Portable Playbook · Framework

The Coca-Cola Inversion

When Your Greatest Strength Becomes Your Greatest Vulnerability

Section III · CROSS-CUTTING PLAYBOOKS: THE IDENTITY LOOP · The Identity Loop

Every organization has a domain where its core competence is systematically misleading it, and that domain is invisible precisely because the organization is using the competence to evaluate the threat.

How It Works

For every major decision, instead of asking "what does our core competence tell us to do here?" ask the opposite: "What would we conclude if our core competence were systematically misleading us on this specific question?"

Coca-Cola's identity as a data-driven organization was correct. It was also the thing that blinded them. The data said New Coke was better. The data was right. And the data was irrelevant, because the decision operated on territory where data has no jurisdiction: brand identity, emotional attachment, cultural meaning. Morgan's partners never asked "what if our social network is not actually a proxy for character?" because the social network had functioned as a character proxy for a century. The inversion is uncomfortable by design. The answer is usually no. But the one time the answer is yes, the practice saves the company, and that one time is the exact moment when every other analytical tool in the organization will produce the wrong answer.

How to Use This Today

The engineering-led company evaluating a brand decision.

The engineers will reach for data. The data will be precise and irrelevant if the decision turns on emotion, identity, or tribal belonging. Run the inversion before the next major product decision: gather your leadership team and ask, "What would we conclude if our core analytical competence were systematically misleading us on this specific question?" Force written answers. If the answer is always "nothing, our analysis is sound," you have not performed the inversion. You have performed self-congratulation. The inversion works only when it produces discomfort. Coca-Cola's taste tests were methodologically flawless. The methodology was the problem: it measured taste when the decision was about identity. Your A/B test might be methodologically flawless too. The question is whether you are measuring the variable that matters or the variable your competence can measure.

The sales-led company evaluating a product decision.

The sales team will reach for customer feedback. The feedback will tell you what customers want today, not what they will want when a competitor redefines the category. The specific inversion: before your next product planning cycle, ask each product leader to write a one-page memo answering this question: "If our most successful sales technique were actively preventing us from seeing the market accurately, what would we be missing?" The sales team's strength is listening to existing customers. The inversion asks whether existing customers are the right signal source for the decision at hand. Blockbuster's customers said they wanted more stores and better inventory. Netflix did not ask Blockbuster's customers. It asked people who hated going to stores. The core competence of listening to your customers misled Blockbuster precisely because Blockbuster was so good at it.

Any leadership team about to make a bet on the basis of their strongest analytical tool.

Before committing, conduct the inversion as a formal exercise: assign one senior leader to build the case that the tool is wrong for this specific decision. Give them a week and a budget for external research. If the case they build is trivially dismissable, proceed with confidence. If the case they build is plausible, you have discovered the one-in-ten scenario where the inversion saves the company, and that discovery is worth whatever the exercise cost.

The inversion requires temporarily suspending the very belief that has produced your success and asking whether that belief might be the thing that destroys you. If this feels comfortable, you are not doing it correctly.