Jul 1, 2019

Don't Forget the Condiments

Woman putting condiments on burger
While people may not forget the condiments when they host a summer barbecue, new research from Numerator shows they clearly forget when they’ve purchased them – highlighting yet again the need to conduct survey research with verified purchasers. Without verified purchasers, you’re seeing as much as half of your study’s budget get, well, flame broiled (and not in a good way).

It must be summer, as emails about the joys of cooking outdoors are filling my inbox. Hardware stores want me to buy a new grill (no thanks, I actually bought one from your store three years ago and am not taking up grill collecting as a hobby) and related accessories. Apparel retailers are coming at me with clothing specific to the barbecue - which I didn’t know I needed, like an idiot I’ve been cooking in just my everyday shirt / shorts combo. Though in fairness I did find my cursor hovering precariously above the buy button for a “Grilling Like a Villain” t-shirt the other day which, alas, I did not purchase.

And I get it, dinner from the grill, eaten on our deck, is the one meal my entire family can agree on – no arguing, no phones at the table, everyone willing to help set up and clean up. Though I fear as a family we may have gone a bridge too far, my kids have now decided that when we grill the hot dog can be counted as a vegetable. To which I say who knows what’s actually in a red hot, let’s just roll with that. In short, our most memorable meals each summer are not from a restaurant, but instead are from the grill.

But you know what turns out not to be memorable? Condiments. I mean that with no disrespect, as a lifelong fan of ketchup in particular. We clearly miss them when they’re not there, like electricity and oxygen. But beyond that we pretty much take condiments for granted.

You know what else isn’t that memorable? Buying condiments. Or, to ensure I’m not just picking on condiments, purchasing pretty much any pantry staple. But let’s stick with condiments, to keep with the summer grilling theme.

Numerator recently conducted a survey where we asked members of our purchase panel to recall which condiments they’d bought in the past one, three, six and twelve months. We then compared those claimed purchases with what we’ve actually recorded based on their verified (meaning they submitted a receipt) purchases during the same time frame.

What we saw was, not surprisingly, the incidence of claimed purchase and verified purchase were most aligned at the one month (because of recency) and twelve month (because in general shoppers know what their households consume) mark. But even at the one month mark, for a product like relish claimed purchase was over 40% (4 percentage points) higher than verified purchase. And the six month time frame, arguably the most frequently asked recall duration in survey research, was easily the least aligned in terms of incidence.

Maybe even more concerning was the accuracy of claimed purchases. Even though at times claimed and actual purchase incidence were relatively well aligned, the accuracy of people’s recall – meaning someone both said a product was purchased during a given time frame and a receipt verified that fact – was as low as 50%. This means companies are wasting as much as half their project budget by surveying people who shouldn’t qualify for a study.


Oh, and if you want a beer with your burger, that creates a different pattern of mismatch entirely, the only category we looked at where the number of people claiming purchase during a particular time frame was consistently and substantially less than verified purchases – potentially suggesting with a product like alcohol there’s a psychological bias at play as well, where people are reluctant to admit how frequently they buy it.


These results highlight yet again the need to conduct survey research with verified purchasers. From a screening standpoint, the main benefit of talking to verified purchasers is clear – only using verified purchasers ensures you are talking to the right respondents. The oft overlooked benefit is that even when not used as a screening variable, verified purchase behavior can be used to profile respondents, resulting in shorter surveys and more accurate analyses.

Interested in grilling up some survey research with verified purchasers? Drop us a note. That's a summer email we'd enjoy receiving.