The Traveling Salesman and the Grocery Shopper
Did you hear the one about the traveling salesman and the grocery shopper? No? Well, it goes like this. Traveling salesmen are always moving from one customer to another looking to stay on schedule and get as much done as they can. The same is true of grocery shoppers and that is why a new study uses a concept known as the “Traveling Salesman Problem” (TSP) to better understand how consumers shop around the supermarket.
Wharton marketing professor and RetailWire BrainTrust member Peter Fader along with
his Wharton colleague, Eric Bradlow, and doctoral student Sam Hui have written a new paper, The Traveling Salesman Goes Grocery Shopping: The Systematic Inefficiencies of Grocery
Paths.
As the research paper points out, TSP-type problems have interested academicians and practitioners, but a lack of in-store data has made it difficult to gain quantifiable insights.
For the study, Wharton researchers worked with Sorensen Associates and its PathTracker system employing radio frequency identification (RFID) tags on shopping carts.
The study tracked shoppers as they moved about a supermarket in the western U.S. The research analyzed the paths taken in 993 shopping trips and put that together with the corresponding purchases. What researchers discovered is that, while consumers were good at choosing an order in which to move about collecting goods in the store, many are taking a longer path than is necessary.
Shoppers “might have the right order, but they are taking too long and going too far to get from point to point,” said Prof. Fader. “It seems like people are doing a certain amount of wandering around regardless of how much they buy.”
Shoppers that do the most wandering, not surprisingly, also tend to be the same ones with the most in their carts when checking out, according to the study.
According to Herb Sorensen, Ph.D., president and CEO, Sorensen Associates, shoppers
“spend only 20 percent to 30 percent of their time actually acquiring merchandise,” leaving retailers an opportunity to determine ways to get customers to buy more in the remaining
time on shopping trips.
“We can’t manage that time intelligently if we don’t know what’s happening,” he said.
Greater use of in-store media may help retailers address the opportunity, according to Dr. Sorenson, who added “$300 billion of advertising money will move into the retail space in the next five years.”
While consumers spending more time in the store tend to checkout with larger market baskets, the researchers pointed out that making shopping more difficult and time wasting can have serious top and bottom line consequences for grocers.
Dr. Fader said the research collected didn’t help definitively answer the question as to why some shoppers spent what appeared to be “needless time” in the store.
“Some shoppers may be hedonic browsers … who like to wander around the grocery store and derive utility in ‘window shopping,’” according to the researchers. “Other shoppers may not have enough knowledge of the store to remember where the products they wish to purchase are located.”
“By no means does this study solve all the burning questions that keep retailers awake at night, but it’s a step in the right direction,” said Dr. Fader. “The main point is that we’re bringing hard science to an area that’s been left to judgment and intuition alone. There are still many steps awaiting us on our ‘path’ to better understand in-store behavior, and we hope we can be fairly efficient — unlike most grocery shoppers — as we move ahead with it.”
To download a copy of the research paper, The Traveling Salesman Goes Shopping: The Systematic Inefficiencies of Grocery Paths –
go to this download page.
Discussion Questions: Is it useful and relevant to study “shopper inefficiency”? Should retailers try to create sources of inefficiency, and how do they
know how far to go in this regard?
It’s an interesting conclusion that shoppers are spending only 20 – 30 percent of their time acquiring merchandise and are taking a longer than necessary path. Now, we ask – why? I would like to see results for the following: What are the percentages of the following grocery consumers for an entire chain, region, market, and even store level?
Strict Budget Consumer:
Consumer #1 will only spend X amount of dollars per trip.
Flexible Budget Consumer:
Consumer #2 will spend between X and X amount of dollars.
No Budget:
Consumer #3 has no restrictions on spending
If a particular market or store location has 80% of Consumer #3, the opportunity exists to significantly increase basket size. On the other hand, if a market or store location has 80% of Consumer #1, this opportunity to increase basket size is obviously minimal.
Wandering shoppers may be wandering because they actually are enjoying the shopping experience. So, we need to be careful that we don’t try to capitalize too much on their enjoyment and overmanage it. I see Costco shoppers wandering all over the place and it seems to be because they see impulse items in the distance or smell the samples being offered. When shoppers are wandering and happy, why mess with it?
This is good information. Anytime we can get an idea as to how a shopper is going about their business, we all stand a better chance of determining the best way to present items for sale.
This study makes me wonder about other facts about shoppers, such as: when they enter the store, is there is something on their mind which they must be sure to buy? Is there something on their mind which necessitates that the shopper specifically look for certain types of items? We know from the FMI/Prevention Shopping for Health study that over 50% of shoppers enter a store with a specific health need they want to address, and that over 40% would like signage indicating healthy choices. Consider the Hannaford “Shining Stars” system; this is the direction things are heading. The consumer is more savvy and knowledgeable and will expect to be treated as such.
