The senior managers of large retail businesses spend a lot of time traveling to visit stores in their chains. One reason for this is that regardless of how strongly central planning dictates the communication content and layout of the store, variations in physical space and local competitive circumstances inevitably introduce differences in how each store presents its contents to the public. This variable contributes to differences in sales from one store to another. Understanding the differences is why managers want to walk each store and read it with their own eyes.

But in the age of the internet and cheap digital cameras, perhaps a better use of the manager’s time would be to tour each store through a web portal. Imagine a window where the manager could sit at his or her desk and look at a series of timely images from each store, where many different stores could be compared and contrasted stores so that differences in in-store communications could be correlated with differences in sales.

The experience of slowly clicking through the pictures in such a web portal, displayed next to a map of the store showing where each GPS-coded picture was taken, would feel very much like you’re taking a virtual tour of the store—conducting a virtual store check—only these would be real stores in the system, not some software creation.

But wouldn’t it be even better if the manager could look at each store through the eyes of customers? What if, as a form of augmented reality, each picture in the portal were tagged with important consumer information—what the consumer actually remembers looking at, how they felt about what they were seeing, what it meant to them in terms of their perceptions of brand values?

As an analytic tool, such a web portal would make it easy for managers to automatically rank different stores in the system on key metrics such as sales or customer satisfaction, or on marketing measures of customer perceptions of the store on dimensions such as Quality, Selection, Value, or an Enjoyable Store Experience.

By using the side-by-side displays in this research dashboard, it would be easy for managers to systematically see the differences between how the two stores are communicating with their customers.


A working example of such a portal is shown in figure 1. To better understand how this portal works, let’s look at the research methodology a bit more closely.

Visually-Charged Exit Interviews

The simplest and most direct way that retailers can determine what parts of a store consumers have shopped is to ask them as they are leaving the store, as part of an “exit interview.” Historically, this was done is by an interviewer with a clipboard standing by the door, who intercepts shoppers and asks a series of questions designed to help shoppers recall what parts of the store they visited and what they recall looking at.

There are three problems with this approach, however. First of all, many consumers don’t want to take the time to have their shopping trip interrupted while they do a research interview. Second, and more importantly, the verbal prompts, which researchers use in the interview to describe the different parts of the store and things that are on display, are necessarily vague and imprecise cues for consumer memory. Third, no two stores from the same chain of retail stores are likely to be exactly the same so it becomes difficult for the researcher to write a universal description of what is going on across different stores.
For retailers, the Internet makes it possible to systematically lower the cost and improve the quality and value of an exit interview—and so provides a powerful new visual-based way of measuring in-store communication.

First, it is much less of an interruption to hand out an invitation to take an online interview as someone is leaving the store than to ask them to stop and answer fifteen minutes worth of questions then and there. Second, with the cheap digital cameras that now come in almost every cell phone it is easy to walk around a store and snap pictures of the store. Third, because it is easy to do, digital pictures for each unique store can be quickly emailed to a website where they can be uploaded into an online interview. Using GPS data the exact location in the store where the picture was taken can easily be identified.

From a communications research standpoint, this “visual sampling” of a store is analogous to selecting the key frames from television commercials for an on-line ad tracking study.

Perhaps a better way to think about these photographs from a store is that they are like close-ups used the movies. Close-ups are a critical part of how films can be used to tell stories because they mimic the psychological process of how we focus our attention as we scan and search our visual field of view.

Finally, visuals are important for conducting exit interviews online because recognition-based memories, as opposed to verbally prompted recall, tend to decay at a very slow rate. A customer who was invited to take an exit interview when she walked out the door two or three days ago will still recognize many of the things she saw as she shopped the store, but will recall the experience very poorly.

This long-lasting quality of visual memory is what makes it possible to develop reliable measures of in-store communication on-line, after the shopping trip is over.

These picture-based metrics have been extensively validated for other forms of ad research. To understand how they work, from a psychological standpoint, to measure in-store communications, let’s first look at how the shopper experience is different from looking at a television screen.

Engaging Flashlight Attention

The first job of advertising is to capture the attention of the consumer. Media companies simply create an opportunity for an exposure to advertising. But sophisticated advertisers understand that this is not the same thing as engaging the mind of the consumer. Because of the psychological process known as selective perception, the link between exposure and engagement can be highly uncertain.

Unfortunately, many advertising practitioners who work on creating screens of advertising believe in a simple mental model of attention that might be described with the metaphor of a light switch—the screen is on or its off. A television commercial is recalled, or it’s not. A print ad gets noticed, or it doesn’t. An ad turns the light bulb on, or else a brand is plunged into the darkness of non-awareness. But is this light switch metaphor, which treats the consumer as a somewhat passive receiver of information, the best model for describing the active form of attention involved in consumer search behavior?

Here’s a simple thought experiment. Imagine for a moment that you find yourself in a pitch-dark room. It is a large and unfamiliar room. You move forward carefully, in slow circles, trying to discover what the room contains. In your hand you are holding a flashlight. You use the narrow beam of the flashlight to scan the darkness, sweeping the wand of light here and there. Occasionally, the spot of light comes to rest for a moment on recognizable objects—an attractive pair of red shoes, a sign with letters calling out a “sale”, shelves full of boxes, an empty leather chair, a cash register.

