By Chris Nichols
Three social scientists, Suk, Lee, and Lichtenstein, walk into a bar…
No joke, they do, and they observe patrons for eight weeks to watch how they order off specially designed beer menus. One beer menu has a variety of beers ordered with prices from high to low. The other beer menu has a menu of beers ordered with prices from low to high. Are the results the same? How does it compare to a randomly ordered menu? Does it really matter? And, most importantly, what happens if you did the same test for banking?
What is reference dependence theory?
The 2012 joint study mentioned above was conducted by the University of Colorado and Korea University to test something called “reference dependence theory.” In other words, it looked at the influence of price order on choices across differing brands. What researchers found out was that it does matter how brands and prices are ordered. When the menu ordered choices from the highest to lowest price, the bar owner made an additional $0.24 per beer as patrons ordered more expensive beers compared to a random menu or a menu ordered from lowest to highest price.
Does this theory work for banking?
In order to test this theory for CenterState, we walked into a bar and conducted a very similar experiment with our retail checking account offerings. We presented a represented sample of approximately 200 patrons with our checking options and asked them to choose one that would best suit their needs.
Side Research Note: At this point, we will grant you that there was no need for us to go into a bar to conduct this research, but the parallel and opportunity for expensed drinks was just too much to pass up. Besides, it was St. Patrick’s Day and you have to go to where the people are. We did our best to ensure everyone was sober when taking this test, but in the name of full disclosure, it is not like we breathalyzed anyone as that would have been awkward. We will also add that we also hope our CEO (and CFO) are too busy to read today’s post. But, we digress.
Getting back to banking. Most every bank in America that displays pricing for checking accounts—including CenterState—orders its checking accounts from cheapest to most expensive.
For our study, we showed patrons one of the following:
- The exact page as presented on our website, in low to high order
- The identical format of accounts but in random order
- The accounts in reverse order, with the highest-priced checking option first
Our first lesson was that having some order to the prices turns out to be important. If you order your accounts in seemingly random sequence, people are not only slower to make decisions (we timed the response rate) when compared to the other two options but a material portion (9%), choose to skip it even when asked to make a decision. Luckily, no bank in the industry does this, so this point is moot.
Our findings.
What we found is not only do we actually like Guinness, but account ordering matters much more than beer ordering. More importantly, our findings were the opposite of what Messrs. Suk, Lee and Lichtenstein found.
When choosing beer, from a behavioral economics standpoint, the first set of prices serves as an anchor or reference point, so that the other prices are compared to the first set. When ordered from low-to-high price, participants perceive the increase in price, with each new choice costing more. As a result, consumers essentially stop closer to the top of the list—which correlates to a choice with a lower relative price in comparison to a high-to-low order. When looking at a beer list that is ordered from high-to-low price, on the other hand, there is a perceived loss of quality as opposed to price. Participants stop a little further down on the list—but that stopping point, depending on demographic, occurs at a higher price point.
It turns out that beer is not the same as banking. When we ordered accounts from high to low, participants chose the cheaper options more often. With this order, our average test-case revenue was $7.97 per month per customer. However, when we ordered the checking accounts from low to high, our average test-case revenue shot up to $10.51, or a gain of $2.54 per account. That is a material 24% difference.
What’s going on?
While we are not psychologists (contrary to what we were saying at the end of the night after a series of Guinnesses), our thesis is that the more complicated nature of account choosing forced participants in exactly the opposite direction as choosing a beer. When ordered from low-to-high, the time it took to make a decision was 15% longer, as participants studied account attributes and based their decision on the quality of the product. When ordered from high-to-low, participants focused more on price and slid down the list at a faster rate, rendering the quickest decisions of any of our three presentations.
Conclusion—and next up.
Our findings and conclusion validate the methodology that almost every bank is already using to present its checking products, and that low-to-high price sequencing is superior compared to high-to-low. Contrary to what we might have implied, we did conduct this study with some rigor, and the demographics of our participants can be found HERE.
In the coming days and weeks, we will be presenting additional findings on what our focus group participants thought about pricing in general, about banking, and about the better way to display product information. Some of the things that we will be gathering data on and researching include:
- Whether the size of your font matters when it comes to price
- Whether pricing should be displayed first or last
- Whether products should be aligned vertically instead of horizontally
We have a long list of items to test, the results of which should make bank readers more effective.
Stay tuned. And if you are interested in similar research for your bank, just let us know where to send the bar tab.
Chris Nichols is a contributing editor to ABA Bank Marketing.com. Located in San Francisco, he is the chief strategy officer of CenterState Bank, which has its headquarters in Winter Haven, FL.
