Retail Customer Experience: Why customer satisfaction surveys don’t work
Commentary
by William Cusick CEO and founder, Vox
Through a special
arrangement, presented here for discussion is a summary of a current
article from Retail Customer Experience,
a daily news portal devoted to helping retailers differentiate the
shopping experience.
Recent evidence
suggests fully 95 percent of our cognitive processing is subconscious.
That leaves all of five percent of the rational mind to do what it does
best: rationalize (some might say “guess about”) our decisions and actions
to ourselves.
What that means
is we’re pretty poor at telling others what we like or don’t like, and
why we feel that way. And the science suggests we’re even worse at predicting
what we might like, or do, in the future. One might wonder then, if we’re
so horrible at this, just why businesses continue to crank out standardized
customer satisfaction surveys. How can you get closer to the truth, to
determining real, actionable steps to drive customer behavior, when you
don’t know what they’re thinking in the first place?
And there lies
the real paradox: our actions are driven by emotion (in our irrational
subconscious) much more than logic. Yet, to really understand your customers,
you can’t look at those emotions. Instead you must take a step back,
stop making assumptions and focus on their behavior. Behavior, it turns
out, leads to the truth.
Paco Underhill,
a self-described “retail anthropologist,” has understood this for a long
time. In his book, Why We Buy, The Science of
Shopping, he makes a compelling case for employing
observation of customer behavior over other techniques like surveys.
To help his clients get closer to the truth, he doesn’t ask customers;
instead he sends his “trackers” out in the field to the actual retail
environments, and observes customer behavior in real time.
It was through
this power of observation that Mr. Underhill discovered what he referred
to as the “butt brush factor.” He noticed that there were serious and
unintended consequences when two product displays stood in close proximity
to each other. If a customer wanted to bend down to take a closer look
at a product on a lower shelf, it forced passersby to turn and shuffle
by, resulting in said “butt brush.” This seemed particularly uncomfortable
for women, and it meant very low sales on the products in those displays.
The behavior, in other words, held the answer to an actionable improvement
to the customer experience, and to desired customer behavior.
This same idea
— that behavior is truth — holds in the online retail world as well.
You can’t just ask customers if your website is “satisfactory,” or what
improvements they’d like to see.
Can you see
just what your customers are trying to do on your site? Where do they
enter, and what path do they start down? Where are the road blocks? It’s
been our experience that, once you look at the behavior, it’s not that
hard to see where customers are abandoning the site, where they are stalling
or backtracking, and more.
So remember,
it’s not what they say, it’s what they do.
Discussion questions:
What are the pros and cons of behavioral research?
How well suited is it to analyzing consumers’ web experience? What
value do you place on customer satisfaction surveys?
With so much of our cognitive processing being subconscious, marketing professionals do work with psychologists to help differentiate consumer impulse and actual consumer behavior, from the more thoughtful conscientious choices made by the consumer that is involved as a study in behavioral consumer research. What we teach in our own consultative workshops is how to ask exactly the right “questions” to get the accurate responses. “What did you do?” is more accurate than “what will you do?” “Why?” is where the psychology comes into play.
I am a firm believer in observation (real and virtual). Recently, while doing research on value-added dairy products, a focus group member described the dairy case as “cold, white, dull, boring metal.” However, such a vivid description did not adequately capture shopper behavior in the dairy section. With a stop watch, pad, and parka I observed customers buying v. shopping this section–time spent measured in seconds. This behavior works for buying milk, scanning for color of cap and expiration date. However, the coldness and sterility of the section was not inviting to customers shopping for value-added products, namely, flavored dairy creamers, soy and organic dairy products, etc.
The result of this observation was to place a small refrigerator in the coffee aisle. Sales of flavored coffee creamers (real value added) went through the roof. As a post-script, the experiment almost failed as the store couldn’t keep the case full.
Survey techniques and focus group research alone would not have produced such a success. As the expression goes, “don’t watch what I say, watch what I do.”
