Retail TouchPoints: In-Store Insights Help Turn Browsers Into Buyers

Through a special arrangement, presented here for discussion is a summary of a current article from the Retail TouchPoints website.

A missed opportunity is walking out of stores every day: the browsers who could be buyers, and about whom most retailers have little or no information.

Most of the analysis that’s done today is based on POS data. If 80 percent of the people that come into the store don’t make a purchase, they are absent from the analysis that retailers use to feed decision-making. Other than door counters used at a small percentage of retail outlets, there has, historically, been very little intelligence or insights available about browsers in-store.

But advances in video cameras and related technologies is making the gathering, quantifying and analyzing of such in-store visual customer intelligence possible. With the latest solutions, store management can see where shoppers stop and spend time, what areas of the store get the most traffic and which the least. Traffic to, and engagement with, promotional displays can be measured and analyzed.

Among the ways visual customer intelligence can be used to better target shoppers:

Labor – Match Staffing and Traffic Levels: Staff scheduling, commonly based on transaction volume, does not account for the highly variable nature of conversion rates by hour of the day and day of the week at both store and department levels. For example, in one retail outlet, a 40 – 50 percent drop in conversions was recorded in the early afternoon; store operations was able to trace that drop to staff members leaving the sales floor to manage inventory in the stockroom. In another store, visual customer intelligence showed that employees were twice as busy on weekends, serving double the number of customers. The store was able to more efficiently adjust staffing levels to better balance weekdays and weekends traffic.

Marketing – Measure Marketing Effectiveness: With a view of how many customers entered the promotional zone and engaged with the product, combined with sales for that product, marketing professionals can see the actual impact of their campaigns. They also will have the ability to troubleshoot, and make small tweaks that turn browsers into buyers.

Merchandising – Know Where Your Browsers Go: Imagine the power of witnessing a 300 percent difference in browsers between the busiest and next busiest branded promotion area, and what that means in terms of the relative value of that space. In one instance, one retail customer was concerned when a heavily promoted product was not selling as well as expected. Analysis of in-store traffic flow and engagement data revealed that the product was located in a promotional zone where only four percent of shoppers ever ventured. The product was moved to an area that recorded high traffic regularly, with a subsequent sales uplift for that product of 35 percent over other stores in the chain.

BrainTrust

Discussion Questions

Discussion Questions: Where do you see the greatest potential benefits of visual customer intelligence as well as the biggest hurdles toward adoption? Do you see such technologies as potentially most beneficial to labor scheduling, marketing or merchandising?

Poll

11 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
Paul R. Schottmiller
Paul R. Schottmiller
11 years ago

Video recognition is advancing rapidly. The software can now predict gender better than you can. Age and ethnicity are not far behind. On the horizon is mood, heart rate, specifically identifying you and who you are shopping with, and detailed shopping behaviors (i.e. comparing).

The real payoff comes when retailers marry this information with other data sources (oops there is that big data stuff again) and create a realtime/neartime capability to react to it — in other words, affect you at the moment of truth during this shopping trip.

Gene Hoffman
Gene Hoffman
11 years ago

Many people browse because they like the experience. They never intend to buy. It gives a little meaning to their life and available time, much like “window shopping” did for people during the Depression days.

Other browsers are uncertain of the value equation between acquisition and money retention, but they are potential customers. Once such a browser is in a store, you should merchandise directly to their perceived whims to balance the scale toward a sale. Of course, you have to research what their whims are.

Robert DiPietro
Robert DiPietro
11 years ago

I think the biggest winners in visual customer intelligence are labor scheduling and merchandising.

1) Labor scheduling is mostly done off POS data vs customer traffic and flow so you could optimize labor on when you have the most flow. It could also help by zone assignments for the associates as the article references … don’t have associates out back when customer traffic is peaking. Retailers could really focus on where in the store that labor is required.

