digital and other media delineated, after Reach, must exhibit both Stopping Power and Closing Power, which ultimately drive sales conversions. But if you are a retailer selling media exposures, your obligation is to deliver measured exposures (Reach). Responsibility for Stopping and Closing resides with the media company, as an extension of the Brand Supplier. Here I am simply making the science of the sale reach the business reality of the retailer choosing to sell exposures to their measurable audience.
‘woolah’ obviously has a good grasp of the mechanics and popular theory of endcaps. However, Glenn Terbeek showed that half the people buying from an endcap were unaware that this was promotional pricing, and half of those that were aware, didn’t care. The abysmal economics have been documented by the Ehrenburg-Bass Institute. There are solid reasons for brands to buy endcap exposure – paying shoppers to buy from them is typically NOT justified.
Whether we would agree on all the rationale, ‘vesper” is absolutely right that “retailers should invest in getting her in and out of the store faster and easier.”
In response to Bill Gerba, “there is no one, single big winner,” the reality is that 40% of ALL sales in supermarkets comes off endcaps and other promotional displays, rather than the main-aisle displays. But for proper perspective on this, we need to recognize that an endcap may have 10 SKUs on it, while the aisle beside it has 1000+ items. This means that the endcap COULD BE the big head of product sales, simply by virtue of its placement – not the pricing. Almost certainly, it would be easier to increase sales by 20% on an endcap by embedding digital media, than increasing the entire aisle by 2%. The significance of this is larger than can be discussed here.