The Problem with Average Drinkers

“Don’t design for average”—Rory Southerland.

Author and human behavior genius Rory Southerland doesn’t think averages are useful tools for designing human experiences.

To support his view, he describes how engineers used averages to design pilot cockpits. They carefully made the seats, controls, and gauges to suit the measurements of an “average” pilot.

As you can see from the chart above, a cockpit designed for the average man would be extremely well-suited for 68.2% of men and not too shabby for 95.4%. Great idea, right?

Well, not so much, really.

When you designed a cockpit for an average man, you were designing a cockpit not for everybody but for a surprisingly rare, or even non-existent, body type. Not a single pilot of the 4,000 measured was within the average range on all ten body measurements. (pg 99 Alchemy)

It turns out that “average” is a mathematical construction that doesn’t suit anyone, especially not a real person.

WOW! She’s really average looking.

Cockpit seats are not the only place where averaging can go awry. Interestingly, attempts to construct “average female faces” don’t produce an average-looking woman. Instead, the mathematically average face is above average in attractiveness. The chart below shows average faces derived from large samples of female faces from various countries. If we asked male college freshmen to rate these faces on an attractiveness scale from 1-10, my guess is that the freshmen would not rate them as fives.

https://petapixel.com/average-faces-of-women/

[FREE TIP FOR MY MALE READERSHIP: Although it is statistically true that a composite female face constructed by taking the mathematical average of hundreds of women’s faces is quite beautiful, it’s probably still not a good idea to tell your sweetheart that you find her extremely average-looking.]

Design for the extremes

In Rory’s opinion, restricting our ideas to ones that fit “average people” will kill great ideas. Rory is convinced that designing for a few outliers will foster the development of better products and experiences.

His idea overlaps with something Luise Perry recently wrote about.

Marketing experts group consumers into five categories: 

• Innovators

• Early Adopters

• Early Majority

• Late Majority

• Laggards

Winning over the innovators is a big deal for anyone wanting to flog a product, because this group has power over the majority. It’s a subtle kind of power, and it’s disproportionately possessed by women – hot women, in particular. That’s how social media influencers make a living: not only are they good looking (which automatically confers status on women), they also possess this mysterious talent for identifying what’s in and what’s out – a kind of enhanced sensitivity to social signals. 

It’s not average people who make or break a product or service. Instead, the path to market success must run through that extremely unaverage group of innovators and early adopters.

There’s no such thing as an average drunk

Drinking has its own form of non-normality. The average drinker hardly drinks at all during each week. It’s those few individuals on the far right-hand side of the chart that are pulling all the weight when it comes to keeping beer, wine, and spirits producers in business.

As you can see, the distribution of drinking does not make a bell curve. It’s more like a hockey stick. The technical way to say it is this:

Drinking is not distributed normally; it follows a highly skewed distribution, where a small percentage of people consume the majority of the drinks, resembling a power-law or exponential pattern. [Authored by ChatGPT]

Just imagine a recovery program designed for the average drinker who drinks between .02 and .63 drinks per week. I think you can see why such a program would be a colossal flop.

What about Resilient Recovery

I’ve seen this “don’t design for average” rule play out when writing vignettes for our Resilient Recovery meetings. In each meeting, we read a short vignette about a person with a substance abuse issue. The vignette illustrates the Biblical principle being discussed during the meeting. Most of these vignettes tell the story of a composite character—a sort of patchwork quilt comprised of bits and pieces of people who have attended Resilient Recovery meetings in the past.

We have tried two philosophies for constructing these vignettes.

• Design for the average. These vignettes were designed to be relatable for people with a variety of addictions and even those who may have a family member or friend with an addiction. Likewise, the discussion questions were as broad as possible.

• Don’t design for average. These vignettes focused solely on addicts and alcoholics, not mentioning any process addictions or other difficulties. The vignettes also describe extreme cases—homelessness, criminal record, suicidal thoughts, and children taken away by the Department of Public Safety.

We saw something very interesting when we analyzed the response to the two types of vignettes. When we designed for average, the extreme cases did not feel welcome. Or as the kids say, they did not feel “seen” by the program.

However, people with less extreme issues were very capable of making the connections between their experiences and the Extreme Cases described in the vignettes. It turns out that the less extreme attendee had the skill of extracting a general principle from the vignette and finding examples of this general principle in the specifics of their own problems. So, often high-functioning people with bipolar, depression, anxiety, or problematic substance use could benefit from a program designed for the extreme case of a chemical addiction.

Why is that?

Apart from the fact that it just works, I am not sure why designing for the Extreme Cases is more successful.

It could be that people who are Extreme Cases resemble the Innovators and Early Adopters who marketers covet. Perhaps there is a general feeling among the Resilient attendees that if a program can speak to the most extreme cases, it must know what it is doing. Thus, writing for the Extreme Cases signals competence and trustworthiness to attendees who may rightly wonder if the program can help them.

It could also be that by addressing the Extreme Cases, we overcome the extraordinary sense of shame and incompetence that the Extreme Cases feel as a result of their addiction. Perhaps these individuals require a greater-than-average assurance that they are welcome—and seeing their experiences reflected in the vignettes gives them that assurance.

My final guess is that the anchoring bias might be involved. For reasons psychologists can’t quite figure out, we rely too heavily on the first piece of information we receive when making decisions. For example, if we are asked to estimate the height of a sequoia tree, our estimate can be greatly impacted by the size of the number we recently saw. So, if we are shown a large number and then are asked to estimate the height of a sequoia tree, we estimate high because the large number “anchored” our thoughts toward a high range. The reverse is also true. If we are shown a low number and then asked to guess the height of a sequoia tree, we will guess that the tree is shorter.

Luxury car makers understand the anchoring bias and use it to their advantage. For example, Rolls Royce and Maserati stopped exhibiting their cars at car shows because a $ 400,000 car gives people medical-grade sticker shock after first seeing the cost of your average Chevy or Ford. The luxury car makers now display their automobiles at plane and yacht shows. As Rory says, “After you’ve been looking at Lear jets all afternoon, a Maserati feels like an impulse buy.”

So it might be that providing the Extreme Case in the vignettes acts as an anchor, making sharing more minor mistakes and embarrassments feel a bit safer. Or, to paraphrase Rory, “After listening to the worst-case scenarios in the vignettes, sharing my transgressions feels socially acceptable by comparison.”

Regardless of the mechanism(s) by which his rule works for recovery groups, I am convinced Rory’s admonition not to design for average is a trustworthy maxim.

NOTE: After this post was written and scheduled to be published, a Resilient volunteer sent me an email that may throw a wrench in my theory about extreme cases. She conducts Resilient meetings in group homes where up to a third of the residents are being treated for severe mental illness, not substance abuse.

It turns out that severe cases of mental illness may require extreme vignettes of the mental health variety. Talk about a beautiful theory being slain by an ugly fact. . . but that is a can of worms to open on another day.

In the meantime here's a song from someone who is far from average: