BrainTrust Query: Is it possible, and advisable, to establish a set of industry metrics to gauge in-store media effectiveness?
By Laura Davis-Taylor, Founder and Principal, Retail Media Consulting
While participating in Ad Week in NYC last week, I attended the POPAI’s At-Retail Media conference. It was a fantastic event focused on the growing emphasis on marketing at retail.
POPAI believes Retail Media is simply a reflection of the retail environment as it exists today. They state that this is comprised of:
- Traditional CPG product merchandising
- CPG direct product sales
- Retailers who are branding and customizing their stores and their chains to differentiate themselves from their competitors.
Dick Blatt, POPAI’s global president, dove passionately into subjects such as: finding research and valuation metrics for store media efforts; looking at technologies that are
currently in our toolkit (as well as on the horizon); and how to bring it all together for positive shopper engagement.
The American Association of Ad Agencies also quietly sponsored a Retail Media event this week, further verifying that they are embracing the channel as a ripe opportunity for
reaching consumers.
Where is all of this going? In listening to brilliant POV’s on Retail Media, it is clear that most of the constituents feel that “media” efforts in the store need to be planned,
purchased and measured via the very same processes and metrics as all other media. This is a world of GRP’s – or reach and frequency – and is based on capturing “eyeballs.” This,
many feel, is the only way that the advertising industry will transfer significant media funds from traditional broadcast TV to new store opportunities.
Last Wednesday, at the In-Store Marketing Expo in Chicago, the In-Store Metrics Consortium unveiled the Pioneering Research for an In-Store Metric (or P.R.I.S.M.) model. Their
goal is to deliver “a common language to determine consumer reach in-store” with a type of Nielson rating for store media. Proctor & Gamble is the primary sponsor, with 3M,
Coca-Cola, Walt Disney, Kellogg’s, Miller Brewing, Alberstons, Kroger, Walgreens and Wal-Mart also involved. Rather than traditional GRP’s rooted in reach and frequency, the P.R.I.S.M.
GRP’s are calculated for specific locations within the store and are based on traffic x compliance x unduplicated impressions.
Peter Hoyt of the In-Store Marketing Institute feels that, via P.R.I.S.M.’s metrics, “retailers will determine what advertising vehicles and communications best meet their objectives,
and potential advertisers will be able to evaluate those opportunities accurately.” He is also careful to note that all involved must evaluate which specific strategies and objectives
are most effective at improving the shopping experience.
Discussion Questions: Should we create a standard GRP valuation for store media touchpoints? Do you feel that metrics will convince traditional media
agencies to pour “advertising dollars” into the store, seeing the environment as a rich opportunity for brand messaging?
Kudos to P.R.I.S.M. for a fabulous effort that is very exciting for our industry. We’ve been struggling with the valuation of the store as a marketing vehicle
for far too long. This is not surprising, as we still struggle with valuating traditional broadcast, TV and print. In fact, the only media we don’t have this struggle with are
direct and interactive marketing, where accountability is more trackable.
If we want to capture a higher percentage of media dollars from large agencies, it makes all the sense in the world to model “retail media” in their nomenclature
and buying processes. However, we need to think hard about how much control the agencies will ultimately have over “buying media” in-store that’s valuated by “eyeballs.” Retailers
have tight control over their store real estate and are increasingly focused on cross-channel merchandising promotions that enhance the consumer shopper experience. Most often,
these promotions are valuated and negotiated directly between the merchant and the brand manufacturer.
Another point is that we must ask ourselves if consumers even want in-store advertising to “capture” them. Over and over again, we hear that due to the
aggressive use of TiVo, ad blockers and consumer control devices, the store is the one place remaining where strong consumer reach is assured. Call me crazy, but why would we
try to replicate the same old “look at me” advertising models in-store that people are rejecting in other channels?
