Is simpler better when it comes to shopping analytics?
Wandering the floors of NRF’s BIG Show a couple of weeks ago, it seemed to me that the retail industry has made a lot of progress analyzing the behavior of shoppers, dissecting it in nearly every way imaginable, and then targeting those shoppers from all possible angles. Yet, we’ve sometimes remarked that all we’re really trying to accomplish is to get back to the days of yore, when shopkeepers knew the names of their customers, what they wanted, and how to engage them.
With that in mind, I paid a visit to the booth of a small company with a catchy name, offering a "Happy or Not®" system for retailers, that harkened back to a simpler time. It really is simple: a small, standing display is installed near the checkout or store exit, giving shoppers a chance to indicate whether their in-store experience made them very happy, sort of happy, sort of unhappy, or very unhappy by pressing a button with the appropriate smiley or frowny face. The data is then analyzed, assessing results by location, day, time, etc. Retailers can get daily reports online or via e-mail, with trend analyses included. The question can also be changed, so retailers can ask about product quality, selection, or anything else the retailer chooses.
![[Image: HappyOrNot]](/public/images/discussions/17299/happyornot-still.jpg)
The company has a number of clients in Europe and is in the process of rolling out in the U.S. Retailers using or testing the system include Tesco, Carrefour, HEB, Office Depot, Boots and Delhaize. The system is also in use at Heathrow airport in the UK, where it provides feedback on the passenger security experience.
Discussion Questions
Could retailers gain something by taking a step back from complex shopper insights and analytics to measure something as simple as whether the shopping experience made the customer happy or unhappy?
There is definitely an issue around the big data movement: too much data. So looking in the other direction of core data simplicity makes sense. However, I vote for a balance.
Getting too granular can lead to both big expenses and overzealous inferences. On the other hand, being too simplistic can fall short of meaningful inferences and usable data. I believe “happy or not” is far too oversimplified and only provides a general temperature check that provides little to no specific actionable data; for example, “Why are shoppers unhappy today?”
Shopper marketing analytics are here to stay. We live in a world of multi-stored, big box competitors, not the simpler times of a single shop keeper.
A tool like this holds some value in listening to the shopper who is currently in your store. However, does it offer a 360 degree view of this consumer — What is her demographic profile, shopping patterns away from the store, how much does she spend within a category, etc?
Those answers about the more complete consumer are available. By understanding some very basic demographic background, and then having the ability to cross-tabulate those data points with another set of data about shoppers — Net Promoter Score (NPS) or purchasing plans by brand (likeliness of a Chevrolet owner to be a buyer in the coming 6 months, etc. — the entire shop keeper’s team can be positioned to develop Propensity Scoring Models that point to predictive analytics.
Prosper has taken that approach in our daily work for clients. Analytics work.
I believe so. Big data and now analytics are the new buzz words and focus of most CPG organizations. However, it does not excuse retailers from the need to continually enhance customer intimacy. My concern with analytics is that we have the ability to analyze data ad infinitum. Unfortunately, sometimes the results of our analysis have little impact on the customer experience or even levels of sales.
Kind of reminds me of the days when CPG manufacturers and retailers hired MBAs from prestigious schools to be brand managers. These talented individuals paid more attention to crunching the numbers than to really trying to better understand their customers. Instead of spending time in the field observing customer behavior, they spent countless hours developing mind-numbing analytical reports, that had little connection to their customers.
Therefore, this return to simplicity and real customer engagement is welcome.
Possibly. Net Promoter Score (NPS), is a pretty simplified measure of customer satisfaction and ForeSee has recently introduced a slight variation on this.
But when a majority are using simplified measures, gaining an advantage on the margin requires more detailed analytics.
Yup – they could gain a lot. Not the answer to all questions, but a great starting place.
I used this at Heathrow recently and thought it was great because it was customer focused. It didn’t ask me a dozen questions to make it easier for the CMO to slice and dice. It was an easy ask and easy answer — and most everyone was using it on their way out of customs, even though they were rushing to get out.
I would take this any time over a Netpromoter score.
I have long been a believer of quantitative results as they provide a much more realistic snapshot of what is actually happening. “Do the math” has become my tagline. However, simplicity is always better, and the approach taken by “happy or not” combines both quantitative results and direct customer feedback. The interface has an additional bonus in that it is not “spying” on shoppers by way of facial tracking, mobile phone signal intercepts, etc.
