An August 6th Investor’s Business Daily article titled Software Stocks Fall As Datadog, HubSpot Earnings Raise Questions Over AI Pricing reported that Datadog stock fell more than 16% and HubSpot plunged 22% after disappointing earnings and guidance. Salesforce, ServiceNow, Adobe, and others pulled back alongside them. Investors are rattled by one question: how do software companies make money from AI?
The market is treating this as a new problem. It isn’t. It is an old problem that AI just made impossible to ignore.
Here is what the selloff is really telling us.
Per-Seat Pricing Is Product Myopia in a Business Model
The article flags growing investor concern about per-seat pricing models. That concern is well founded, and the reason goes deeper than AI.
Per-seat pricing charges for access to the product. It says: value comes from having a license. That made reasonable sense when a human being sat behind every seat and did the work. The seat was a proxy for the value because a human using the software was a human getting the result.
AI breaks that proxy. When an agent can do the work of ten seats, the per-seat metric stops tracking value and starts tracking something arbitrary. Buyers notice. They start asking why they are paying for seats when the work is being done by something that does not need one.
Per-seat pricing was always a product-first metric. It charged for the thing, not the result the thing produced. AI did not create that problem. It revealed it.
HubSpot’s Decision Is the Right One
HubSpot deliberately lowered its AI revenue guidance. According to the analyst quoted in the article, the company made pricing and go-to-market changes to show customers proof of value before they make a buying commitment. That includes free trials, lower entry price points, and outcome-based pricing for agents.
That is exactly the right move, even though it hurts short term.
HubSpot is admitting that its buyers cannot yet see the value clearly enough to commit to paying for it. Rather than pushing harder on a price buyers cannot justify, they are doing the work of making the value visible first. Free trials let buyers experience the result. Outcome-based pricing ties the metric to the result once it is visible. Lower entry points reduce the risk of committing to something unproven.
A company that lowers its price or reaches for a flexible mechanism is confessing that the value has not yet been made legible. HubSpot is making that confession openly and doing something about it. That is more honest than holding a price nobody can defend.
Consumption Models Have the Same Problem
Datadog and Snowflake charge based on how much software buyers consume. That feels more value-aligned than per-seat, because at least usage tracks activity. But consumption is still a product metric. It measures how much of the product the buyer used, not what changed in the buyer’s world because they used it.
When buyers pull back on consumption, as they appear to be doing now, it is because they cannot connect the usage to a result they care about. The consumption went up. The outcome did not become more visible. So buyers do the rational thing and spend less.
A metric that tracks consumption will always be vulnerable to that question. A metric that tracks the result the buyer gets answers it automatically.
What the Winners Will Have in Common
The companies that figure out AI pricing will share one thing. They will have done the work of finding the result first, before they set the metric.
That means talking to buyers about the specific problems they are living. Documenting what changes when the problem is gone. Naming the result in the buyer’s own terms. Once the result is that specific and that visible, the right metric becomes obvious. It is the unit that best tracks what the buyer actually got.
The market is learning this lesson through stock prices. Your buyers are learning it through their own AI budgets. The companies that get there first, that find the result and build the metric around it, will hold their prices, earn commitment, and stop watching revenue evaporate every time a buyer questions the value.
The selloff is painful for investors. For everyone else it is a signal worth paying attention to.
Share your comments on the LinkedIn post.
Now, go make an impact!
Tags: AI, ai pricing, outcome pricing, pricing, pricing metrics, pricing models, pricing strategy, SaaS, value, value pricing
