Glasses and police lineups: Leveraging comparisons for good and evil

How Warby Parker’s strategy will drive sales

I’ve been in the market for a new pair of glasses for a few weeks now, so when my buddy Adam told me about a relatively new company called Warby Parker that sells glasses, I jumped aboard.What I didn’t expect was for the behavioral scientist in me to get involved!

In case you don’t know, Warby Parker has an interesting sales strategy. Because they have few physical stores, most everything’s online. A potential customer picks out five pairs of glasses that interest him or her and has them shipped to the home. The customer then has five days to wear the glasses around and hopefully decides, sooner or later, to purchase one (or more) of the pairs.

I estimate that Warby Parker enjoys a very high transaction rate for those customers who have glasses shipped to their homes. It’s not the marketing literature that makes this suggestion, though, but rather the forensic psychology literature — specifically, the work on eyewitness memory. Here’s why.

Simultaneous and sequential lineups

Researchers have long known that similarity plays an important role in both true and false identifications in a police lineup. To make a long story short, if you see an individual commit a crime, chances are good you’ll be able to identify that individual successfully in a later lineup. However, if you replaced the true culprit in the lineup with someone who looked very similar, you would also have a high chance of identifying that innocent individual, too (what we cognitive psychologists call a “false alarm”). Researchers have found that the individual that looks most like the perpetrator is the most likely to be recognized. Importantly, eyewitnesses, upon given a lineup, make relative comparisons among individuals in the lineup and select the best match (and do so with high levels of subjective confidence).

Obviously, the criminal justice system is very interested in maximizing the number of correct identifications and minimizing false alarms. So some researchers have suggested that a better way to administer lineups — so that judgments are made in an absolute fashion, rather than a relative one — is to show a witness face after face, one at a time, for possible identification. The hitch is that once an individual identifies a face, he or she may not see any more faces. Imagine this difference and you’ll realize that this strategy likely decreases both the number of false identifications and eyewitness confidence.

(Does this method actually lead to improved discrimination, however? Psychologists disagree.)

The power of relative judgments

So how does this apply to picking out a pair of glasses? Perhaps this image illustrates at least one similarity.

When a customer receives a box of five glasses, just like in a lineup, he or she is encouraged to make relative, rather than absolute, judgments. The question asked by the customer is “Which of these five frames look best on me?” and not, “Which of these five frames make me look good?” or even “Which of these five frames make me look better than other glasses out there might?” Just like an individual in a police lineup is singled out for being the best of the given options, a pair of glasses is purchased eventually because it is the best of the given options.

It should be clear that we should not send an innocent man to jail because he’s the best approximation of a murderer or thief. Likewise, we probably shouldn’t purchase a pair of glasses if they only approximate attractiveness.

So the moral of the story is clear: When purchasing from Warby Parker (or any similar company), the question is never, “Which of these glasses should I buy?” but rather “Should I buy one of these glasses?” Don’t let the power of of relative judgments take you for a ride.