Almost a year ago I had the good fortune to get introduced to Dr. Yuping Liu-Thompkins of Loyalty Science Lab, an incredible academician, researcher and good person. Her research focuses on real world issues of more than intellectual interest, research that suggests both problems and solutions. And Dr. Liu-Thompkins, in conjunction with David King and Dr. Bonnie Holub — both of Teradata– have done it again.
This post might seem to be aimed at retailers — and to some degree it’s a wake up call for them – but I think you’ll see it’s pointing to important changes in consumer behavior that have occurred during the pandemic – changes that for the most part are likely to survive it. Their case for analytics usage — which is not the obvious case — is compelling. But I’ll let them tell you about it.
A year ago, the toilet paper section down your grocery store aisle was just a couple of boring stacked-high shelves, where words like “aloe vera” or pictures of a satisfied bear were about as exciting as it got. Who knew in a few months, it would become the talk of town, not only mentioned repeatedly in the news but also inspiring many hilarious YouTube videos.
If you work in the retail industry, however, the toilet paper shortage may not have been that funny to you. Since the start of the COVID-19 pandemic, you most likely have struggled with numerous challenges, including store shutdowns, supply chain issues, accelerated digital transformation, and changing consumer behaviors.
To tackle all these challenges, we believe smarter analytics adapted to the current environment can go a long way. We will explain why and how. But first, let’s take a look at what really changed and what remained the same in how people are buying.
Changes in Consumer Purchase Habits
The most large-scale change we’ve seen so far is the shift toward digital. The US Department of Commerce reported a 44.5% increase in e-commerce sales in Q2 of 2020 over the same period in 2019. This is really impressive considering that overall retail sales actually declined 3.6% during that same time. Today, ordering online and jumping in the car a few hours later to pick up at curbside has become immensely popular.
Besides the channel shift, consumers are also changing the brands they buy and the stores they buy from. According to McKinsey, 75% of consumers have tried a new brand during the pandemic, and 17% of consumers have shifted away from their primary stores. What drives all these movements? Product availability certainly plays a role. But research from the Loyalty Science Lab suggests that it may also be the uncertainty that many people feel during this time. To get rid of the uncomfortable feeling of uncertainty, people will try to get away from their current situation and gravitate toward new things, even in small things like what to eat or what brands to buy. This can be a problem, since long-term customers may break up with your business in search for something new. But it also offers the rare opportunity of getting a whole new line of customers that you may have never seen before.
What hasn’t changed in all this is consumers’ expectations. Regardless of supply chain disruptions and the channel shift, consumers don’t care about these back-end differences and expect the same frictionless experience. A recent survey from the Wharton Baker Retailing Center and WisePlum showed a dramatic 20-point difference in the Net Promoter Score between consumers who experienced problems with a retailer in their most recent purchase and those who did not have any problems. Unfortunately, 66% of the consumers surveyed said they experienced problems.
Tackling Customer Challenges Through Smarter Analytics
The numbers above tell us that many retailers still need to play catch-up to the new reality. The question is how. We think part of the answer lies in the smarter use of customer analytics. When so many transactions have shifted online, clickstream will become much more critical, and e-commerce related analytics will experience a meteoric rise. Analytics is no longer something nice to have but will be the primary portal for engaging with customers.
As consumers zig and zag across channels for even a single transaction, the need for data integration is more critical than ever. It is simply not possible to create a seamless customer experience without having connected data. So data integration across different platforms will be a high priority during this turbulent time. Think about the example of a consumer who orders online, picks up at curbside, and then calls customer service to report a problem. As the retailer, you need to connect the data from all these touchpoints to form a holistic view of the customer.
This integration is challenging for retailers late to the digital retailing game because of traditional data silos, such as separate inventory management systems at the distribution centers versus in stores. In some cases, even though consumers can clearly see a product being available in stores, they still have to wait days for the product to be separately shipped to the store for them to pick up later, simply because it’s an online order. You can imagine the consequence: customer frustration at the very least and disappearing customers at the worst.
If you are one of those lucky retailers already with integrated data, the next challenge is likely to tackle the tremendous quantities of data available. This will require advanced analytics at speed and scale so that you can leverage your data to create hyper-segmentation and personalized communications. Traditional customer segmentation often generate a few large static customer groups. In contrast, hyper-segmentation dynamically generates a large number of very small groups of customers based on similarities in behaviors at a given point in time. The best-of-breed retailers today are able to train their machine learning models on millions of observations, and to score more than 250 million observations multiple times per day. Such capabilities allow these retailers to personalize their communications to hundreds of thousands of most valuable buyer hyper-segments in real-time.
One really thorny issue in customer analytics will be making forecasts. Projections are never certain, but now more so than ever. You need to keep in mind that most forecasts are like looking into the rearview mirror of a car. They predict what will happen based on what happened in the past. For example, your previous forecasts may take into account the fact that purchases in some product categories always surge during certain times of the year. In the new reality, these existing models are probably no longer valid because consumer behaviors have changed so much. It is necessary to re-examine your existing predictive models and make adjustments based on the changing customer purchase habits we talked about earlier.
Another necessary adaptation is to re-map out customer journeys that have emerged during this time. For example, many parents are now buying groceries remotely online and have them delivered directly to their kids. In-person social gatherings have also turned to virtual parties, changing the nature of shopping for these gatherings. Your customer analytic models need to recognize and take into account these social customer journeys (customer journeys that involve others’ influences in the decision-making process) and other new journey patterns.
Even with these adjustments, it will be flawed to forecast into the future (e.g., to the year 2021 or 2022) based on the current reality. No one can predict when the pandemic is going to be over, but the second wave of infections across the world suggests that the end is probably not near. The high level of uncertainty associated with the pandemic means that the current behavioral patterns may or may not be applicable in a year or two. So we are likely to see new algorithms and proxies emerge. These new algorithms and proxies need to address what it will be like when things get back to “normal” or the “new normal”, when hopefully toilet papers will just be boring toilet papers again.
Thank you. Hey, retailers — and pretty much a lot of others — get on the case. Now.
Announcements
The CRM Playaz are having a Watch Party for Marc Benioff’s Keynote at Dreamforce To You on December 2 at 1pm to 3pm ET. Not only will we be commenting on the speech in a Mystery Science Theater 3000 kind of way, but we will have Brian Solis, Salesforce’s Global Innovation Evangelist, thought leader, analyst, digital anthropologist, author, and – most importantly – an alumnus of the CRM Playaz Playa in Residence program, on to answer questions and respond to comments from you. Then, analysts from the business technology industry will discuss their take on the speech and direction of Salesforce. If you are interested in registering, click here and sign up as 400 of your brethren have already done. .
Also, watch for announcements shortly on BYOB 2020, Playaz Productions Battle of the Tech Company Bands. BYOB 2020 stands for Bring Your Own Band 2020, BTW. If you are interested in some preliminary information, come to this page and from here you can begin to see what this is all about. Please keep in mind that both the site and the band pages are a work in progress so be merciful. A few of the bands don’t have their pages up yet. But the bands, representing (in alphabetical order) are all in:
We’ll see you all at the watch party!