This is a fascinating study. The challenge for retailers and brands alike is to capitalize on the 70%-80% of time that shoppers are not actually acquiring merchandise. As Dr. Sorenson points out, there is a wonderful opportunity to introduce media-based experiences that are properly designed to engage consumers during their ‘wandering’ time in such a way as to allow them to discover and be be surprised. All too often we’ve seen a ‘visual speed-bump in the form of a plasma screen placed in the consumer track simply because its location increases the ‘opportunity to see.’ How uncreative! Rather, design, create and engineer a magical experience that keeps the consumers delighted and in the store longer. This, as the study indicates, will increase the size of the basket. Perhaps, as the recent Hollywood film reveals, the Prestige of this magical experience could be unveiled through an online experience that would extend the relationship between the brand and consumer while inviting them to return.
I think there may be a causality question here – travel inefficiency is correlated with more shopping (items in the basket, areas visited, etc.). It doesn’t mean that traveling inefficiently causes increased sales. It may be that buying more stuff and visiting more areas causes one to be less efficient. A forced walk (like in Stew Leonard or the old days of one-way aisles) would remove this efficiency – that would, in theory, reduce purchasing. We know that’s not true, so the increased exposure of a forced walk has positive effects too.
While it is no doubt true that shoppers who spend more time wandering around the store buy more, encouraging this seems like a very dangerous idea. There is a lot of research suggesting that most people feel they are under tremendous time pressure. Many shoppers are even willing to pay more to save time. Focus on making it easy for people to be inspired with a great spur of the moment idea; don’t try to slow people down.
I think we need to be careful about what we are measuring. Are we tracking wandering time here or the response to an inefficient layout? Are shoppers happily milling about waiting to find something new to buy or are they frustrated because they can’t find what they are looking for? I don’t know too many people who feel they want to spend more time shopping for groceries.
I seem to remember that the original Stew Leonard supermarket in Norwalk Connecticut had a unique layout: it wasn’t a grid. The customer had to walk down every aisle to get to the cashiers. She had to pass by everything in the store. It was the ultimate in customer-path control. I cannot think of any other retailer who’s done that.
Yes, understanding shopping patterns is useful and the inefficiency component is interesting but I would think there would be some observational work with selected shopper discussions included in the study to answer the inefficiency questions. Maybe they are comparing labels, or maybe they are looking for a specific product that is lost in the SKU jungle? With marketing dollars moving from media to consumer and retail, understanding what is going on in “shopper heads” is key. This kind of information would help everyone across the value chain develop better products and packages rather than leaving us wonder about shopper inefficiency.
Is it being suggested that because customers that shop “around” the store spend more than those that shop more efficiently, we should make sure everyone is less efficient?
The inefficiency of shopping is symptomatic of a shopper that is browsing for things to buy and is more likely to pick up extra things for their cart. The inefficient shopping route isn’t the causal factor driving up purchase rates. Why then, are we talking about rerouting customers to get them to buy more?
Shouldn’t we be looking at what those “wandering” shoppers are buying and what stimulates them to make those extra purchases? It seems to me that understanding the merchandising elements or feature pricing that drives purchases among those wandering shoppers is where the real insights might be.
This is a terrific study. Having the quantifiable information of how long it takes shoppers to complete the trip and the path they travel provides significant new information about what is going on in the store. Like any good study it raises significant new questions.
We know that there are many different kinds of shoppers and some of them do enjoy the shopping experience (not many but some). We know that some shoppers look for new products. We know that some shoppers frequent more than one store — those shoppers may forget the layout of a particular store and have to backtrack while doing their shopping. Trying to map which shoppers take a longer time is an important next step.
Another piece of information not identified in this study is what people are doing as they go through the store. Are they reading labels? Are they stopping to talk with people they know? Do those conversations remind shoppers of something they forgot so they have to go back and get it? Have some of the shoppers identified a particular meal but find that they are unable to get some of the necessary ingredients so they change the meal and have to go back for other items? Are some of the shoppers just not focused on what they are doing and have to keep backtracking because they are not focused?
Eighty-nine percent of all grocery shoppers (sample: approx. 8,000) say that finding the items they want to buy is “very difficult” or “somewhat difficult.” Eighty-five percent say that they find layouts and looking for items in stores “very frustrating” or “somewhat frustrating.” From this I draw the conclusion that shoppers care about being able to find what they want (but our data also shows it is not in the top four expectations), that they are not able to find what they want easily, and that this bothers them quite a bit.
Yes, let’s study scientifically all of these things, and more. I hope we can use our cleverness to focus on consumers, what they need and want, and how to provide those things at a consistently very high level. Retail, including grocery, has some fair way to go in meeting consumer expectations — the overall score is 65%. That is not something to cheer about. Let’s find a way to improve this score, and find a way to get retailers off the idea that “we’re doing good enough.”
This was a good study that brought some obvious issues forward. However, there is a caveat to some of their data. We don’t know how familiar with the store the shoppers were. We need to know how many times they had visited that store during the last week, month and even several months.
