The 21st Variable

 

There is an old saying in management that what gets measured gets managed. For years, I have preferred a slightly different version: what gets measured gets valued.

Organizations measure the things they consider important. Revenue, expenses, productivity, customer satisfaction, employee turnover, wait times, error rates, sales per square foot, cost per transaction—once an organization becomes large enough, numbers become essential. A small business owner can stand in a dining room and see what is happening. The CEO of a company with 20,000 employees cannot.

But there is another half to the principle that is easy to miss:

What gets measured gets valued, but what does not get measured becomes harder to perceive.

That does not mean organizations deliberately ignore everything they cannot quantify. It means that scale creates a sensory problem. The larger an organization becomes, the less its decision-makers can experience it directly. Eventually they must rely on representations of reality: reports, surveys, dashboards, key performance indicators and spreadsheets.

Those representations can be extremely accurate.

They can also be incomplete.

And sometimes what gets left out is the very thing people loved.

Something About the Place Changed

Most people have probably experienced this.

There was a restaurant you loved. Or a shopping mall. A tavern. An arcade. A park. A school. A workplace. Maybe even an entire neighborhood.

Then something changed.

Perhaps you return after several years and most of the recognizable pieces are still there. The building is there. The tables are there. The menu is mostly the same. Some of the same employees may even remain.

Yet you find yourself saying something frustratingly vague:

“The feel of the place has changed.”

That sounds almost meaningless. What is a feel? Where is it? How would we measure it?

And yet another person who knew the place might immediately understand exactly what you mean.

The interesting possibility is that “the feel of the place” isn’t meaningless at all. It may be an emergent property—a real experience produced by the interaction of many smaller things.

Imagine that we could identify twenty variables contributing to a restaurant experience. There is the quality of the food, portion sizes, prices, lighting, music, cleanliness, wait times, furniture, staffing levels, employee friendliness, noise, menu selection and a dozen other things.

But the customer doesn’t experience those twenty variables as twenty separate columns in a spreadsheet.

They experience them together.

And together, those twenty variables produce something else.

Call it the 21st Variable.

The 21st Variable is the emergent quality of an experience that arises from many interacting features but cannot be adequately understood by examining those features separately.

We already have words for some versions of it.

Atmosphere.

Culture.

Character.

Charm.

Vibe.

Soul.

Feel.

These are remarkably imprecise words for experiences that can have remarkably powerful effects on our behavior.

Nobody Designed the 21st Variable

This is where the problem becomes especially interesting.

The 21st Variable may never have been deliberately created.

Imagine a small family restaurant. The portions are generous because Dad likes feeding people. Mom knows half the regular customers by name. Their son occasionally sits down to talk with an elderly customer when the restaurant isn’t busy. The servers have worked there for years. The chairs don’t match perfectly because they were purchased at different times. The homemade bread is slightly different every day. Nobody rushes customers out after dinner.

Perhaps nobody sat down and designed any of this as a customer-experience strategy.

It simply happened.

Yet all those little variables interacted to produce something customers loved.

The place had a feel.

And this creates an unusual vulnerability because the same emergent process can operate in reverse.

Nobody has to decide to destroy the restaurant’s character.

They merely have to make a long series of individually sensible decisions.

Could we reduce the portion of potatoes slightly?

Could one server handle another table?

Could we purchase the bread from a supplier?

Could we replace this expensive ingredient with a cheaper equivalent?

Could we encourage customers to leave a little sooner so we can turn the tables faster?

Could we standardize the menu?

Could we reduce training time?

Could we replace the old furniture with something easier to clean?

Every decision might make perfect economic sense.

Every spreadsheet might show an improvement.

And eventually a regular customer walks in and thinks:

This place isn’t the same anymore.

Nobody deliberately destroyed the 21st Variable.

Its destruction emerged too.

The Laser-Tag Problem

Consider something as simple as laser tag.

Suppose a laser-tag facility owns twelve working units. Sending a single damaged unit away for repair is expensive, while sending six at once qualifies for a substantial maintenance discount.

Someone notices this and makes an entirely reasonable optimization:

Don’t send the equipment for repair until six units are broken.

The maintenance numbers improve.

Unfortunately, the customer may not actually be purchasing “twenty minutes of laser tag.”

The customer may be purchasing the experience of running around a dark arena in a huge chaotic science-fiction battle.

Twelve players can create a six-on-six battle.

With only ten working units, the experience may still be excellent.

Eight starts feeling different.

Six means three-on-three.

Technically, the company is still selling laser tag. The arena is open. The game lasts twenty minutes. The customers receive working equipment.

But something important has disappeared.

The real product was the big battle.

Maintenance costs are easy to put into the quarterly report.

Good luck putting epicness into the quarterly report.

And yet epicness may be one of the most important things the company sells.

This is how optimization can become strangely self-defeating. An organization can become increasingly efficient at producing a technically recognizable version of its product while becoming less effective at producing the experience that originally made people want the product.

The Asymmetry of Measurement

Suppose a large restaurant chain discovers that reducing every serving of potatoes by a small amount will save $1.2 million per year.

That $1.2 million is beautifully visible.

It can be calculated. Reported. Graphed. Compared with last year. Someone can be rewarded for discovering it.

The corresponding loss may be almost invisible.

A customer receives dinner and thinks:

“Huh. They used to give you more potatoes.”

He doesn’t complain.

He doesn’t ask for the manager.

He doesn’t fill out a customer survey.

He doesn’t leave a one-star review.

He simply enjoys the restaurant slightly less.

Then another small thing changes.

And another.

And another.

Eventually he stops going.

Perhaps even he cannot explain precisely why.

This creates an enormous information asymmetry. The organization can possess extremely high-resolution information about what it saved while possessing extremely low-resolution information about what it lost.

