How Data Science Beat Human Mind?
A lesson in behavioral economics will tell you that consumers don’t always practice what they preach. So what better reason is there for product marketers to make testing a key phase in their product lifecycle?
Unpeel the Facts
Consider the unique predicament of KitchenApplianceCowboy and his Super Peeler.
Q: Dear SuperAnalyticsMan, Our new and unique peeler was meant to revolutionize the world! Okay, maybe just the kitchen. It’s easy-on-the-hands, ergonomic and efficient design was the result of a feedback-based redesign project. Not only did it look better, it even felt and peeled better. The focus groups we exposed the product to loved it so much that they didn’t seem to care that it was priced $5 more than its predecessor. I was fast adrift on cloud nine.
Fast-forward to now. It’s been two months since the new peeler hit the stores, and I learned that Super Peeler sales are turning out to be worse than its notorious predecessor. The consumers who promised to adorn their kitchen drawers with this revolutionary gadget didn’t show up to purchase it! What’s happening?
Sincerely,
KitchenApplianceCowboy
A: Dear KitchenApplianceCowboy,
First, SuperAnalyticsMan doesn’t exist. So you’d do well to empower yourself with some data science prowess.
Second, studies have shown that the way we respond in a buying situation is often quite different from how we say we would. Our rational mind is not as much part of the decision-making process as we like to think it is!
Sincerely,
TheDataScienceHelper
Testing — The Anchor that Holds
In a recent experiment, I and my co-trainer conducted with a small group of participants at Aryng’s Data Literacy workshop, we held up a state-of-the-art projector in our conference room and asked all the participants to answer three questions on a piece of paper:
- What are the last 3 digits of your social security number (SSN)?
- Would you be willing to pay that amount (answer to question 1) for the projector (Y/N)?
- What is the maximum amount you are willing to pay for the projector?
The findings are shown in the figure below. There was an uncanny and direct correlation between their SSN and the amount they were willing to pay for the projector. Imagine that! Although the social security number is a random number that bears no relationship to the participant’s wealth, behavior or any other attributes, those who happened to have high SSNs were willing to pay more for the projector, and vice-versa. How would you explain that?
Noted psychologist and eminent behavioral economist Daniel Kahneman describes this phenomenon as anchoring — a psychological heuristic that influences the way people choose by comparing to a nearby reference point. In our experiment, the participants “anchored” to their SSNs, and it influenced the price they were willing to pay for the projector. In common speak, consumers may say something in a focus group that is completely contradictory to their actual response in the field (under the influence of an unknown anchor!).
Super Peeler — Revisited
Now, back to the Super Peeler. I take a stroll through the kitchen aisle at a neighborhood store and find that the Super Peeler is co-shelved with some of its cheapest competitors and other kitchen gadgets half its price. Given that a prospective buyer is not immediately aware of its superiority relative to others, it does look a tad bit on the expensive side. Could this inventory placement and/or relative pricing explain the revenue dip? Possibly.
Moral of the Story
Consumers often aren’t conscious of what they want, what they don’t want and how they would react in a particular situation. So, FIELD TEST. Test your hypothesis with a sample of your customers before you roll out to the entire population. One way to test would be to roll out in select stores for a short period and study what happens. If it doesn’t sell as expected, test your next hypothesis by moving the placement of the peeler (store’s merchandising layout permitting) or changing the price (even if temporarily). With testing, you will be able to mitigate the risk of full product roll-out while still knowing with much greater certainty whether the new product at the new price with new looks is a go or a no-go.
To summarize, if you want to understand what your customers want, confirm your hypothesis by asking them but “know it” by testing it with them.
