Your Brand’s Own Data is the Missing Link for Consumer Insights

Original post (here)

Earlier this year, a team of data scientists got their hands on a dream dataset. The cache contained a huge amount of data points about consumers’, well, consumption habits: millions of time-stamped orders through the delivery service Postmates. Using this data, the analysts were able to uncover previously hidden patterns about how, what and when consumers get their fix.

When do people most crave chocolate? (Wednesdays) What’s the busiest day for Bud Light delivery orders? (The 4th of July, natch.) What do consumers most seek out in the winter? (Chicken noodle soup and Emergen-C.)

These findings are a goldmine for food and CPG brands that want to better understand their customers’ habits and preferences. But in many ways it’s just the tip of the delivery-on-demand iceberg. Despite its huge storehouse of customer ordering data, Postmates only has access to a fraction of the data points necessary to get a true 360-degree view of consumer preferences for the products and brands they peddle.

Where’s the rest? Public online data. Consumers may order their favorite food, drinks and CPG items directly through Postmates, but they talk about their general preferences on social media, blogs, forums and more.

What can we learn by combining these two invaluable datasets? Quite a bit, it turns out. We analyzed millions of consumer conversations about the top products consumers order for delivery, and uncovered the three critical insights:

  • Top orders by hour
  • Biggest alcohol-consuming cities
  • Most popular restaurants by region

Top Orders by Hour

One of the most interesting findings from the original Postmates dataset was consumers’ most-ordered items at various points throughout the day — doughnuts and coffee in the morning, sandwiches and MacBooks (?) in the afternoon, burritos and booze in the evenings, and Advil, energy drinks and Nyquil late at night.

Are these patterns similar in the general online conversation? It turns out there are a lot of overlaps.

The differences in what people eat and drink throughout the day is reflected in the discussion about what to order. We divided the day into the following categories: breakfast, lunch, dinner, early party hours, and late party hours.

Cravings change significantly throughout the day. During breakfast hours, consumers tend to discuss healthier foods, like organic meals. They also discuss ordering flowers and classic breakfast foods like coffee and donuts. The lunchtime discussion generated the most overall conversation, as people may be ordering lunch for their workplaces. Chinese food and pizza were the most popular options. During dinnertime, people order pizza, McDonald’s, chicken dishes, and burritos. During early party hours people order Chinese food, pizza, and alcoholic drinks like wine and beer to get the party started. The late party hours generate the second highest conversation levels. Drunk and in need of food, people take to their apps to order McDonald’s, ice cream, and Thai food. For non-partiers, or during the week, people order groceries, opting for apps instead of in-store shopping.

Looking at order data from Postmates, one can see the results: what consumers ordered. But the ordering process is not that straightforward; people do not select their restaurants and food or drink randomly. What a consumer orders is dependent on their location, the time of the day, the situation, peer influences, past experiences with the restaurant and the food or drink, willingness to try something new, and more. There are many discussions that take place, many factors that lead to the order. Using online data can help inform brands the consumer’s thought process and purchase journey.

While many of the items people order are foods, a significant part of the app and online ordering conversation is alcoholic drinks, a popular option for Postmates delivery (for people over 21). Plus, with liquor delivery apps like Drizly and Saucey, it has become easier than ever for consumers to find their favorite liquors and have them delivered to instantly.

What can the social media conversation tell us about consumer preferences about alcohol?

Blame It on the Alcohol

There are certain cities that are more known for alcohol consumption than others. Napa Valley, California, is home to expansive wineries. Portland, Maine, contains breweries galore. Does established drinking culture correlate with consumers’ alcohol ordering habits?

Alcohol preferences differ geographically. Some states, like Tennessee and Vermont, strongly prefer ordering beer. In fact, more states are beer states than wine states. Other states, like New Hampshire and Montana, strongly prefer ordering wine.

Bu alcohol analysis is not the only interesting geographic data. Looking at select cities, different restaurant preferences emerge.

What Reigns Supreme in Regional Cuisine?

We looked at the ordering conversation in Los Angeles, California; Phoenix, Arizona; New York, New York; and Houston, Texas to better understand the food ordering habits of consumers in different regions.

Perhaps it is unsurprising that McDonald’s is a hit in all cities, making up at least 40 percent of the conversation for all. But there are regional preferences. In both Los Angeles and Phoenix, Chipotle is the second most popular option. In Houston, Chipotle ranks fourth. In Los Angeles, California burger chain In-N-Out received 16 percent share of voice. Local favorites, Roscoe’s House of Chicken and Waffles, and sushi restaurant, Sugarfish, rounded out the top five.

In Phoenix, the most discussed orders are chain restaurants. Whataburger, a Southern burger chain, had 4 percent share of voice. Burger King had half the conversation compared to Whataburger. In New York, burger chain Shake Shack was second to McDonald’s, making up 15 percent share of voice. Regional favorites The Halal Guys made up 2 percent share of voice and Katz’s Delicatessen made up 1 percent share of voice. Healthy vegan chain by CHLOE. Was also in high demand, with 1 percent share of voice. In Houston, burger chain Whataburger, which was also popular in Phoenix, had 24 percent share of voice. Donut chain Krispy Kreme is also a favorite among Houstonians, with 17 percent share of voice. Chipotle followed Krispy Kreme, then regional Mexican restaurant Torchy Tacos made up 2 percent share of voice.

The data demonstrates that people discuss large chains the most, but regional favorites appear.

Conclusion

Online data fills the gap needed for a comprehensive understanding of consumers’ app and online ordering habits. While company ordering data is valuable, its value increases when you pair it with unfiltered consumer conversations. The online conversations reveal more than just their actions: they include the discussions that happen as consumers try to decide what to order and when, their experience using the app, what they are thinking as they wait for their order to arrive, and the food itself. By looking at the online conversations, the picture is so much clearer.

Request a Demo

Ready to transform your business?

Get a walkthrough of Crimson Hexagon and learn how consumer insights can help you make better business decisions.