These “anonymous” technologies provide useful and valued insights only if the retailer acts on these insights to bring more value to the shopper. Unfortunately many of these implementations are used to invasively inject another advertising, promotion or merchandising scheme. Combining quantitative and simple is a great combination and the results can be easily understood by the executive committee. That should make everyone happy!
Many retailers can gain a great deal if they applied the principle of “Occam’s Razor” … which generally states to choose the simplest solution if two or more are available.
Knowing that more customers are happy, or not happy, is better than knowing nothing about them at all. The core growth question is WHY … Why are they happy or unhappy?
The real questions for retail survival is HOW do they shop and WHAT can I do differently to change their experience in my stores?
The successful stores today are using much more than HappyOrNot buttons. In store traffic analyses and location analytics will become critical success factors for the store of the future.
You know how I measure a happy customer? It is when I see them week after week, coming in to support my business. There is so much high-tech stuff that is supposed to make your store like Disneyland, but it doesn’t do one thing to create the one-on-one moment when a customer needs to talk with you about their unique situation.
It is extremely tough to stay in business today, and to me it is more important for me to be available when my customers need guidance on catering, or how to budget their money. Sometimes they like to vent to me about their kids, or the weather, whatever, but either way for me, it is the relationships I build over the years that may keep me around for many more years.
State of the art checkout systems, SKU management, procuring the right product at the right deal, and having extremely friendly honest workers is the cornerstone to survive against very tough odds in our digitally overloaded world. Don’t agree? That’s fine, but try to run a business without the special human touch needed to make the difference vs. the BIG BOYS, and see how long you can stay profitable.
One primary objective of “complex” analytics is to treat customers as they may have been treated two hundred years ago in the country store, where the owner knew each of his or her customers personally.
But how do you do this when you have not hundreds of customers in one location, but millions of customers worldwide. The answer is, you need sophisticated analytics that learn and remember your customers’ preferences in a way that emulates what our mom-and-pop store owner of days-gone-by could do. And further, these analytics need to intelligently and automatically act upon these preferences – in ways that our mom-and-pop store owner could not have imagined – in order to enhance each individual’s shopping experience.
What is described above is the essence of Artificial Intelligence. Creating algorithms that embody the knowledge and savvy of a human domain expert (mom-or-pop), enabling the application of such expertise to be scaled up to millions of customers in thousands of locations. This is no simple task! Surveys that ask the customer if they had a good shopping experience may be of value, but the analytics needed to emulate for our millions of customers the personalized shopping experience found in the old country store requires analytics at an entirely different level.
Like it or not, the shopping experience has become more complex and the required metrics more sophisticated.
The emergence of factors such as digitalization, omnichannel, mobile marketing, and dynamic pricing are rapidly changing the game. The availability of big data and complex analytics are providing a more granular and multidimensional view.
While there may still be a role for simple measures, the winning retailers will need to adapt to the new reality.
While I certainly believe it is very important to ask shoppers how happy or unhappy they are on an ongoing basis, this is a bit too simple for any deeper understanding. While it tells you whether the shopper is happy or not, it doesn’t do anything to help you understand the “why” behind their response, or provide any actionable insights to fix a problem. A well-constructed feedback process that provides the needed information, and allows action to be taken to resolve the negative and reward the positive, still has a place in today’s world.
It really depends on the business. I do believe we are so utterly consumed by metrics that we can lose sight of what is really important. However, working in a multifaceted industry, if customers simply say yes or not, how do we know what part of their experience they enjoyed? Was their fitting room experience up to par? Were they greeted? I believe that the OSAT covers the question of whether or not they were satisfied.
I’m reminded of the old Pete Seeger quote “Any darn fool can make something complex; it takes a genius to make something simple.”
The challenge is that any data/research you do needs to be actionable. If the only analytics you collect about the shopping journey is happy or not, what action will you take if they are happy? Just keep doing what you are doing?
What if a customer came in to your store to make a $1000 purchase, but the store layout was so confusing that the customer never found the right department. On the way out of the store, the customer discovered that you carry refreshing beverages, and had a great checkout experience with a cheerful cashier. After checkout, that shopper is likely to click “Happy” (with their $2 purchase) and you’ll never know that your store layout cost you a $1000 sale.
You need analytics that give you timely feedback about what is working and what isn’t, but those analytics need to be granular enough that you can rapidly test, measure, and evolve.
