What Is a Product–Market Vector (PMV)?

Why generic visibility tools miss the failures that matter

Generic AI visibility tools often treat a company as one brand, one category, and one flat list of keywords. That structure hides the differences growth teams need to see: a company can be highly visible for one offering and nearly absent for another, visible to executives but invisible to the technical evaluator, strong in a broad market prompt yet missing in the precise scenario that signals purchase intent.

A stronger benchmark starts with a hierarchy. Each level below narrows the question "who is being asked about what" until it matches a real buying moment.

The strategy hierarchy

Offering

The product or service line the market can recognize — a cybersecurity platform, a data observability product, a vertical SaaS product, a managed service.

Product–Market Vector (PMV)

At GridStrat, we call this a Product–Market Vector, or PMV: a specific offer aimed at a specific ICP, in a defined geography, for a particular use case.

In plain English, a PMV answers:

What are we selling, to whom, where, and for what use case?

For example:

AI visibility diagnostic for founder-led B2B SaaS companies in North America that need to understand why competitors are being recommended in AI-generated answers.

That is more precise than "AI visibility services." The PMV is the smallest unit that carries real commercial meaning — and it is the level at which most visibility failures actually occur.

Target market

The broad pool of potential buyers or users. For example: B2B technology companies.

ICP

The specific, targetable segment inside that market. For example: founder-led B2B SaaS companies with complex offerings, long sales cycles, and a need to improve how they appear in AI-assisted buyer research.

Persona

The human role inside the ICP. A useful persona includes role, seniority, influence in the decision, goals, pains, decision drivers, search behavior, trusted sources, and preferred content formats.

A job title alone is not enough. A CFO, VP Marketing, technical evaluator, procurement lead, and founder may all investigate the same category differently.

Scenario

The moment that causes the persona to search now: a failed vendor, a renewal deadline, a budget cycle, a compliance deadline, a new strategic initiative, a board question, an urgent operational problem, a competitor gaining visibility.

This distinction matters. "Cybersecurity software for mid-market companies" is not a sufficiently precise test.

A better test might examine:

A compliance leader at a regulated mid-market firm, after a failed audit, exploring vendors that can reduce remediation time.

That prompt context is closer to a real buying moment. It also produces a more actionable result.

Benchmark current external perception before future ambition

A benchmark should use the company's current product and market reality, not only its internal growth aspirations.

Strategic goals are valuable for deciding where to invest, but feeding those ambitions into an external-perception audit can bias the baseline. The audit should first measure how AI systems represent the company today. Then leadership can decide which gaps deserve correction based on strategy, commercial value, and execution capacity.

In other words:

Measure current perception first. Use strategic ambition to prioritize what to correct.

How the PMV structure is used in the benchmark

The hierarchy is the scaffolding of GridStrat's 21-day AI visibility benchmark. PMVs define what gets tested; personas and scenarios define the prompts; and results are analyzed at every level of the hierarchy, so a single company-level score never conceals a strong core product and a weak growth product.

The same structure carries into correction: findings such as "visible at company level, absent for one PMV" or "visible for executives, absent for practitioners" each map to a specific, narrow fix rather than a generic instruction to produce more content. And because each PMV's approved claims live in the company's governed facts, corrections propagate consistently across every source AI systems read.

Frequently asked questions

What is a Product–Market Vector?

A Product–Market Vector, or PMV, is a specific offering aimed at a specific Ideal Customer Profile, in a defined geography, for a particular use case. It is the smallest unit of analysis that carries commercial meaning, and the level at which most AI visibility failures actually occur.

How is a PMV different from an ICP?

An ICP describes a buyer segment; a PMV joins an offering to that segment, in a geography, for a use case. One ICP can appear in several PMVs, and one offering can serve several ICPs — the PMV is the intersection where a specific offer meets a specific buyer situation.

Why benchmark AI visibility at the PMV level?

Because company-level averages hide the product-market combinations where growth is won or lost. A company can be highly visible for one offering and nearly absent for another; only benchmarking by offering and PMV exposes that difference so resources can be allocated to the gap that matters.