What Transfers Across Different Businesses

Experience transfers when the mechanism survives a change in context.

Over 20 years, my work has crossed Amazon, Ancestry, Stanley Black & Decker, and the Bed Bath & Beyond portfolio, including Overstock.com and buybuyBABY. The customer, business model, technology, and organizational constraints changed. The useful operating questions remained remarkably stable.

What decision are we making? What customer evidence supports it? Who owns the result? How will we know whether it worked? When will we review and correct it?

Company vocabulary does not transfer cleanly

Every organization develops shorthand. That language can make work efficient inside the company and nearly useless outside it.

A reusable lesson needs to be separated from its internal acronym, tool, meeting name, and organizational chart. What remains should be an operating mechanism another practitioner can evaluate in a different setting.

That is one reason the public Playbooks library describes purpose, inputs, ownership, outputs, measures, and failure modes. The goal is not to recreate a former employer's process. It is to expose the decision logic and make the boundaries visible.

Customer evidence still comes first

Retail, subscriptions, ecommerce, and consumer brands produce different behavioral signals. The operating discipline is the same: use the minimum evidence needed to improve a decision, then review what customers actually did.

That requires distinguishing observation from inference. A purchase, renewal, service interaction, or lapse can be observed. Intent usually has to be inferred and tested.

Measurement has to match the claim

The strongest number is not always the most useful number.

An attributed result may help explain channel credit. A holdout or other comparison may be needed to estimate causal lift. A customer measure can move while the business result remains unchanged, or the reverse. The claim should remain inside the evidence.

Teaching improves the mechanism

I have tested these ideas in conference panels, team sessions, and university guest lectures, including two Braze Forge panels, eTail Palm Springs, and Northwestern Medill.

Speaking is useful because a room exposes hidden assumptions. If a mechanism needs ten minutes of company history before it makes sense, it is not portable yet. If an audience cannot identify the owner, input, output, and measure, the model needs another pass.

The classroom version of the Growth Operating System uses four moves: diagnose the operating problem, design the mechanism, implement it at a sustainable scale, and audit the loop. Students apply the model to one operating problem rather than leaving with a list of ideas.

A test for transferable experience

Take one lesson from a prior role and remove the company name, internal vocabulary, and organizational details.

Then write:

1. the operating problem;
2. the evidence available;
3. the decision made;
4. the mechanism used;
5. the result that could be measured;
6. the conditions under which the lesson would not apply.

If the lesson remains useful, it may be a mechanism. If it disappears with the company vocabulary, it is still an anecdote.