Now more than ever, the ability to understand and predict change is a key competitive strategy. Businesses that can observe change in real-time can take both offensive and defensive action to survive (and thrive). Our simple framework will help you understand the impact of major market changes on your business and your customers.
As Charles Darwin said, “it is not the strongest of the species that survives, nor the most intelligent; it is the one most adaptable to change.”
This is true not only of biology but also of retail.
In today’s retail environment, the pace of change is higher than ever—and the dimensions of change that impact retail businesses are far-reaching. For example, climate change has impacted seasonal weather patterns, so apparel companies might notice their customers buying different types of clothing in different geographies during different times of the year. There’s greater variance in COVID-related restrictions in different states and cities around the world, so many brands with multiple brick-and-mortar sites are opening up in different locations at different times.
The world has far from stabilised, and these changes will continue into the foreseeable future. Businesses with the greatest agility and speed-to-action are the ones in the best position to expertly navigate risks or opportunities that arise following these disruptions.
A Customer Data Platform partner like Lexer can help you solve your data challenges so you can reorient your business around its greatest asset: the customer. Customer-centric businesses with predictive analytics capabilities are more resilient in times of crisis, better positioned to preserve and strengthen customer relationships, and ultimately more profitable.
In this blog, we’ll:
- Introduce you to Lexer’s simple, three-pronged approach to measuring change in consumer behaviour, and
- Provide a real-life example of the framework in action for one of our clients.
This approach to measuring change is not designed for a data scientist, but rather an in-house marketer or a business user. Its simplicity is what makes it effective and functional.
Using this framework, you can understand the business impact of major market changes, enhance the customer experience, and make smarter decisions to drive your business forward.
The 3-Pronged Framework for Measuring (and Adapting to) Change in Retail
1. Who do you include in your analysis?
First, consider the population of customers you want to analyse to understand the impact of change.
In this framework, we advocate for separating your customers into two groups: Your newly-acquired customers and your returning customers.
Newly-acquired customers are those making their first purchase during the period of change, whereas returning customers are coming back for an additional purchase.
2. What do you measure?
Once you’ve defined the customer segments you’d like to include, you need to determine which attributes or characteristics you’re going to compare.
The most common attributes we use for a change analysis are:
- Demography: This includes attributes like age, gender, and location.
- Transactional KPIs: These KPIs traditionally include metrics like average order value, average spend per product, discount rates, items per purchase, conversion rates, and more.
- Products purchased: This attribute is a really meaningful lens for tracking change and understanding buying motivations. Which products are customers buying now that they didn’t previously? Which categories are moving faster or slower than others?
- External data sources: To gain the most holistic insight into the impact of change on your customer base, include third-party data sources in your analysis too. For example, Experian’s Mosaic groups, which segment consumers based on income level, are a key attribute to consider right now due to the economic shifts occurring in the U.S. and around the world.
3. When did the change occur?
Finally, you need to set the time period for your analysis.
This is one of the most important aspects of the framework for measuring change. Because certain changes happen seasonally, you need to ensure that the trends you’re picking up are not natural seasonal trends, but directly resulting from the change you’re considering.
Be mindful of the comparative periods you look at. Typically, Lexer’s approach is to pick three periods to compare:
- Now: Choose a period to represent the “now.” The length of time representing “now” can range from this month to this year, but it should encompass the change you’re looking at.
- Pre: Once you’ve set your “now” period, you also need to pick the prior period for comparison. For example, if you set your “now” timeframe as “this month,” then the prior period would include the prior month.
- Prior: Finally, you need to set a comparative for the same time period of the prior year. For example, if you’re analysing a change that occurred in April of 2020, you’d need to compare it against April of 2019 as well.
Using these three measures of time, you can distinguish between normal seasonal fluctuations and unexpected disruptors such as major economic or political events.
An example: Measuring the impact of the COVID-19 pandemic on sales
Here’s an example of the change analysis framework we used to help a client understand the impact of the COVID-19 pandemic on their sales:
- Who we included: Newly-acquired customers vs. returning customers.
- What we measured: Total volume of orders, average order value, average spend per product, items per purchase, first product categories purchased, and household income.
- When the change occurred: The WHO officially declared COVID-19 a pandemic in March, so we included the first month following the declaration, the prior month, and the same month the prior year.
As you can see in the chart below, this brand’s sales volume took a big drop from the pre-COVID to the COVID time period. The prior year’s volume during the same time is also fairly low, so the year-over-year change is not as drastic.
Additionally, the average order value hasn’t changed, but the average items per purchase has decreased, and the average spend per product has increased. That means that customers who are being acquired during the COVID period are buying fewer, higher-value items, pointing to a meaningful change in transactional behaviour.
Knowing that newly-acquired customers were buying fewer, higher-value items during the COVID period, we looked at the first product categories purchased next to try to gain deeper insight into their buying behaviour.
As shown in the graph below, there was a significant drop in accessory purchases during the COVID period. There’s a similar (although less pronounced) drop in headwear sales. In the prior year, headwear sales grew with the change of season, so this brand would’ve expected headwear sales to increase had COVID not happened.
On the other hand, you’ll notice that the tops category grew from 5.8 per cent the prior month to 8.5 per cent during the COVID period—but the tops category from the same month in the prior year was at 9.3 per cent, so this change can be attributed to seasonality instead of a major effect of COVID.
Most notably, sales of active wear increased dramatically, from 3–4 per cent in the prior month and year to over 11 per cent during the COVID period. This threefold increase can likely be attributed to the rise of at-home workouts caused by gym closings and stay-at-home orders enacted around the world.
We also looked at returning customers during the same time period.
As you can see in the graph below, this brand had a strong decline in returning sales volume during COVID. Average order value and average spend per product dropped as well, whereas the last order discount percentage increased by more than twofold.
The movement of these metrics tells us that this brand is using discounting as a means of stimulating customer behaviour. One of the questions this brand might ask in response to these insights is: Which customers are engaging with your brand as a result of these discounts? Is discounting actually driving buying behaviour that will benefit your brand in the long term?
We strove to answer those questions by next looking at the household income of this returning segment of customers.
As you can see, customers in low-income households have dropped away during the COVID period, whereas customers from higher-income households have increased. The only reason that the percentage of higher-income customers has increased, however, is that the low-income customers have dropped—so the conclusion is not that this brand is attracting more high-income consumers, but that the distribution of income levels has reorganised within their returning customer base.
In this instance, we wondered if this brand was discounting for no reason. If high-income consumers would purchase from the brand at full-price regardless, then this brand might actually be reducing margins and diluting its brand proposition by offering additional discounts during the COVID period.
Our recommended next-step was to A/B test a range of discounts for the high-income segments. This test would help the brand understand whether or not discounting was truly stimulating buying behaviour.
Customer insights like these can be massively helpful in determining next-steps during times of great change.
The difference between detrimental change and meaningful evolution? Insight.
The retail market will continue to evolve, and with it, the behavioural patterns and preferences of your highest-value customers. Therefore, surviving in retail is not a matter of being the biggest or most resource-enabled company, but being the most adaptable to current and future trends.
Lexer’s CDP unifies your data into holistic customer profiles and provides effective tools to track change and quantify impact. With unified data and sophisticated analytics tools, you can predict and measure the impact of market changes on your business and customers. Keeping a pulse on these changes as they’re occurring gives you the ability to make smart decisions and respond as quickly as possible for the greatest chance of success.