2) Merchandising can be helped by not only heat mapping customer flow but measuring browse time and potential sight line optimization (e.g.. if most customers stand here, can they see all the HOT offers?).

Nikki Baird
Nikki Baird
11 years ago

From the case studies I’ve seen and the retailers I’ve talked to, the biggest predictor of sales success is average time per customer. When matched against a certain level of traffic and a target level of conversion rate (that whole turning shoppers into buyers thing), the store can become very predictable in terms of generating sales. So to me, that would be the most important piece of the visual analytics puzzle. The marketing and merchandising stuff is nice, but I feel like those two departments already have enough data to wrangle already, and I’m not sure that either one is best positioned to turn video analytics into insights and then act on them.

Bob Phibbs
Bob Phibbs
11 years ago

The biggest benefit is discovering the dead zones in a store, if that is possible. The other two used to be known by an engaged manager, one who had free rein over scheduling. Most today do not have that freedom nor are they measured using realistic goals.

Steve Montgomery
Steve Montgomery
11 years ago

Video research has come a long way since the days when we had to review hours of tape and manual note what we saw the customers and browsers doing. The ability of software to translate video into actionable data continues to progress remarkably.

The greatest value for each retailer will depend on where they are with the use of their other data. I would expect the application order for most would be merchandising, labor and then marketing.

Mike Spindler
Mike Spindler
11 years ago

Interesting topic. However, using video recognition on consumers is only half the equation. We have determined that WHAT you are offering to the consumer is more of a variable than you, or your plan might think. Broadly, the carefully thought out and trading partner-agreed-to plan is only 50% delivered to the shelf and consumer. 20+% of assortment is not as planned, and well over 30% of facings deviate from plan. Without delivering a predictable merchandising plan to the consumer, how can you evaluate anything?

Ed Rosenbaum
Ed Rosenbaum
11 years ago

Give the browser what they want and they might make a purchase. However, there is always the “time killer” browser who has nothing better to do than look. Once they are approached, they will move to another department or even another store and resume the browsing. What a waste of productive time this is.

Ted Hurlbut
Ted Hurlbut
11 years ago

In this world of retail data overload, where retailers are struggling to digest all of the data they’re already collecting, I think retailers would do well to focus on what the customers who are buying are trying to tell them. Frankly, I’m more interested in what I can learn from those customers than I am from customers who aren’t buying.

Herb Sorensen, Ph.D.
Herb Sorensen, Ph.D.
11 years ago

We began collecting with our PathTracker(R) methodology in 2001 — first with RFID, and within months moving to video for stores where carts and baskets were NOT good surrogates for shoppers. And beginning a few years ago we began experimenting with image recognition software — which does have a lot of potential. But a huge problem with real-time data is that management cannot possibly respond in real time, so there is a heavy investment, capital wise, in making real time available, with close to zilch ROI.

Instead, we developed personal audit methods that allow us over a few days, to determine precisely the information being discussed here, at a tiny fraction of the cost of technology solution methodologies. Even fitting a small sampling of shoppers with unobtrusive video-glasses-cams can add a layer of explanatory insight beyond anything you will ever get from observing shoppers — instead, see through the shoppers eyes what the shopper sees.

The BrainTrust discussed some of these issues in 2008. The original paper is at Long Tail Media in the Store.

Ralph Jacobson
Ralph Jacobson
11 years ago

Wow, for some reason this whole topic strikes me as a very old challenge. We’ve been trying to crack the code on this for years. Realizing there are different types of browsers/shoppers (some whom had no intention of buying anything, etc.), I think this category of technology has its merits for merchandising/store layout after “hotspots” in the store categories are better defined. Effective deployment and coaching of staff is critical. How many times have you walked in and then out of a store to never have been asked if you need help? By the time you reach the POS, the customer has already “checked out” of the store in their own mind, so the cashier asking if they found everything is virtually moot. These technologies can help the areas mentioned and more. Gotta integrate the staff participation, though.