There is a moment that happens in every organisation that uses metered analytics — where someone pauses before asking a follow-up question, does a quick mental calculation, and decides the question is not worth the cost. That pause is where insight goes to die.

Per-query pricing — whether structured as queries, tokens, compute units, or API calls — introduces a friction into the analytical process that looks small on a spreadsheet and is enormous in practice. It changes the behaviour of the people using the system in ways that are almost never measured.

The third question problem

In any meaningful analytical investigation, the first question establishes a baseline. The second question explores a dimension. The third question — the one that follows from something unexpected in the second answer — is where the real insight lives.

The most valuable insight in any analysis is usually the third follow-up question. Metered pricing kills it.

When analytics is priced per query, executives and analysts learn — quickly, implicitly, without anyone telling them — to plan their questions in advance. They ask fewer exploratory questions. They stop following threads that feel expensive. They optimise for efficient use of the system rather than thorough investigation of the data.

The result is that organisations pay for an analytics system and then use it in a way that deliberately avoids the most valuable analyses. The pricing model has inverted the incentive.

What metered pricing actually costs

The real costs of per-query analytics pricing

  • The follow-up questions that reveal the real cause of a trend — not asked
  • The exploratory analysis that would have found the anomaly — skipped
  • The second opinion on a number that seemed right — never sought
  • The deep-dive into a segment that looked interesting — deferred indefinitely
  • The audit of last quarter results against current methodology — too expensive to run

None of these costs appear on any invoice. They are invisible — the decisions made worse, the risks missed, the opportunities not identified. They are also, in aggregate, far larger than the query costs that generated them.

The organisational ratchet

Metered pricing creates an organisational ratchet. In the first month, usage is high — people explore the system, run lots of queries, discover what it can do. The invoice arrives. The CFO flags it. A usage policy is created. Usage drops. The system is now used only for the questions that have been pre-approved, pre-planned, and pre-budgeted. Which is to say — only the questions someone already knew to ask.

Analytics that can only answer pre-planned questions is not analytics. It is reporting. And reporting can be done with a spreadsheet.

The case for flat pricing

Flat analytics pricing — per seat, unlimited questions — is not just a commercial preference. It is a statement about what analytics is for. If the purpose of an analytics system is to help organisations make better decisions, then every barrier between an executive and a question is a failure of the system.

When questions are free, behaviour changes. Executives ask the third question. Analysts explore threads that look unpromising. Teams revisit assumptions that were never questioned because questioning them used to cost money. The analytical culture of the organisation changes — from cautious and rationed to curious and thorough.

Curiosity is the input. Better decisions are the output. Pricing that taxes curiosity taxes decisions.

What to look for in analytics pricing

When evaluating any analytics platform, the pricing structure reveals the vendor's assumptions about how the tool should be used. Per-query pricing assumes that questions are expensive to answer and should be rationed. Flat pricing assumes that questions are valuable and should be encouraged.

The second assumption is correct. Every question an executive asks of their data is an attempt to make a better decision. The analytics platform should make that as frictionless as possible — not price it like a utility bill.