Mr. Underhill is working on some interesting concepts. That said, the approach appears to be very expensive if done for one specific client. What would be interesting is if Mr. Underhill created a report including 10-15 findings that he could then package as a solution for clients. With some small customization the work Mr. Underhill has done would be more widely used and ultimately accepted. It would also allow smaller retailers and manufacturers to invest in his work as well.
Survey’s do have their limits, but they are a cost effective way to get some insights into what the customer is thinking if done correctly. I recently moved away from NetFlix. When I closed the online account they asked me to participate in a 4 question survey. One of the questions was very relevant and I think helpful to Netflix understanding their business better. It asked “where I was going to get movies in the future from?” The options included stores like Blockbusters, their online competition and vending machines like RedBox. I gave then an honest answer that I was moving to RedBox. That information will hopefully help Netflix improve their competitive offering in the future which is beneficial to me the consumer.
Knowing your customer better is great for everyone involved including and most importantly, the customer.
I’ve said this many times before in these forums, but customer’s actual purchases are the best window into a person’s preferences (or as Richard George put it: “don’t watch what I say, watch what I do”). Tracking shoppers in the store, as discussed in this article, can give you additional insights, but the existing shopper loyalty databases are the real gold mine. As a marketer who is introducing, say, a new low-fat ice cream, you are far more likely to sell your ice cream to someone who actually buys a lot of low-fat products than to someone who says they think they would buy it.
The challenge in working with this data is one of scale. Unlike surveys, which can be summarized and tabulated relatively easily (even for big studies), homing in on shoppers based on their shopping behavior can often be like finding a needle in a haystack. A typical supermarket location can generate 50 MB/day of T-Log data (multiply that by the number of stores in a chain), and most BI/analytics tools that can handle that quantity of data require highly technical users in the driver’s seat.
In order to get the most from this resource, you need a technology platform that can handle the massive data volumes and that can be used by a non-technical marketer, salesperson, or category manager.
This is a meme that my mentor, Bob Stevens, was leveraging at P&G in the ’60s and ’70s. As a scientist, I “accidentally” got involved in in-store research in the ’70s. Probably well less than 1% of “shopper” research is actually conducted in-store, and of that, probably well less than half is actually observational research. For my own firm, I doubt that as much as 20% of our in-store research is observational. And there is a reason for that: once an observation is made, right then is the time to ask the shopper why, and what were they thinking. So marrying observations with asking questions will always yield more insight than simply asking questions.
Nevertheless, most of what pretends to be “shopper” research is just ask-type research conducted far from stores, essentially consumer research in new clothes.
I am a firm believer of integrating behavioral data with consumer verbal responses. We are able therefore to observe what consumers do and develop insights as to why. Marketing research should be about selecting the appropriate tools, perhaps in combination, to obtain the desired information and insight. It does not have to be an “either or” choice.
I’ve recently had the opportunity to contribute to the online marketing research course “Principles of Marketing Research” sponsored in part by the University of Georgia. The modules of the course emphasize how important it is to relate the end use of the research with the methodology selected. Every buyer of research has decisions to make about time and budget as well as research objectives. I therefore find it unsettling to read an article that suggests one size fits all.
If you want to develop or enhance an existing website to deliver visitor satisfaction and more sales you need to learn what it is that will deliver the end result. Click through’s can tell you what has happened, eye tracking can show you how it happens, but only the verbal question and answer can deliver the initial steps to understanding why it happens. Now the superior research analyst (dare I say “me”) has to align all this information to create guidelines for future development. This requires setting a goal at the outset, defining the methodology and the process to achieve the desired results. You can’t grab a technique off the shelf and expect it to adapt to the nuances of your situation. Well, I guess you could expect it to, but you’ll be very disappointed.
Similarly, observing the synapses in the brain can shed light on emotional responses. However, these factors are not stand alone analyses, but rather contribute to greater understanding. In some cases you must have the time and fortitude to keep searching for the answers when they are not clear. Again, it takes defining the problem, developing the appropriate methodology (which means you have to decide why this one is better than another) and depends on an appropriate analysis.
