Where Native Alpha Actually Lives, and How It Decays

Executives look for company-wide competitive advantage and miss the useful stuff, because the useful stuff is usually smaller than a company.

A business unit can hold advantage the parent does not. So can a founder, a single salesperson, a dataset, one relationship, one workflow, one feature, or a combination that only appears when two portfolio companies sit at the same table. A narrow workflow in one customer segment is often the seed of something large. Search at the level where the evidence actually is.

Advantage is always advantage for something

A famous brand matters in one market and means nothing in the next one. A proprietary dataset is priceless for one decision and useless for another. A relationship opens one buyer and no others. A patent blocks one competitor and leaves ten substitutes alone.

So when somebody says our data is our advantage, I wait for the next sentence. Advantage for what? Who cares? Which decision does it improve? Why can nobody else reproduce the outcome? What gets stronger when we use it? Without those answers we are admiring assets rather than allocating around them, and admiring assets is a pleasant way to spend a board meeting and a bad way to spend a year.

Protect it where it lives, not where it is convenient

Not every advantage belongs in a patent. Depending on where it actually sits, the right fence might be a trade secret, a contract, a data-rights clause, workflow integration, a customer relationship, a distribution agreement, brand reputation, access, accumulated evidence, operational know-how, employee knowledge, or the structure of a partnership.

Figure out where the advantage lives before choosing the protection. Legal protection matters, and so do architecture, incentives, records, permissions, exclusivity, portability, security, and continuity. A sharp strategy with sloppy rights is not much of a strategy, and I have watched more advantage leak through a badly drafted data clause than through any competitor's engineering.

Publishing can strengthen the thing it reveals

Ideas can hold advantage, which makes publishing them look like giving it away. Sometimes it is. Often it is not, because the framework becomes more valuable once people understand it: reputation improves, relationships arrive, access widens, the thinking gets sharper under criticism.

Execution, judgment, accumulated evidence, and the ability to keep extending the idea are hard to copy from a blog post. Choose deliberately, and choose based on whether publishing strengthens the next use.

Everything decays

Technology moves. Employees leave. Contracts expire. Competitors learn. Customers consolidate. Regulations change. Channels weaken. Data turns up somewhere else. AI turns scarce expertise into a commodity. Reputation can disappear faster than it was built.

Put the question on a calendar rather than waiting for a crisis: does yesterday's advantage still exist? Sometimes the answer is no. That is information, and it is cheaper to act on than to discover during a renewal cycle.

Stopping is part of the discipline

I do not treat continuation as evidence of conviction. Often it is evidence that nobody wants to be the person who says the thesis failed. Stop when the advantage turns out to be ordinary, when the customer does not care enough, when the asset cannot be legally or operationally used, when competitors reproduce it cheaply, when AI commoditizes it, when implementation destroys the economics, when the workflow cannot change, when the advantage does not compound, or when the same resources have a better home.

And keep asking whether each use strengthens the next. If the tenth customer is no easier than the first and the hundredth is no harder to copy than the tenth, growth is not building anything durable. Better to know that now, while the decision is still cheap.