Guy Vaughan from Retail Marketing Services was clear at the At-Retail Media conference that camps like the In-Store Metrics Consortium are aiming to create
audience metrics first, then standards for ROI and message effectiveness. Understanding that this is the long term goal is key, as we see that their end game for valuation is
rooted in solid ROI.
It’s very, very important that we all keep this in mind as retail media gains popularity. Traditional advertising is struggling with their desire to hold
onto the “value of the eyeball” and we can’t risk hurting in-store consumer relationships. The agencies have fantastic value to bring to retail media, but not if they come to
it with old models that are already suffering.
I agree with Robert Leppan; I don’t foresee traditional advertising agencies re-channeling ad dollars to retail even if a comparison of GRPs proves it to be more viable media alternative. From my experience, traditional ad dollars and in-store advertising dollars typically stem from different budget sources. More importantly, agencies would have a great deal of catching up to do to speak coherently regarding the in-store environment.
Where this new PRISM research will prove to be of immediate use is among manufactures who have forever struggled to justify the expense of in-store advertising often required of them by retailers. This new metric system will afford manufacturers the opportunity to assess, determine and pursue the most proven in-store strategy for their product. It will also provide manufacturers with a more objective approach in positioning their products within a given department.
There is ample reason to be excited about the PRISM research and other parallel efforts of the In-store Metrics Consortium; but still much remains to be considered. Previous comments that the customer must be considered in this equation are directly on-target. Although compliancy is a key part of the new metric system, the method for which it will be assessed is still undetermined. That, in itself, could be an entirely separate project that would require extensive effort and research.
For more than a dozen years I have been advocating that we regard “the retail store as a communications environment for brand messages.” In a supermarket, those messages begin with 30,000 or so package labels and extend to a host of static and digital advertising signs, as well as a variety of promotional media including demos, coupon dispensers, displays and coupons. Mass merchants may be even more complex.
New efforts to find the in-store GRP equivalent are commendable, but they barely scratch the surface. In the first place, all these media are not interchangeable in their absolute effects on consumers. Second, In such a complex messaging environment, consumers let their eidetic filters take over, screening out communications they perceive as extraneous, annoying, or even stalking. The real work of understanding consumer perception of these messages and linking them to attitude and actual purchase behavior has yet to be done.
I propose that in-store communications may be classified in a pyramid. At the base are communications that have reach only – this is what the P.R.I.S.M. initiative has learned how to measure. This is a “page view” metric, to use a Web metaphor – great in number but low in individual value.
Next up the scale are communications that stimulate some kind of interaction, but not sales. This may include pressing a touch screen for further information, taking a coupon or “take-one”, trying a sample. This is a “click-through” metric, fewer in number but of greater value to marketers.
Next up the pyramid are communications that may be directly related to product trial or sale. This “purchase” metric will be more scarce, but even more valuable.
At the pinnacle are in-store communications that contribute not just to a single purchase but to enduring affective and behavioral change. We call this loyalty, and it is rarest and dearest of all.
As a marketer, I would require that all these layers be measured and modeled so that I can truly understand the ROI of my in-store communications. As a retailer hosting these messages, I would require that I get paid in accordance to the value delivered at each of these levels. As an in-store network operator, I would seek a way to justify compensation at each level as well.
It has been an exciting week, filled with promise for this industry. The Shopper Media work is just beginning.
No doubt about it. In-store POS/media should have stringent measurement standards and its great to see a heavyweight industry group (In-Store Metrics Consortium) come together to address this critical issue. How to effectively measure store media will be a challenge since there is such a a wide variety of in-store messaging; signage on displays, floor decals, murals on walls, electronic media, etc. As a packaged goods marketer I knew, kind of intuitively as well as through stats from POPAI, that messaging to shoppers at retail was a key link in the total marketing process. It’s the last chance to influence consumer behavior before they buy. (I also agree with Ryan Matthews that the value of label/package to influence consumers should not be under-estimated). But a point to think about: traditional agencies are often not in the loop with respect to store-level POS/media. This is “below the line” spending and I don’t see many ad agencies being gung ho about re-channeling ad $ into retail. Plus it’s my experience that marketers and their sales team are the ones buying and negotiating promotional packages (POS, in-store media) directly with retailers or their 3rd party representatives.
