5 Common Mistakes Marketers Make When Analyzing Shopper Data

Stephen Marosi, 84.51°
Current average rating    
Key Takeaways

​What? Analyzing shopper data can be tricky if careful attention is not paid to details.

So what? There are some common mistakes in analyzing data, including data overreach, loss of humanity and ambiguous context.

Now what? Carefully avoiding common mistakes will lead to smarter company decisions that reflect customer nuances.

​Nov. 28, 2017

Successful data analysis requires both careful attention to detail and vision for maximizing insights

In some ways, analyzing shoppers is not much different from analyzing product movement, manufacturing quality, advertising expenditures or weather patterns. Data is data, right? Yes, but shopper data comes with its own set of common mistakes. Here are five mistakes to avoid:

1. Data Overreach

Since some shopper data can be less abundant than other sources, such as shipments or scans, it’s tempting to stretch what you have beyond the bounds of good analytics. A generous sample size can shrink into statistical insignificance as you slice and dice by family status, ethnicity, age, household income, basket size, geography, specific product size, etc. The danger lies in extrapolating special cases into broad conclusions.

Another overreach is to combine trends from multiple shopper data sources collected at vastly different levels of granularity; for example, you might try to tie a national trend from a syndicated source with a unique shopper group at your retailer. These shoppers aren’t the same, however, and there’s really no way to confidently mix them without risking false correlation. It’s best to know the limitations of the data and be satisfied with more modest (but accurate) conclusions.

2. Convenience Over Correctness

Sometimes it’s more convenient to reuse data from another analysis because it’s already there and you can avoid the hassle of querying your shopper system. But do you really know how the data was pulled? Are you confident it applies to your business question? Your reduced effort may lead to wrong (and costly) decisions for your company.

Another blunder is to “save time” by relying too heavily on one data source over all others—even if it’s shopper data. Analysis may show that shoppers are trending down for your brand at a particular retailer. Does this signal declining interest in your brand and the need for a revamp? Maybe. But store operations data might reveal other drivers like insufficient ordering or rising out-of-stock rates. For a complete picture, take time to draw insights from all relevant sources.

3. Loss of Humanity

As you crunch the numbers, it’s easy to forget that shoppers are not data points but willful humans who can surprise you. People have a way of slipping in and out of neat, non-overlapping shopper segments. Millennials don’t all think the same way. Hispanic families cannot be lumped into one segment. Shoppers are not algorithms or rules of thumb. Focus solely on numbers, and you’ll miss important causes and deeper consumer needs. 

 

 How stores track your shopping behavior | Ray Burke | TEDxIndianapolis

 

As an example, consider a study of price points and elasticities for a handful of frequently shopped-for items. The analysis recommends cutting prices by 15% to attract more price-sensitive shoppers. After price reductions, however, market response disappoints. Why? Chatting with a few dozen price-sensitive shoppers reveals something you won’t find in the numbers alone: In their quest to stretch limited budgets, these shoppers feel threatened by what they perceive as a win/lose game where the retailer tries to entice them to overspend. They experience anxiety just walking into a store, feeling that the retailer doesn’t have their best interests in mind. For many of these shoppers, visiting a no-frills discounter presents a lower-stress shopping experience because they see everything as transparent and full of empathy for their plight. Interviews yield phrases like “on my side” and “outsmart the system” and “no stress.” With this better understanding of the problem, it’s obvious that reducing prices on key items must be only one part of a larger strategy to win over these shoppers.

Attitudinal, qualitative insights (aka “thick data”) can be alarming for the decision maker accustomed to orderly columns of figures and statistical significance, but they give power to answer questions that won’t yield to purely numeric analysis.

4. Insight Tack-on

Have you ever sprinkled shopper insights onto an otherwise finished analysis? Your audience expects some shopper perspective, so you toss it in, possibly ignoring data points that inconveniently contradict your previous analysis. This, of course, is backward and hazardous. Do you really want to expand a product beyond its test market without knowing who is buying and how they compare to shoppers in other markets? To avoid market catastrophes, discipline yourself to begin and end your analysis with the shopper: her needs, wants, attitudes, behaviors, trends and exceptions. Shoppers are giving you their data, so you’d be advised to listen to it.

5. Ambiguous Context

The scope, assumptions, and even vocabulary you use must be crystal clear to your audience. While this is true for any analysis, non-standard shopper terminology and widely differing segment definitions can be especially troublesome. For example, you might find that a wine brand selling at $5.99 over-engages with the lowest income quintile of wine shoppers. But these “lower income” wine shoppers might be in the upper income quintile when compared to your total universe of shoppers across all categories.

The solution is to annotate your tables and charts unambiguously and completely—label everything so another analyst with the same data can reproduce your work. Your audience will thank you as they confidently interpret your insights and make informed decisions.

Yes, data is data—but taking special care to avoid these common mistakes will yield smarter decisions for you and your company.


Related Articles:
AMA PCM Digital Marketing Exam The Business of Actionable Shopper Insights What Nuances Are You Missing With Big Data Segmentation?

AMA PCM Digital Marketing Exam

The Business of Actionable Shopper Insights

What Nuances Are You Missing With Big Data Segmentation?


Author Bio:

 
Stephen Marosi, 84.51°
Stephen Marosi, director of client advisory at 84.51°
Add A Comment :
 

Become a Member
Access our innovative members-only resources and tools to further your marketing practice.