Worse, the measurements may continue showing success.

Food costs improved.

Labor efficiency improved.

Revenue per customer increased.

Waste declined.

Customer satisfaction remains respectable.

Nothing about those numbers has to be fraudulent. The organization really did improve the things it measured.

The problem arises when incomplete measurement is mistaken for complete perception.

An organization can optimize twenty variables successfully while accidentally destroying the twenty-first.

Human Beings Are Measurement Instruments Too

This doesn’t mean the 21st Variable is completely unknowable.

It may simply resist the kinds of standardized measurement organizations prefer.

A customer survey might ask:

“How satisfied were you with your visit on a scale from one to ten?”

Useful information.

But consider sitting down with a regular customer and asking:

“Why did you used to love coming here?”

That is a radically different measurement instrument.

Or ask:

“Has anything about this place changed in a way that’s hard to put your finger on?”

And perhaps the most revealing question:

“What could we change that would make this place stop feeling like this place?”

That question is unusual because it isn’t asking what should be added.

It is searching for the invisible structural supports of the existing experience.

The answers might be surprising.

Customers might not care about the expensive renovation being contemplated for next year. They might care enormously about the bartender who remembers their name.

They might not care about expanding the menu from thirty items to fifty. They might care deeply about one peculiar sandwich that has been there for twenty years.

They might not want a sophisticated new digital ordering system.

They might just want the giant pile of home fries.

Humans can perceive these things remarkably well because we don’t experience environments as spreadsheets. Our brains continuously integrate enormous numbers of signals into overall impressions.

Sometimes “Something feels different” may be the conscious output of an extraordinarily complicated unconscious calculation.

Optimization Inside a Protected Boundary

None of this means businesses should stop optimizing.

Quite the opposite.

A restaurant that ignores food costs may close. A laser-tag facility that ignores maintenance expenses may go bankrupt. A company that refuses to standardize anything may become chaotic as it grows.

Businesses that ignore economics disappear, taking their wonderful 21st Variables with them.

The question therefore isn’t whether to optimize.

The better question is:

What must optimization not destroy?

Before optimizing a successful experience, an organization could first attempt to identify the conditions necessary for producing that experience.

In the laser-tag facility, perhaps management discovers that customers consistently have a great time when ten or more units are operating, but enjoyment drops rapidly below that.

Excellent.

Ten working units become a boundary condition.

Management can now optimize maintenance costs as aggressively and creatively as it likes—with one constraint:

Never knowingly operate below ten functioning units.

Now the optimization is working inside a protected boundary.

This suggests a broader principle: preset limits to optimization.

A restaurant might establish that certain signature dishes will never be reduced below a particular quality.

A workplace might decide that efficiency improvements cannot increase a particular employee-to-supervisor ratio.

A game developer might optimize monetization while declaring that purchases will never affect competitive ability.

The protected boundary says:

Improve everything you can, but don’t cross this line, because beyond this line we may begin destroying the reason the system exists.

That requires knowing what the system is actually for.

And that question can be surprisingly difficult.

What Are We Really Selling?

The laser-tag company isn’t necessarily selling twenty minutes.

The restaurant isn’t merely selling calories.

The tavern isn’t merely selling drinks.

The mall isn’t merely renting retail space.

The video game isn’t merely producing engagement minutes.

The workplace isn’t merely converting employee hours into output.

Each of these systems produces experiences, and experiences contain emergent properties.

That means some of the most important things an organization creates may never appear as individual variables at all.

They exist between the variables.

This also explains why the early stages of an organization can sometimes feel so different from the later ones. In a small operation, the owner may not need a sophisticated measurement system because the owner is standing inside the system.

She hears the complaints.

She sees the regular customer who suddenly stops coming.

She notices that Saturday nights aren’t as lively.

She hears employees joking with customers—or notices when they stop.

She doesn’t necessarily measure the 21st Variable.

She is immersed in it.

Scale changes that.

Eventually decision-makers cannot directly experience everything they control. Information has to travel upward through managers, surveys, reports and dashboards. That isn’t necessarily dehumanization or greed. It is partly an unavoidable information-processing problem.

The organization has become too large to perceive itself directly.

So it builds instruments.

And instruments can only detect what they were designed to detect.

Protecting the 21st Variable

Perhaps this gives organizations another task as they grow.

Don’t merely ask:

What should we measure?

Ask:

What might our measurements be unable to see?

Don’t merely ask:

How can we make this more efficient?

Ask:

What exactly are people coming here to experience?

And before changing something that appears inefficient, ask:

Could this apparent inefficiency be contributing to something important that we haven’t identified?

The mismatched chairs might be irrelevant.

Or they might be part of the charm.

The oversized serving of potatoes might be wasteful.

Or generosity might be part of the brand.

The experienced employee who doesn’t look extraordinary on six standardized performance metrics might actually be carrying fifteen years of institutional memory around in her head.

The old laser-tag equipment might genuinely need replacement.

But the six-on-six battle might be sacred.

Not everything old deserves preservation, and not everything measurable deserves suspicion. Optimization, measurement and standardization have produced extraordinary improvements in modern organizations.

The danger lies elsewhere.

It lies in forgetting that reality contains more variables than the dashboard does.

And sometimes the most important variable isn’t one of the twenty things we’re measuring.

It’s the mysterious thing those twenty things create together.

It’s why people came.

It’s why employees cared.

It’s why customers returned.

It’s why somewhere became that place rather than merely another place providing the same service.

And sometimes we don’t discover how important it was until we’ve optimized it away.

Everything appears to still be there.

The building is there.

The product is there.

The numbers may even look better.

But the people who knew the place before walk in, look around, and somehow know.

It just doesn’t feel the same anymore.

The first twenty variables survived.

The 21st did not.