I remember being quite impressed at the inroads PRN made with their in-store network for Wal-Mart, and now other retailers, back when it was understood that such activity could not be measured. It was wish-and-a-prayer marketing at its best and that was back when the delivery system was a teenie monitor with bad sound quality (since upgraded and in a checkout aisle near you). In spite of this, vendors and retailers stepped up and gave it a whirl. This inability to quantify results plagued RFID for a time (and still does, depending who you ask), pre-programmed store music, and in-store merchandising as well yet it shouldn’t keep vendors and retailers from stepping out and embracing differentiating tools. Creating a GRP valuation system would help the pioneers solidify their case and as these things go, will most likely embolden the timid, armed with success data, to ride on coat tails with me-too efforts.
The PRISM measures might be useful to compare one in-store network against another. But the absolute effectiveness level of in-store media can only be measured over time, by testing actual messages versus actual sales. It would help if PRISM chose another acronym, since it’s likely to be confused with PRIZM, a longstanding market segment measurement owned by Claritas.
I believe that we must quantify the “advertising value” of in-store communications via closed-circuit TV, display advertising, or other creative communication efforts that might be on the horizon. This is needed for three reasons:
1 — so marketers can make informed choices about how to spend their brand-building dollars.
2 — so that the full impact of creative merchandising efforts can be taken into account.
3 — so marketers can make the right choices about what product ideas to launch.
Along the lines of point 3, I was involved with a new product forecast where a traditional modeling service provided a very low sales estimate for a new product that would have a permanent display. When my team accounted for the traffic that would be seeing the permanent “billboards” at point of purchase and equated this to GRPs, it became clear that the traditional forecasting approach led to a sales estimate that was about one-fifth of the true sales potential.
In store advertising and brand building must be quantified in a way that is “apples to apples” with other forms of communication.
A big question here is do the retailers want their stores measured? If they did, they would be happy to provide POS data, which they are not. The amount of bodies in a store is irrelevant. What is relevant are the bodies that are making purchases. The study all over the media last week is fine, but seriously flawed — laser beams? Our company currently tracks displays in-store in a real-time environment. We can tell if a display is up, where it is in the store and we are tied into the retailers POS data to evaluate sales and profit effectiveness; we are doing what the study is trying to do.
I echo Mark Lilien’s assertion that effective media metrics need to be linked to sales performance, not just a GRP model. While standardizing metrics across all media makes a great deal of sense, we should not overlook the special nuances of the retail environment that will ultimately determine if the messages we are sending are received or even welcomed by the consumer.
I have been involved with in-store messaging through TV monitors and other digital signage for a number of years. It seems to me that “in our rush to get the party,” we have once again forgotten to invite the consumer. Many of the same issues shoppers had with in-store media when the Checkout Channel and other early renditions were on the scene, a decade ago, have largely gone unaddressed. Those include ambient sound controls, relevant messages which often include immediate gratification in terms of a price/item proposition in the store, ample content in the rotation to avoid annoying repetition of ads, et al.
Running generic “equity advertising” to an audience that already feels time-starved and overwhelmed with signs, programs, nutritional information, and the like, will crash and burn as did the early efforts in this arena. Shoppers will only take the time to “look up” if they regard the messaging as a consistent conveyance of relevant value.
While these issues may not be of immediate importance to the advertising community, they will be extremely important to the retailer who becomes the “owner” of this new medium in the consumer’s eye and will ultimately be the decision maker as to whether in-store media becomes a pervasive, exciting new medium or a limited niche opportunity.