We also don’t know which visit represented a “traditional” visit, or which were driven by ads. We also don’t know how shopping duration was impacted by visit frequency. Finally, we don’t know how lack of familiarity impacted the shopping experience, including some basics about the shopper (age, income, family size and the use of a cart vs. basket).
This study did reaffirm the obvious things that we know about retail including: 1) The longer a person remains in a store, the greater their propensity is to purchase something. 2) Every time a shopper purchases something they are just that much more likely to purchase an additional item. I am skeptical about the rest of the research since it left many questions unanswered.
Being that I am in the business of shopping, I don’t encourage any grocery retailer to make my shopping experience inefficient. I founded a nonprofit called Shopping Cart, Inc. that strives to promote independence for frail elderly, disabled and homebound persons by providing low cost shopping and home delivery services so they don’t have to rely on friends and family members to do the shopping for them, who by the way do not have the time nor luxury to mill about the grocery store. They want in and they want out.
There are simply too many different types of shoppers out there to lump them into one, two or three categories. If I chose to spend an hour more than I ought to milling about the store, for whatever reason, I would prefer it be my choice to do so.
Interesting discussion — it is always fun to think about using new data. What occurs to me is that it would be interesting to analyze where the most traveled areas are and compare that to sales generated in those areas. This could be combined with interviews with the shoppers to learn what they purchased that they did not intend to determine where the opportunities for incremental sales are.
I am, for the most part, a wanderer in the supermarket! One would think it is because I like to look around and see what’s new but in reality I’m like a lot of shoppers who shop WITHOUT a shopping list. This causes me to wander around for a number of reasons: 1) I do not know what I want for dinner or an occasion; 2)I wander around trying to recall what I might be out of at home; or 3) I did not take the time to read the retailer’s ad and am shopping the store to view the special sale items in various departments. When I have a list it’s definitely a hit and run operation…when I don’t have a list it’s browse the store and be victimized by the treasure hunt of taste and smell that Al referred to in his Costco store visit.
Although not surprising, I think by quantifying the findings, Professor Fader has given supermarket designers some good reason for considering the “wandering around time.” The challenge is that it is difficult to accommodate all the various reasons this might occur. As the discussion indicates, various customers will have different reasons for taking their time as they shop. And even these shoppers will have days when they “just have to get things done” so how do you build a physical environment that meets all their requirements?
The answer is you don’t. This is where the next wave of online shopping is going. Maybe the first question the online experience should ask is “How much time do you want to spend on shopping today?” The response will dictate the type of experience the “virtual store” provides the shopper. If they have time, they will be offered recipes and substitutions. If they’re looking for a “quick pick up,” they can go directly to the desired category and select their items. This is where the online experience can excel.
Tried to read the report, but I don’t have an advanced degree in engineering.
It is nice to see someone translated the information into something a grocery person could understand.
I think a lot of this is related to being unprepared for the shopping trip. Lots of people have short lists but they try to maximize the shopping experience once they get to the store. It also has to do with the frequency they shop a store. If they are repetitive shoppers, they are more likely to be more familiar with product location than if they only go to the store periodically.
In addition, I would be anxious to know if there was a significant difference between the behavior of the shopper due to age and sex. I do know from my years of study (my wife and myself) that she is a shopper and I am not, and that has to play in the mix somewhere. Bringing science to this matter is encouraging but it will be complicated and I think we have to be careful in interpreting the results.
The systematic way of measuring shopping trip efficiency in supermarkets proposed by investigators Hui, Fader and Bradlow is certainly valuable. The triangular diagram included with their working paper is of particular interest. Great to see these strong academicians working with Sorensen’s trip data in this way.
Now that we have this better way to characterize and compare shopping trips, we face the subtler questions surrounding how retailers should choose to apply these learnings to influence trip efficiency. Re-mapping store departments might be one action, but this raises the questions: Should we attempt to make trips more or less efficient? For which shoppers? On which trips and need occasions?
More areas for further investigation: Does interacting with Shopper Media tend to cause shoppers to extend their trips? Is this good or bad for sales? For loyalty? For total customer experience?
Inefficiency strategies abound in retail. Consider the mall operator who locates its two best shoe stores on opposite corners of the layout to force comparison shoppers to criss-cross past other stores and kiosks. I’d submit that shoppers sense when they are being played. They probably don’t like it very much. Before we try to lead them down a meandering garden path in our stores, we’d best carefully examine the effect this will have on our shopper relationships.
Who, what time, what day, and where are these grocery shoppers
being tracked would be my first question?
I would be interested to know more specifics, i.e.: demographics, especially age, and gender, etc.
Isn’t it curious that certain shoppers have idle time, and their valuable time is spent in the supermarket?
I wonder if these shoppers are the frequent visitors of the supermarket? Hmmmmmmmm
Bernice Hurst pointed out one other retailer, like the original Stew Leonard’s location that forces its shoppers to see everything: Ikea. Ikea and Stew Leonard’s have a similar profile: great success and unique merchandising.