GenAI’s gravity moment: why your value metric will make or break you

We’ve finally hit the inflection point for generative AI.

The magic magic that once accompanied each model release is now fading. Perceived innovation per update is flattening. Leading models now feel largely interchangeable to most end users.

This is the gravity moment for GenAI:
When the technology stops selling itself, and the economics of how you package, price, and deliver it start deciding the winners.

When the model is no longer the differentiator

If you’re building on or with LLMs today, you’re in a market where:

• Switching costs are lower than ever: customers can move between providers with minimal friction
• Marginal improvements in speed or accuracy are no longer headlineworthy (tf i still don’t believe much of those benchmark reports each LLM provider cited when claiming they were “best in class” and eQuiVaLeNt of XYZ pHdS).
• The moat is not in the model itself: it’s in the business around it (and this is where my initial bet was wrong re how well LLM wrappers could really do)

This shifts the competitive conversation from “whose model is better/bigger?” to “whose business is built to monetise value most effectively?”

Value metrics: the core of genAI monetisation

A value metric is the measurable indicator you bill against: compute hours, API calls, storage, tokens, seats, or a hybrid.

Get it right, and you have:

• Aligned incentives: customers pay in proportion to the value they receive
• Expansion revenue built-in: usage growth drives ARR without contract drama
• Churn defence: pricing feels fair because it scales with outcomes

Get it wrong, and you’re subsidising your most expensive accounts while overcharging light users.

As Peter Zotto put it this week when I tagged him in Iwo’s LinkedIn post: pricing isn’t inherently complicated, but it takes deliberate effort. The key is finding a curve that feels fair yet still scales revenue: often by segmenting buyers, combining baseline commitments with usage flexibility, and tying tiers to outcomes customers actually care about. The challenge for scaling GenAI companies is doing all of that globally, without adding operational drag: something that requires a commercial layer capable of handling changes across currencies, markets, and tax regimes as easily as changing a price in one market.

In GenAI, where infra costs can climb just as fast as customer usage, ignoring your value metric is basically ignoring your margin.

Value-based pricing: worth watching

Beyond metered billing, value-based pricing, charging in proportion to perceived value, is worth watching.

Why? Because in GenAI, the same model can be:

• A curiosity for one customer.
• A mission-critical workflow for another.

That gap in perceived value can swing willingness to pay by multiples. The companies that can measure, segment, and price against that perception will capture far more upside than those locked into one-size-fits-all rates.

Localisation: the multiplier

Once your value metric is right, localisation multiplies its impact.

Selling in USD to customers in Dubai, Paris, or Jaffna quietly erodes conversion. Switching to local currencies has been shown to lift conversions by 6% to 51% depending on the market.

It’s not just optics, in fact it often:

• Builds trust: you operate in their market
• Removes friction: no mental conversions at checkout
• Protects fairness: shields against FX swings that make your product look more expensive overnight.

If the value metric isn’t tied to actual value delivered, localisation only scales the mismatch across currencies.

Infrastructure: the quiet competitive edge

Optimising value metrics and localising pricing are only as effective as the infrastructure underneath them.

Without a foundation that can:

• Handle tax compliance across multiple jurisdictions,
• Support the payment methods customers expect in each market,
• Manage FX without eroding your margins, and
• Let you run pricing experiments without rewiring your stack,

…you’re capping the upside before you start.

The fastest-scaling GenAI companies aren’t treating this as a back-office admin. They’re building a commercial layer that abstracts complexity, protects margins, and lets them monetise globally without slowing product velocity.

As the current model race levels out, the next era of GenAI will be defined by those who can execute monetisation globally, turning value into revenue in every market.