Um…er…should there be metrics for in-store media? Well, I suppose the answer is a fairly obvious, “Duh…YES.” There ought to be clear metrics for measuring any media. Now…the tough question is, “What should those metrics look like?” Sadly, we’re not all that good at measuring traditional advertising, let alone in-store media. One problem is that the store is already awash in media (end caps, displays, packaging, PA system, etc., etc.). Not only can’t we measure any of these individually all that well, but we also have even less ability to measure them in situ, i.e., interacting with everything else. Oh, and for those who don’t believe packaging is media, just walk down the soup aisle.
Aren’t these the same guys who declared the marketing model broken and who want a new consumer centric model to help drive marketing ROI? How can anyone expect some infrared blips to be a suitable substitute for the consumer? You can’t build a new model by recreating the old. What do the blips represent? Are they men or women? Are they young or old? Are they upper income or not? Are they looking at the promotions or are they influenced to buy by the same? If the infrared blips can tell us that, then maybe they are meaningful, but I doubt it. Sounds like a lot of hype with no substance. but what else is new?
Should we? Of course! Anything that makes it easier for media planners to evaluate retail media compared to other media for possible inclusion in a brand media plan is a good thing. It represents new money.
Can we? Maybe. Retail audio, because it is broadcast throughout the store, is fairly simple. It merely requires knowing 1) How many shoppers; 2) How often; 3) How long per trip. The first two are available from several syndicated research sources including Simmons and MRI. The third is available from the Department of Labor Services.
Other media require knowing how many people visit each section of the store, making it a bit more tricky. Solutions are available using RFID or cameras, but none that are syndicated. Syndication represents a problem because it requires enough potential retail, advertiser and agency clients to be financially viable. There simply aren’t enough players yet.
However, recent developments like MediaEdge:cia’s creation of MEC:Retail and Publicis’ recent deal with Simon Malls show that the major agency players are taking retail media seriously. Overseas, many agencies are creating retail-focused shops as well. I expect to see many more of these developments in the next 6-12 months.
The metrics created in this space must initially be comparable to other media, i.e. Reach, Frequency and GRPs. These are the metrics media people require and the metrics that allowed Internet media to finally flourish in the late 90s. Once retail media is able to be evaluated on the same terms as other media, we can begin to introduce ROI, ROO and other metrics that will hopefully increase the value of our media versus traditional media that can’t be evaluated on those levels.
Don’t underestimate the agencies. They want ROI, but it simply can’t be delivered for most media. The retail environment represents a perfect laboratory in which we can measure the effects of nearly any variable.
GRP’s are loosely related to actual exposures to potential buyers. Early efforts to move this into the store environment were founded on the Opportunity-To-See concept (OTS,
www.arfsite.org/downloads/InstoreAdMeasurement2003.pdf) which essentially said that if you put media in the store, however many people are in the store can be counted as exposures, on the basis that they have an “opportunity to see” the media. Since the typical shopper sees less than 25% of the store, this gives grossly exaggerated GRP’s.
P.R.I.S.M. represents a quantum leap forward by attempting to look at the traffic in the various aisles. (And yes, converting non-buyers to buyers is what advertising is all about, so traffic IS the relevant issue to advertising and GRP’s.) However, P.R.I.S.M. is still based on opportunity to see, in this case defining opportunity as at least being present in the aisle.
For what should be obvious reasons, any one selling media is NOT going to enthusiastically (at least unilaterally) endorse ANY method that minimizes their numbers. However, just as monitoring the clickstream allows accurate, DIRECT measurement of viewership, so too can in-store shoppers be measured in terms of their actual exposure to displays. However, all this moves us beyond the opportunity concept to direct measurement of the field of view.
As significant as the advance of field of view measurement is over “opportunity,” of even greater significance to the marketer is the tracking of the shopper’s eyes, superimposed on their field of view. Although in-store eye-tracking is unlikely to ever be practical on a scale to serve as the engine for an ongoing syndicated media metric program, it is crucial from a research point of view in understanding a display’s relationship to individual fields of view. This will serve to validate and clarify the meaning of field of view metrics. Ultimately this validation will put into context:
* The actual exposure of shoppers who directly looked at a display
* The larger number of shoppers whose field of vision included the display
* The even much larger number of shoppers who were in the immediate vicinity of the display (opportunity-to-see)
* The vastly larger number of shoppers in the store, most of whom may have had a very doubtful opportunity to see.
