The next dollar should buy an answer

Imagine three projects arriving at the same investment review. Each is described as promising. Each has a large market slide, a roadmap, and a request for another quarter of funding.

But the first project does not yet know whether its technology works. The second has working technology and no evidence that a buyer will pay. The third has both, but no credible way to integrate into the customer’s operation. Calling all three “early stage” hides the most useful difference between them.

They do not need the same next step. They need different answers.

A real narrowing¶

My sensing work began with a broad attraction to material signatures. The archived experiments made the uncertainty more specific: geometry, composition, solvent, and time could all influence a response. The useful decision became which reference measurements and controls would distinguish those explanations. A narrower drying question gave the next experiment a clearer purpose than another claim about a universal scanner.

A portfolio of questions¶

My interest in strategy is the machinery between an interesting possibility and an actual commitment of people, money, or attention. A useful system makes that machinery inspectable. What are we betting on? What do we know? Which unknown could invalidate the plan? What are we going to do about it?

The next dollar does not always need to purchase immediate revenue. It can purchase an answer that improves a later decision. A technical experiment that rules out an expensive route may be a good investment. A successful demonstration that leaves the central buying objection untouched may be a poor one.

The distinction is not between research and execution. It is between learning that changes what we do and activity that makes the project easier to present.

March’s work on exploration and exploitation helps explain why this is difficult.¹ Exploring unfamiliar possibilities and improving established activities compete for resources, with different timing and uncertainty. A review process that rewards only near-term, easily measured outputs can systematically favor what the organization already knows how to do.

I do not infer that exploration should receive an unlimited budget. I infer that its budget should be defended in terms appropriate to the uncertainty it is meant to resolve.

Write the decision before the experiment¶

Here is the small record I would want beside every substantial exploratory commitment:

Field	What belongs in it
Decision	The commitment we are actually considering
Critical assumption	What must be true for that commitment to make sense
Evidence	What we have observed, separate from what we infer
Next test	The cheapest credible way to resolve the important uncertainty
Branches	What we would do after a positive, negative, or ambiguous result
Owner and date	Who decides, and when the question returns for review

Rita McGrath’s discovery-driven planning provides a closely related discipline: identify important assumptions and use staged work to convert them into knowledge.² The practical addition I care about is the last half of the record. If we cannot explain how a result would change the decision, we may be proposing an experiment for reassurance.

“Interview ten customers” is an activity. “Determine whether a named buyer controls a budget for this problem, before building an integration” is a decision-oriented investigation. The interviews still require judgment. A polite expression of interest is not a purchasing commitment, and ten conversations do not automatically constitute a representative sample.

Stopping is an output¶

Organizations can make stopping feel like the absence of a result. That creates a predictable incentive: preserve the project, rename the milestone, and ask for one more increment.

I would rather record a stopped project with a useful explanation than a permanently green dashboard. The explanation should say which assumption failed, what evidence caused the update, which assets remain useful, and whether any condition would justify reopening the work.

A stop rule agreed before results arrive can reduce the temptation to improvise a new justification. It cannot eliminate judgment. Sometimes an unexpected result reveals a better problem. Sometimes the original test was badly designed. A good process allows a change of direction, but makes the new rationale visible.

The same discipline applies to success. If a project passes a technical test, the next commitment should correspond to what was learned. Passing one gate is not a blank check for every downstream assumption.

The strongest objection¶

A system like this can become a machine for killing unusual ideas. Some research needs time before its questions become well defined. If every project must predict its next discovery, exploration becomes a performance of certainty.

The answer is to distinguish kinds of commitments. A protected research allocation can fund open-ended inquiry. A commercial expansion budget can demand evidence about adoption. Both need accountability, but they do not need identical milestones.

I would also examine the portfolio over time. Are all the stopped projects the unfamiliar ones? Are the same sponsors always exempt from evidence? Does the process recognize valuable negative results? A formally neutral template can still encode a strong bias toward the familiar.

The meeting should change something¶

A strategy meeting earns its cost when it changes a decision, resolves a disagreement, exposes a consequential unknown, or assigns a test someone will actually perform. A meeting that only updates the audience may be useful communication; it should not be mistaken for capital allocation.

The test of the system is not the elegance of the scorecard. I would look for fewer repeated debates, clearer reasons for continuing or stopping, and evidence that resources move when beliefs change. I would also watch for false precision: a weighted score can conceal a disagreement as easily as it can clarify one.

The goal is a record that a future decision-maker can reconstruct. Here is what we believed. Here is why we spent the next dollar. Here is what it taught us. Here is what we did afterward.