The monetization conversation your team keeps deferring

There is a pattern I keep seeing in B2B SaaS right now.

A team ships an AI feature. It works. Users engage with it. The feedback is positive.

Then someone asks: "How do we charge for this?"

And the room goes quiet.

Not because nobody has an opinion. Because everyone has a different one, and nobody wants to get it wrong, and the product is changing so fast that whatever you decide today might not make sense in three months anyway.

So the decision gets deferred.

The feature gets slotted into the highest-priced tier, because it's expensive to run and someone needs to justify the margin, or it gets bundled into everything because "we don't want to block adoption." Either way, the monetization model stays static. And the team moves on to the next feature.

I understand why this happens.

Pricing has always felt like a one-way door in SaaS. You change it, you upset someone. You raise prices, you risk churn. You experiment with packaging and suddenly CS is fielding angry calls about features that used to be included. So most teams touch it as rarely as possible.

The problem is that AI broke the assumptions that made that caution reasonable.

The margin equation has changed

In traditional SaaS, your marginal cost per user was close to zero. You built the feature once. Every additional customer using it cost you almost nothing. So you could afford to be generous, bundle features broadly, worry about monetization later, and trust that margin would follow.

That equation is gone.

Every time a customer uses an AI feature, you pay for it. Tokens, inference, API calls. The cost is real, it is per-usage, and it shows up early. Not when the customer renews. Not when they expand. When they first try the thing.

Which means the monetization conversation can no longer wait until "we have more data." The data is being generated right now. The margin is being consumed right now. And a pricing model designed for a world where usage was free is probably already working against you.

Two common non-answers

Most teams respond to this by doing one of two things.

The first is to put AI features behind the highest-priced tier and call it a premium offering. The logic makes sense on paper: high cost, high value, charge accordingly. But in practice this often blocks the adoption that would have justified the price. Customers who might have discovered the value never get close enough to feel it. The feature becomes a selling point in a sales deck rather than a habit inside the product.

The second is to include AI features in everything and hope the cost comes down fast enough. This feels generous and growth-oriented. It is also how you quietly destroy your margin while building a customer expectation that is very hard to unwind later.

Neither of these is really a strategy. They are both ways of deferring the actual decision.

The actual decision is harder

It is not "which tier does this feature belong in." It is "what behavior are we trying to create, and what pricing model reinforces that behavior rather than blocking it."

Elena Verna wrote recently that monetization experiments at Lovable are not focused on revenue increases; they are focused on engagement lift. That the revenue-neutral experiment is often the biggest winner, because engagement today compounds into revenue later.

I think that is right. And I think most B2B SaaS teams are not even close to having that conversation yet. They are still treating monetization as a pricing page problem. A conversion rate to optimize. A tier to fill. A number to hit this quarter.

The model is not the moat. The monetization model is.

Treat the model as the experiment

Here is what I think is actually true. The companies that figure out AI monetization in the next 18 months will not figure it out by getting the price point right. They will figure it out by treating the monetization model itself as the thing to experiment on.

Not endlessly. Not without discipline. But with the same rigor you would apply to any other product hypothesis.

What happens when you give more away at the bottom and charge based on output rather than access? What happens when you let customers feel the value before they hit the paywall rather than after? What happens when you design the upgrade trigger around the moment of realized value rather than the moment of feature discovery?

These are not pricing questions. They are product questions. And they require the same thing every good product question requires: a hypothesis, a way to measure it, and enough patience to let the signal emerge.

Patience is the harder part

A monetization experiment does not resolve in two weeks.

You change a credit limit or a tier boundary or a paywall position, and the first month looks terrible. Engagement might drop. Conversion might dip. Someone in a quarterly review points at the number and says it is not working.

But the cohort that went through the new experience is still maturing. And when the experiment was designed to create a habit rather than capture a transaction, the signal will not be visible until that habit has had time to form.

This is why most teams run monetization experiments once, declare them failures, and go back to optimizing button colors. The experiment was not wrong. The time horizon was.

Why the room goes quiet

I said at the start that the room goes quiet when someone asks how to charge for an AI feature. I want to be precise about why.

It is not because the team does not care about monetization. It is because monetization has been treated, for most of SaaS history, as someone else's job.

Pricing is finance. Packaging is marketing. Revenue is sales. Product owns the feature, not what it costs the customer.

That division made sense when the product was static and the margin was comfortable. It does not make sense anymore.

When your AI features consume real margin per use, the person deciding what gets built and how it gets used needs to be part of the conversation about how it gets priced. Not because product should own pricing. Because the product is the pricing model now.

The way a customer experiences a credit limit, an upgrade prompt, a usage threshold, that is a product experience. It creates a feeling. It shapes behavior. It either reinforces the habit you are trying to build or interrupts it at the worst possible moment. A team that treats those decisions as an afterthought is leaving both revenue and retention on the table.

Better questions to start with

I do not think there is a universal answer to how B2B SaaS should price AI. I think there are better and worse questions to start with.

Not "which tier does this belong in," but "what behavior are we trying to unlock and is our current model helping or blocking it."

Not "how do we protect margin," but "how do we design the experience so that value comes before the ask."

Not "let's test the pricing page," but "let's test the moment, the motion, and the model, and give it long enough to tell us something true."

The companies having those conversations right now are the ones I think will look back on this period as the moment they figured something out that their competitors missed. The ones who deferred it will look back and wonder why their AI features never quite translated into the revenue they expected.

This is the work I do now. Not three months of data extraction and a slide deck. Two sessions. A conversation about how you currently think about your customers, your tiers, and your pricing. A one-page hypothesis about where the real opportunity is, and what to check to confirm it.

If any of this is the conversation your team keeps not having, I would be glad to have it with you instead.

Book the Revenue Readiness Conversation →