Accurate measurement is crucial because it is the EXPOSURE that leads to an IMPRESSION on the shopper’s mind, such that there is a resulting SALE. This is the inexorable track we are on in measuring media and understanding its impact on shopping and sales.
The industry metrics that measure the effectiveness of in-store media are very different depending on whether the advertiser has products available for purchase inside the store or if the advertiser is looking for a traditional media buy.
The company that I work for is one of the pioneers in this industry. We have been producing in-store television networks for major retail chains for the past 7-8 years. Our retail clients believe that the consumer experience is most important and that requires informative and entertaining content that will be welcomed by the consumer, not more ads that the consumer will want to block out. While our retail partners are open to a limited amount of non-endemic advertising (commercial spots that do not relate to products sold in their stores), they value their airtime and want product-oriented content where they control the messaging. The measurement that matters most to them is the consumer experience, followed closely by the impact on sales and the ability to communicate with store associates. Our retail clients share their POS data with us and together we program the network to achieve these goals and validate the ROI from the network. There is a big difference between measurement of GRP points to attract traditional media dollars and measurement of actual sales increases from promoting the retailers’ products and services on the network. If the network is effective at increasing sales, it becomes a fundamental part of the retailers’ in-store marketing and the amount of airtime that will be made available for traditional non-endemic advertising will be limited. GRP points don’t mean anything to marketers with products available for sale in these retail stores unless you can get away from marketing budgets and actually tap brand advertising budgets, something that has yet to be done in this industry. As was stated in one of the earlier comments, in-store networks that are perceived by consumers as unrelated “ad dumps” will be tuned out the same way consumers avoid unwanted advertising on other mediums.
GRPs for in-store media sounds like it could be useful. But when I looked under the hood at what they actually did, I was not at all impressed.
First, the plan is to use POS sales data to infer what the traffic was in the category and the store as a whole. So they created a statistical model to do that, but the model turned out to be inaccurate R2=0.76. By the way, is Wal-Mart going to give up their data to the syndicator of this service?
They’re also just measuring walk-bys. In the cluttered retail environment, is that good enough? Aren’t a lot of people going down an aisle just trying to get from the front of the store to the back?
They haven’t explained how they get a reach and frequency measure. Supposedly they use known shopper behavior. In other words, they drew on prior research about how often a category gets shopped and made assumptions.
And, they readily admit that they have no solution for proof-of-performance. They also say they have no way to use GRP measures get to an ROI calculation.
Traffic and store journey studies have been around for a while. Combine this info with what I know about how often my category gets shopped and I could estimate GRPs with more accuracy.
Should we count in-store media impressions?…Sure. Why not?
Will this lead to an accurate and comparative measurement of effectiveness versus traditional advertising media?…Perhaps. There’s no reason we couldn’t apply the same standards for “accurate measurement of traditional media effectiveness”…as soon as we come up with those in the first place.
Will traditional agencies embrace a shift of ad dollars to in-store media once measurements are available?…Sure. As soon as retailers start an upfront auction and advertisers start paying commission on the buy.
On the surface, messaging is messaging. The store is just another location for messages to be posted. So traditional methodologies and processes can be applied — just as well as they are applied in other locations. We can measure how many people walk by a display just as well as we can measure how many people drive by a billboard. But who knows how “effective” (i.e. influenced a purchase decision) that billboard was?
The real challenge for measuring in-store effectiveness is to go BEYOND what we do with traditional media! We’re in the store. People are actually implementing their purchase decisions here. We have the holy grail of “effectiveness” (an actual purchase) available to capture. Shouldn’t we be trying to calibrate the link between in-store messaging (and other in-store marketing variables such as price) and behavior? Not just counting eyeballs?