Retailers and manufacturers know what they want in terms of business goals and revenue. If they explain their wants and needs to their marketing research consultants, they can establish effective research programs that will facilitate productive insights and business growth.
Mr. Cusick raises an important issue and several considerations. People’s/shoppers’ cognitive processing comes from their background, personality, areas of interest, social economic situation, etc. Indeed, as he says our “actions are driven by emotion” which comes from realities of personality, prior experiences–that is how the best stores accommodate shopper personality and expectations. Shoppers expect a different experience when they visit Walmart than say, Bergdorf Goodman. My point is that shoppers’ service expectations are formed and influenced by how they are treated in the stores as compared to how they are expected to be treated. Retailers have the power to influence this dynamic by adopting continuous concept improvement of in-store methodologies.
Well I thought we were going to get into a discussion of what I believe is the next frontier of understanding and being able to create the circumstances of our dreams. We came oh so close.
Yes, 95% of our life experiences (including how your body works) are controlled by the subconscious. The “sub” part of that word is key. Certainly looking at what people actually ‘do’ is a lot more on target than asking them what they ‘think’ which is almost pointless. That’s why it’s called a SUB conscious–you’re not aware of it any more than you’re aware of your liver functioning.
Our world will change ONLY when our subconscious beliefs are brought into alignment with our conscious desires.
For example, we all ‘say’ or ‘think’ we’d like to be rich. But if your subconscious has been programmed with the message that “money is the root of all evil” and that rich people are going straight to hell, the subconscious will actually prevent you from getting rich. That’s because the only job of the subconscious is to look after and protect you. The reason so many lottery winners (who typically don’t make much money) throw away their winnings is because their subconscious says “Wait a minute, you’re not supposed to have all that money! It’s bad for you throw it away!” This kind of programming is all around us. You’ve never seen a western where the rich rancher was loved by the whole town and got the girl have you? And I don’t even want to comment on the lack programming from news channels.
Through recent discoveries in bio-energetics we’ve learned to help people identify the disabling subconscious beliefs they have and replacing them with enabling beliefs that move them toward their conscious desires. It is easy and it is instant. Imagine what would happen if every retail sales person could minimize or eliminate limiting beliefs! You’d see instant change.
But boy is it hard to get organizations to see this. Why? Because the subconscious beliefs of the those who make such decisions prevent it. They believe if you can somehow make people ‘do’ certain mechanistic things, circumstances will change. But that’s not going to happen. The subconscious rules! It’s a challenge equal to convincing people that their beliefs that the sun revolves around the earth are wrong.
Change subconscious beliefs and you change everything.
As a food retailing student in the ’60s at the University of Massachusetts, Dr. Theodore Leed had us map supermarket consumer patterns to see their “shopping path.” When I first met, and talked with, Mr. Underhill I realize what he was doing was similar in concept but at a much higher level. Everything I leaned in my career supports the fact the statements made by others on this topic–what people say and what they do are not likely to be the same.
We have also learned that the wording and sequence of the questions asked in a survey can bias the results. The development of the survey instrument often gets the short shift in the process. That being said, I believe traditional market research has its place.
I would like to twist the data presented here and look at it from a different angle. That is how we can apply the research to improve the hiring of better employees.
Just as customers are making a buying decision not based on what they know but based upon how they feel. That is the way most of us end up making hiring decisions.
Based on the article I come away with two key takeaways:
1. We need to understand more about how we determine whether some one is going to be a good employee or not.
2. We need to do a better job at behavioral interviewing…since past behavior is the best predictor of future behavior.
So apply the research not only to what you are selling but also look at how you are buying.
To say that 95% of shopping is subconscious and therefore that “customer satisfaction studies don’t work” is hogwash.
Consumers are looking for value and products that make their lives easier, better.
