Tapping into various types of consumer data is essential for identifying potential customers, meeting existing customers’ expectations, and ultimately gaining market share. Collecting information on consumer behavior is nothing new, but with different types of data available, making sure you are accessing the right data means the difference between a mediocre sales performance and a commanding lead of market share. To help you find the best information for your business, TraQline Content Marketing Manager Corinne Clements sits down with Eric Voyer, Vice President of Sales and Marketing at The Stevenson Company, to discuss the growing popularity of receipt data and how it compares to the information available in panel data.
CC: Eric, thank you for spending some time with us today.
EV: Sure, absolutely. I’m happy to do it.
Before we talk about the strengths and weaknesses of receipt data, would you tell us more about your background and how your role with The Stevenson Company has evolved over the 23 years you’ve been with us?
I spent about six years as a research designer and analyst working with clients to solve problems such as how to address product marketing and design challenges. I designed custom research based on the client needs and analyzed those results. We then worked together to identify data-driven solutions to their challenges. That role led me into account management and working with retailers and vendors to help them understand how TraQline could add value to their companies.
As the company has grown, there’s been a need to build a team of account executives, and today, I help manage those individuals, ensure clients get what they need out of TraQline, and continue to pursue improvements to our product.
Is that one of the reasons you began to incorporate receipt data, among other data sources, as one of the resources in TraQline?
Yes, receipt data is something that’s really come about recently because everyone has a camera on their smartphone, plus retailers are more often emailing purchase receipts to product buyers. Receipt data allows us to better augment our insight into the precise products consumers are purchasing, where they make the purchase, when they’re purchasing it, and how much they are paying for it.
Which scenarios are most ideal for using receipt data?
Receipt data works well for FMCG (fast moving consumer goods) because of large sample sizes and a very diverse retailer mix but is still in the early stages for consumer durables. There are three key scenarios that receipt data is good for: the first would be fast follow-ups. For example, you can ask a customer to take a picture of the receipt for the product they just purchased. Then you could follow up with that person in a day or two and gather a range of valuable information. One drawback is that these types of follow-ups can be expensive.
The second scenario involves getting really detailed static information about a purchase. On a receipt, you have all kinds of information — everything from the time of purchase to the actual physical address of the store to the method of payment.
And finally, you can get basket attachment rates, that is, what other items did a consumer buy when they bought your item?
Again, because of the frequency of the transactions in these markets and the resulting sample sizes collected, receipt data is a good tool for fast moving consumer goods (FMCG) and Consumer Packaged Goods – but, ensuring the data is representative for Consumer Durables is very challenging because the frequency of purchase by an individual consumer/household is low.
The type of information gleaned from receipts can be very beneficial to companies, but are there limitations to what receipt data has to offer?
There are limitations for sure. For example, it is challenging to get a representative sample size in a short period of time for products that don’t sell in the quantities and frequencies, across many retailers that CPG products sell in. This makes the sample that is available good for qualitative questions but not as much so for timely quantitative data that is representative of the whole market. Also, we’ve been enthusiastic about a time when receipt data may allow us to really dive deeper into analyzing market share. Unfortunately, despite all the information that current receipt data does provide, it isn’t representative enough and doesn’t contain the robustness of information required to accurately relay market share data.
Why can’t you get market share from the data found on receipts?
Market share is challenging to calculate from receipts because representation matters. It takes a large and sample of non-skewed receipts, balanced across the appropriate online and brick and mortar retailers, which translate into a diverse population of people – from young adults to elderly – male and female – and all parts of the country to get a representative sample of the population. Wow, that is a mouthful! As we started digging into the available receipt sources, which is exponentially larger than our competitors, we found that there still weren’t enough receipts available to get, say, freestanding range share across the country.
Retailer representation is also a huge problem. Assuming you get the issues I just mentioned solved, you’ll need to still “map” an ever-changing receipt to not just your products but to your competition’s as well. To better understand the complexity, let’s say two thousand retailers sell tires. If all sell Yokohama, every retailer’s receipt will have a different SKU and name for a specific model. You have to know what’s out there and map individual names with the product and then analyze the data across all of the retailers.
For an individual store, receipt data works great, but in aggregate, it would not be representative due to the thousands of permutations, which are constantly increasing. With TraQline, we don’t try to get to model number; we calculate at the brand and category level, which offers better control of representation. This means that when you’re calculating market share, a survey like TraQline, with over 600,000 completes a year is still the best way to capture the entire marketplace.
Considering this limitation of receipt data, is panel data a better option in some situations?
As we’ve discussed, receipt data provides valuable and detailed information about a specific purchase. But because receipt data cannot give market share or show path to purchase, it should be used primarily in the ways I mentioned earlier. Panel data, however, gives more insight into how and why people shop because it gathers data over a greater span of time and a larger cross-section. It is representative of the whole market. As a result, it offers the best opportunity to track trends and any changes in those trends over time.
What kind of insight does pairing receipt data and panel data provide?
You’ve hit on the future of consumer insights. Pairing the data provides the best of both worlds. We’re working toward a model where we use consumer panels to get behavioral trends and a complete representation of the rest of the market. Receipt data (plus information available about consumers web behavior – a topic for another day!) helps us understand the SKUs, and the brands purchased at key retailers.
How does receipts data affect the clarity of the TraQline survey?
More data, more opportunity for clarity. For years, survey data has been clarified by using both proven statistical weighting methods and improvements in technology. For instance, surveys used to be pen and paper and sent through the mail. Using technology (computers and web browsers) to collect the data improved the reach and speed. Next, skip patterns became automated. Then images and video were added. There are hundreds of examples of using nascent technology to clarify survey results and providing visibility into consumers buying behaviors through receipt data is just another iteration of this.
What inspired the TraQline team to incorporate receipt data alongside panel data?
TraQline is always on the lookout for new technologies and ways that we can continue to make it easier for respondents to answer our surveys and then also make the data more actionable on the TraQline side. Just like TraQline was the first national panel to go all online, we’re the first in our space to use a combination of these techniques.
Thanks so much for taking this time today to go over how TraQline is incorporating more receipt data into the research and analysis that it’s doing. I really appreciate it.
Great questions, Corinne! It was fun chatting with you, and I look forward to our next conversation.