Perspectives Blog

At Helpfully, one of our favorite things about research is that it challenges and often upends assumptions. One assumption we see again and again, is the assumption that customers are totally rational when they shop. Taking it a step further, people are driven purely by rationality. This is false. 

‍

So many things drive human experience and human behavior. Humans don’t exist just in their minds. Humans exist in a lot of ways. We love exploring these ways using different lenses to help us understand why people do what they do, what drives them, what makes them tick, what’s important to them. As researchers, we try to understand people and their behavior using lenses from anthropology, sociology, behavioral economics, etc. 

‍

‍

A story from field work….

This assumption that customers shop based purely on rationale came out during one of our research projects with a grocery store chain in the Midwest. We were tasked with researching customer behaviors, focusing on grocery lists, to aid the grocery store chain in developing a tool or experience to assist shoppers with their lists. Before we began researching, we discovered these assumptions at play: 

‍

1. People put the real thing they’re going to buy on their grocery list. 
2. People shop based on recipes.
3. A grocery list is used to shop quickly and efficiently. 

‍

With these assumptions in mind, we dug into the research. 

‍

‍

Insight 1: a list has a purpose 

We used two methods to conduct our research. The grocery store chain already had a ton of data from all their stores, registers, receipts, etc. There was so much deep insight, so much segmentation, showcasing the different kinds of shoppers, different kinds of families, etc. They had the data on how long people spent in stores, at the register, etc. Along with this very granular data, we had shoppers submit their grocery lists. We got screenshots, pieces of paper, apps, post-it notes, wire-bound notebook, scraps of paper, whatever! Some lists were very organized, very consistent, and with check marks. Other lists included personal notes and reminders in addition to the grocery items. 

‍

We had the opportunity to visit people at home, watch them shop, and observe them write their lists. After this field work, we would follow up with them to see what they ended up buying. We believed there was a connection between what shoppers want, what they put on their list, and what they actually shop for.

‍

What we found was a total surprise. 

‍

It was shocking the number of people who threw away their list as soon as they entered the store. We thought this was a sacred document guiding a shopping trip. When asked, people explained the list was more of a planning tool at home. It was a tool to think about, “'What could I cook for dinner this week? What are the things that I like? What do I want to try?” But once they got to the store, they figured, “I’ve memorized the layout, I know where everything is, I don’t need this piece of paper.” The list is tossed aside and instead shoppers buy based on memory or their mood at that moment. 

‍

Many people didn’t need or want to shop as efficiently as possible. This insight led to a shift in the design of the tool. It became less about efficiency and more about a serendipitous exploration of the store. 

‍

‍

Insight 2: a list is personal 

So much of food and eating is personal, cultural, location based. People wrote the most personally relevant things on these lists. Instead of writing Cheerios, they might write, “Jessi,” with Jessi might be the member of their household who loves Cheerios. Many people in the Southeast just wrote, “salad fixings.” They didn’t write every item. They didn’t need to, because for their home they know what “salad fixings” means. Interpreting this kind of personal data can be difficult. 

‍

‍

Insight 3: caution when interpreting data using AI

Trying to interpret this kind of household “slang” is difficult for researchers and even harder for AI or big data science tools. The people writing ‘salad fixings’ know what that means for their households, but AI won’t. There are so many benefits to using AI and big data science machines to interpret tons of quantitative data, but there is also a danger. Sitting in the cockpit with tons of data, it can be tempting to assume an understanding of what the data means and based on those assumptions build products and experiences that aren’t solving the right problem. On the other hand, interacting with participants and gathering more qualitative data - there is a temptation to interpret their slang and culture at best, incorrectly, and at worst, in ways that are terribly misleading. 

‍

‍

Conclusion: 

We love the tension between qualitative and quantitative research methodologies. Despite the benefits of each approach, it's essential to engage in fieldwork and directly interact with participants to gain deeper insights. Some of our deepest insights for this project came from shopping alongside people and seeing how they used their grocery list. The value lies in finding a balance between qualitative and quantitative data, recognizing the importance of both perspectives in understanding complex human behaviors and experiences.

‍