Ask me what I’m looking for in a pair of pants and I’ll tell you. I want to look good and not pay too much money. I hate trying on clothes so I want to get in and out of the store as quickly as possible. I want to get treated right by the staff. It doesn’t take a psychoanalyst to figure out what’s really driving me.
Any researcher worth his or her salt knows that there is always a difference between what consumers say on a survey versus what they actually do. There are numerous market research techniques that get beyond the difference between stated and actual behavior. Derived importance analysis, for example. Ask me what attributes I associate with a battery of brands and then look at what I’m buying most frequently. Chances are, those are the attributes that I’m looking for.
Now don’t get me wrong. Observational research is a wonderful tool. My company uses it all the time. But it can be as subjective as a survey. Let’s not forget observer bias. I wonder what was going on in Mr. Underhill’s subconscious when he came up with “butt brush factor.”
My point: Don’t throw the baby out with the bath water.
I agree with Ben Sprecher that the real gold mine lies in shopper loyalty data. Why ask customers what they will do if you have detailed information on what they actually do in the store? Loyalty data offers an opportunity to genuinely understand customer purchase behavior and more importantly, the dynamics of behavior change.
As Ben indicated in his comments, the challenge is making sense of the overwhelming amount of purchase data in a loyalty database. This is especially true when the goal is evaluating behavior change over time or in response to promotional stimulus, but that’s also where the significant learnings are.
Great comments thus far. As the author of the article being discussed (and of the just-released book that further discusses this, among other aspects of customer experience, “All Customers Are Irrational: Understanding What They Think, What They Feel, and What Keeps Them Coming Back”), I had a quick clarification: nobody is arguing that you don’t ever want to survey customers. But, as some folks noted, you need to create the surveys with a nod to the customer’s irrational subconscious. Focus on behavior, whether through observations or questions (as Netflix does in the example above, not asking the customer “why are you leaving?” but rather “where are you going?”), and you get a clearer more truthful window into future customer behavior and inclinations.
I’m a firm believer that customer observations provide a lot more value in most cases. However, if you are looking to support specific findings or test results, the shopper survey can be a valuable tool but shouldn’t stand alone as the only measure. Think about how many retail locations are using the “phone number on the receipt” as a way to gather feedback. If this is there only means of identifying purchasing patterns then they will ultimately fail as there is really no way to tie the patterns into shopping behaviors, it’s just a blind test.
I like having multiple tools to assess results and not lean on only one; to that end, I like the behavioral observations but would still use feedback from a shopping survey as part of an analysis.
When I read this I think of the parallels in “The Paradox of Choice: Why More Is Less” by Barry Schwartz. We need to examine the choices and use behavioral models, much like in loyalty or engagement marketing programs. Yet I think the number of surveys, products, websites, CHOICES has caused Americans to retreat. When customers have two choices, they can derive great value from one outcome. When they have 3, they derive a little less, when they have 500 they receive little in any.
Consumers are overloaded and they are retrenching. Surveys may not be representative if there is not a way to analyze the response, compare it to actuall behavior and a control group.
The behavioral-driven approach has been shown in a number of research studies (and in our clients’ experience) to be highly predictive of future behavior, while attitudinal research is much less successful. While this is true for customer experience, it is even more so for loyalty and retention.
Want to know who is most likely to buy tomato soup in a grocery store? Don’t do attitudinal surveys on the role of tomato soup in their lives–the most likely customers to buy tomato soup are customers who bought tomato soup three weeks ago. Past behavior is the most predictive of future behavior. Want to see which customers are most likely to repeat–look at past repeaters and when they would be usually repeat–they usually do.
So the best satisfaction analysis (not survey) combines attitudes–specifically likelihood to repeat and refer, with behavioral metrics related to the customer experience–wait time to check out, return and refund level, hold time on the phone, etc. These are the hard-and-fast metrics that can be related to loyalty and changed to improve the customer experience. What gets measured is what matters.
Make satisfaction (or better, loyalty, since satisfaction has been shown not to correlate to retention) your key metric and then measure the drivers.