I’m all for keeping it simple, but simple still needs to be specific enough to be actionable.
Happy customers is the point of the best shopper analytics solutions! What assortment do they truly need? What pricing is fair and effective? Which promotions engage shoppers productively and which don’t? The bottom line: does my store know me and what does that mean? It means having what I want at fair prices with deals and offers that engage, excite and reward me. Finally, can I measure this with shopper level reporting? That’s what it’s all about!
Sometimes simple is better. The official definition of big data is that it’s too big. Sifting through too much data can create analysis paralyses. Good leaders know which questions to ask and what answers they want. I love the Happy Or Not product. It gives simple and direct real-time feedback. I put it right up there with the simplicity of the Net Promoter Score question. (On a scale of 1-10, what is the likelihood of you recommending us to a friend, colleague, etc.?)
Sure, I’m open to more detailed reports, but give me the big picture first. That’s what will make the big difference. Follow that up with more data and details and I’ll tweak my “big picture” as necessary.
Let me take the customer’s perspective instead of just talking about what the store gets out of it. So, obviously it is not a big deal to just ask the customer if they are happy but…outside of having a real person standing there and asking if I am happy, I get to “hit” a button.
Think about it. What if I were frustrated? With one quick hit, I can tell the store how I feel and relieve a little anxiety. Perhaps if they could also detect the force of each “hit” that would boost the value of this data even more! Ya think?
In my opinion, Paul Helman answers the first paragraph best – complex analytics fuel the customer communications quality possible in “mom and pop” stores. Truth be told, while the technology to do this well is maturing nicely, implementation is lagging at most retailers.
The Happy or Not system introduced in the second paragraph raises an additional point. Most of the “big data” retailers create is behavioral, while “Happy or Not” produces an attitudinal measure. Integrating attitude and behavior benefits customer understanding. I don’t know whether Happy or Not supports customer identification. Though this would likely alter both usage and response somewhat, it would allow meaningful integration and a clearer picture of why customers feel happy or not.
Yes they may be able to gauge shoppers satisfaction in a general way but in this technology driven age, businesses must understand what the shopper wants, what products complement that and what additional sales can be generated by comparing shoppers purchases with others to drive trial of new items or substitutions of existing patterns. None of this can be done with a “Happy Button”! Doing the analytics correctly by interpreting the data to drive customer satisfaction should result in the “Happy Button” being pushed.
Sure, I think this approach is better than nothing. However, the root cause analysis will struggle to produce any actionable insights. It just takes a bit more to create meaningful responses to shopper sentiments. The tools are out there now in the marketplace. I think investments in them will produce far more ROI than a too-simple approach.
Measuring whether customers are happy or unhappy certainly is a start, and certainly these kinds of devices can provide some quick “in-the-moment” feedback as the author suggests. Yes, you can get an indication of satisfaction from the customer on an element of your experience, but customer satisfaction can be the Trojan Horse of loyalty.
The real power of analyzing shopper behaviour starts when you can track and analyze customer transactions over time. Truly understanding how your customer interacts with your brand over time is the only way to create the framework you need to fully realize a strategy that builds engagement with your brand.
There is definitely something to be learned here – unhappy shoppers often don’t return, and communicate the unhappy experience to a number of others. Customers can feel overwhelmed with the number of surveys, social media invitations, etc. sent their way. Customers also appreciate an easy way to say they are happy that can be recorded and collated. Not quite as satisfying as a smile, but a nice step in that direction.
Shopping, particularly grocery shopping, isn’t necessarily the high point of the day for time-strapped consumers.
The challenge is to take the tedium out, make shopping less of a chore, and make it a more engaging experience. Illogical store layouts, irrelevant offers and uninspiring loyalty programs, do not a happy customer make. Continually meeting customer needs as those needs change is what makes customers happy.
And the reality for retailers is that to win in today’s increasingly intense and diverse marketplace, they need to satisfy the needs of shoppers better than the competition.
Using shopper insights based on data from loyalty programs enables retailers to make better marketing, merchandising and store decisions. At NRF’s Big Show in January I co-presented with Loblaw, Canada’s largest food retailer. Loblaw has implemented a fully digital loyalty program that creates a customized experience for shoppers through personalized offers. The results they shared at NRF are compelling. For example, program members using targeted offers increased their visits 12% year over year and enlarged their basket sizes by 5%. That’s a good indication, I think it’s fair to say, that those customers are happy.