Which is closer to the way we normally attend to information in the real world, the light bulb or the flashlight? Do we read our physical environment in the same linear way we turn the pages of a book?

The ancient Greeks thought that the way vision worked was that the eye emits particles of light, which we shine on objects in the world in order to see them. This misconception was created by astute introspection. These philosophers reflected on how eyes functioned as a search engine for the conscious mind, and they arrived at our flashlight metaphor for attention.

In fact, the flashlight metaphor is a good representation of how vision works on a biological level. The only part of our field of vision with clear resolution is a narrow area of about 1 degree of visual angle around the retina’s center, an area the width of our thumb as it looks when held at arm’s length. Outside that region, resolution drops off sharply. To compensate, when we look at our physical environment we constantly move our eyes to bring the sharper region to bear on different portions of our surroundings we wish to observe. The brain then assembles all of these fragments together to create a holistic perception of what we are looking at.

For advertising in the physical environment of a store, therefore, the first job of advertising is not to turn on the light bulb –that comes later, with the delivery of an idea–rather, it is to grab hold of the wavering beam of consumer attention, to hold on to it as long as possible, and to direct and focus it on objects and information in the room.

Finding Focal Points of Attention in the Store


What kinds of information can be collected from an on-line store exit interview? Figure 2 shows some of the data that we collected from the picture-sorting part of exit interviews of a discount shoe store. This was a fairly large store as shoe stores go and we used fifty photographs to represent the in-store communications.

The first picture sort we conducted was whether or not the shoppers in the store remembered seeing particular shoe displays in the store. What we found was a wide variation in how well different displays were recalled: best recalled was an eye-level display women’s shoes with the brand name and price point-prominently displayed; less well recalled were women’s shoes that were on “sale” but without an obvious price; and least recalled were a display of men’s shoes—because, not surprisingly, the majority of our sample was women.

One of the things that we could then do analytically was to look at the differences in perception as a function of whether or not the shopper actually bought something while they were in the store— “walkers” versus “buyers.” One of the interesting findings was that sale displays that had the shoes stacked underneath where the customer could actually try the shoes on by herself had a much higher conversion ratio of walkers-to-buyers than displays where the customer had to ask for help in getting the shoes. These focal points of attention were, in fact, “action points” in the store.

One of the uses of this research was to help our client understand the differences between high performing versus average performing stores in terms of quarterly sales. Because there were quite a few differences between the stores in how the store manager arranged the in-store communications, the photographs taken from each store were also different—though we kept the number of pictures constant from store-to-store. One simple metric of in-store communication effectiveness—one of several ways of “scoring” each store in the system in terms of its communication effectiveness—was to look at the average number of pictures shoppers recalled from the store.


As you can see in figure 3, stores that generated above average sales also generated a level of image recall that was one third higher than the average. In other words, the more successful stores were more effective as communication channels for in-store advertising. In a sense, they were more easy-to-read, or more searchable. Consistent with this finding we also note, in the same figure, that, on average, buyers read more of the store than walkers who didn’t buy anything.

Identifying Emotional Hot Spots

Of course, just as with other forms of advertising, knowing that a shopper paid attention to something doesn’t tell you how they felt about it, or what thoughts it associated with the store brand.

When we use picture sorts to deconstruct the consumer experience of advertising, whether of a television commercial, a print ad or a store, we normally conduct three picture sorts. The first sort, based on recall of the image, is used to measure where consumer attention was focused—which we illustrated with the shoe store example. The second sort is designed to measure emotion engagement, both positive and negative.

Another study that we did, this time for the media company MTV, illustrates how the second sort, based on emotions, can be used as part of an exit interview out-of-doors, as part of a PR event.

As part of their event marketing activities, MTV sponsors a school bus that tours college campuses, putting on tailgate parties that promote viewership of MTV and their sponsors’ brands.


Even in an outdoor setting it’s easy to photograph the advertising-charged elements of the event with a digital camera. And as before, it’s easy to pass out invitations to consumers leaving the area of the event to take an on-line interview describing their experience. From a data collection standpoint, therefore, this was just another on-line exit interview.
An illustration of the consumer emotions evoked by the different advertising-like elements at the MTV tailgate party is shown in figure 4. In this example, Slim Jim’s were emotionally hot, but Visa cards were not.

An interesting application of this on-line exit interview research was that we could link it to other on-line TV research data in order to analyze cross-platform engagement—to study the powerful interaction between the brand perceptions created by including Slim Jims in this exciting live event with the perceptions created by running a commercial for Slim Jims on the MTV channel.


It should be obvious from our two examples that our definition of in-store “communication” is not simply limited to signage. Physical displays, visual iconography, prominent brands that cue quality and other things in the store are potential “cues” for consumer perceptions of a store in terms of the brand values the store itself wants to stand for. Digital photography can create an enormous and highly flexible visual vocabulary for describing potential points of interest in the store in a way that can easily be measured and uploaded to a web portal so that managers can see each store through the eyes of the customer.