As we all know, COVID accelerated several trends in retail. Notably, last-mile technology and omnichannel investments as a hedge against the behavior shift to less frequent in-store visits with bigger baskets. As a result, many retail executives have commissioned projects with the objective of "increasing shopper frequency."
However, the issue is shopper frequency and basket size have an inverse relationship. The more you shop, the less you need, and the less you shop, the more you need. For retailers who have padded COVID revenues with bigger baskets, the question now is, how do you increase frequency without negatively impacting your gains in average basket size?
The ability to increase shopper frequency without negatively impacting the average basket size will be an elusive journey for most retailers. Possible, but elusive. Why? Some won't have the data required to support this granular objective. Others have the data, but competing priorities will limit their ability to invest in the necessary data preparedness to solve for frequency.
Conversely, even the most innovative retailers with pristine data pipelines can fail to properly frame this business problem as an analytics objective and unravel any limiting assumptions before jumping to an ideal solution. It is easy to say that people are shopping less frequently but harder to substantiate the why behind the broader issue.
Pricing can be a powerful tool to build long-term price perception, indirectly increase shopper frequency, regain market share, and drive profit. But in the short-term, personalized offers based on customer history is a direct way to boost both basket size and frequency. However, you need the data to anticipate your customers’ needs.
Recommending a customer something they are likely to buy based on evidence found in their data creates a huge upside for basket building and increasing purchase frequency.
For example, I recently received a push notification from Amazon recommending a bathroom faucet. A few weeks ago I ordered a new vanity set and showerhead. I hadn't thought to replace the sink faucet until Amazon suggested it. I took the bait for two reasons. The product recommendation was relevant, and it uncovered a need and even solved a problem that I hadn't yet realized. I even added a few other items into my cart at checkout.
You can do the same thing for your shoppers. More than ever before, we can connect the dots across a broad array of internal and external data sets to enable deep, granular insights into the shopper's journey. Our digital economy's rapid advancement only exacerbates your customer's digital footprints, equipping retailers with premium fuel for frequency-focused personalization initiatives.
According to a recent study commission by Forbes, 33% of executives using personalization have seen increases in transaction frequency, increase sales, and boosted customer lifetime value. It’s pretty amazing to think that with a little scientific rigor, we can now make 1:1 offers at scale to ensure each customer experience is unique, relevant, and a key differentiator between you and the competitive landscape.
Let’s assume you don't have the time or the budget to implement a new solution, but you still need to solve for frequency. To avoid doing things in the wrong order, we need to first formulate a data-driven perspective on why shopper frequency is important. From there, we can outline some common problem-solving obstacles and how to translate our business problem into a data-driven objective.
In retail, even the most temporary and anomalous impact on total sales revenue can lead to emotional and reactive decisions that negatively affect employee bonuses, executive compensation and even long-term strategic vision. The truth is, frequency has many external factors that just aren't within a retailer's control.
An essential step in translating the business problem into data-driven objective is to gather functional experts across the business and IT department and collect as many data facts as you can about a particular customer. The below list is a start, but not exhaustive:
Once you've gleaned what you can from your internal data assets, think broader.
For this exercise, I would recommend you start with an employee (with their permission) who frequently shops with their employee discount. For two reasons - with an employee, you already know some of the answers, and they can validate the data.
However, you want your data to do all the talking. See what level of effort it takes to find or infer the answers to these questions from your data. Measure how long each step takes, the number of disparate data systems, the number of team members required to pool the data, servers, databases, and the necessary number of tables to extract the data.
Be sure to document your data pulls in a programmatically repeatable way. The key is to start small, automate, and then validate as you scale your analysis. The aim is to align the business experts with your data and infrastructure experts while translating the business problem into data-driven solutions.
You are probably thinking this will take a long time, but innovation takes time, data, and a spark of imagination. Imagine what is possible when you can build a 360-degree view of your customers' journey. For those short on time and creativity, I've found some literature that suggests using customer surveys as another option to understanding dwindling shopper frequency. However, have you heard the saying, "There are no shortcuts in life?" In my opinion, the same applies to retail science.
Consumer surveys ask questions to a select subset of shoppers to estimate characteristics about all shoppers. By definition, it is a shortcut. With this approach alone, retailers run the risk of leveraging expensive and poorly constructed surveys. That leads to inconclusive decisions based on exercises of coincidence without any actual causation.
Whenever possible, we want to avoid data-driven shortcuts. Often, these surveys struggle to account for inherent bias. Making decisions on aggregate data with simple averages presents a significant flaw as pricing is executed at the item level. It's vital to point out that actual transactional data is more indicative than survey data because it illustrates actual consumer behavior and not what a subset of selected consumers think.
Insights that drive frequency decisions should be pulled from item-level transactions through observations and experiments. When surveys are used, they should be in conjunction with transaction-level analytics since shopper frequency is a lower-level metric. This means that we should go to the lowest level of granularity to identify metrics that, when modeled, help us to achieve our objective through actionable insights.
In business, all decisions should be backed by data, not just the readily available data. Sometimes we have to dig a little deeper to make sure everyone wins.
Maybe you’ve already put the cart before the data horse, or taken some survey shortcuts. You wouldn’t be the only ones. Here at Revionics, we've helped global retailers across multiple verticals to realize that increasing shopper frequency should be an outcome, not a solitary objective. The true objective is building on your relationship with the customer and leveraging internal and external data to gain a 360-degree view.
When you understand the purpose of their visit, you can relate to their underlying needs and drive both frequency and basket size. Fortunately, you do not have to reinvent the wheel. We've done the work and can provide a reproducible modeling framework to help you know your customers and gain confidence in your decisions surrounding frequency-based initiatives. Learn more with these Pricing Strategies to Drive Shopper Frequency, or reach out today to talk to one of our pricing experts.