- Revenera's survey of 415 established software producers found 47% put predictable pricing for customers ahead of maximizing revenue, against 16% who favor revenue.
- Revenera's survey also found 41% of vendors feel margin strain from serving AI. It does not say who bears the cost; vendors absorbing it is this article's view.
- Buyers should ask how AI is priced, what happens if usage doubles, and whether staff use the AI features before paying more.
A steady price on a variable cost
When a software company adds AI, it takes on a cost it did not have before. It also takes on a customer worry it has to calm. Revenera's 2027 Monetization Monitor suggests many vendors lean toward a steady bill for the customer.
This article's view is that AI pricing is less a calculation than a promise. The promise is a bill that does not surprise you. Someone has to pay for keeping it. The survey does not say who. One possibility is that vendors absorb the difference. That is an argument, not a survey finding.
Who was asked, and the limits
Revenera polled 415 leaders at established software producers and makers of intelligent devices. The survey ran in June and July 2026. Of these, 69% are software or SaaS companies. The other 31% make products such as industrial equipment, networking gear, IoT and medical devices. Revenera says 66% earn more than $100 million a year, and 75% sell at least ten software products.
The answers are self-reported. Invitations went to Revenera's marketing database and to a third-party panel, and 8% of respondents are Revenera customers. Some questions went only to a more qualified subset. Revenera sells software monetization tools and published the survey itself, so read its conclusions about usage data with that in mind.
Building AI was the easy part
Revenera found that 96% of respondents have added AI to at least some products, up from 80% last year. Another 4% plan to within 18 months. About 61% say most or all of their products contain AI features.
Customer uptake lags, according to the vendors. Among the 380 respondents asked, 48% report that fewer than half of eligible users actively use their AI features. Revenera sorts these into limited (under 25% of eligible users) and low (25% to 49%). Only 11% of that group report high adoption, meaning more than 75% of eligible users.
Why predictability wins
A traditional subscription costs the vendor little each time a customer clicks. AI features differ, because each use draws on computing power the vendor pays for. A director of product management at a firm with $250 million to $1 billion in revenue called balancing "high API compute costs" against a fair, predictable subscription a "tricky puzzle."
Pure usage pricing would match price to cost. But the bill would move from month to month. InfoWorld says reports of surprise AI bills have become more common over the past year.
Revenera found that 47% of respondents prioritize predictability for customers over near-term revenue, against 16% who prioritize revenue. Another 37% take a balanced approach. Among the 90 respondents asked about pricing models, the most common are a higher-priced tier (40%), AI at no extra cost (39%) and a hybrid of subscription plus usage or outcomes (37%). Revenera notes that each of these gives customers a more predictable price.
Who may be holding the risk
Picture an all-you-can-eat buffet. The diner pays one price, and the restaurant covers the cost of heavy eaters. If AI is sold at a steady price while its cost varies with use, a software vendor may sit in the restaurant's seat.
Revenera's figures are consistent with that picture. They do not prove it. Most of the picture is positive: 69% of respondents believe their AI offerings drive revenue growth, and 52% report better profitability. Belief is not a measured result.
The cost side shows strain too. Revenera reports that 41% of respondents feel serving AI to customers is weighing on their margins. Revenera does not tie that figure to any pricing model. It could arise under flat, tiered or metered pricing. That may mean some vendors are absorbing cost swings, though the survey does not measure this.
Asked what most threatens future growth, respondents pointed to rising AI delivery costs (47%), rising cloud costs (43%) and being unable to launch new ways of charging (37%).
Vendors also lack the data to price well. The top blockers to matching price with value, each cited by 49%, were data availability and understanding what customers value. Revenera found that 41% still collect usage data through manual processes or engineering effort. Only 27% use automated alerts to spot churn risk.
Pressure on price, not yet on seats
One expectation is that AI agents will cut the number of paid seats. So far that shows up little in the survey. Only 7% of respondents report seat loss. But 45% report downward pricing pressure from AI. Revenera cautions that the picture could change as adoption grows.
InfoWorld adds that some buyers use the option of building their own AI tools as leverage in negotiations. Growing AI budgets are also pushing customers to look harder at what they spend elsewhere.
Questions to put to your team
For software you buy, ask how each vendor prices AI. Is it included, sold in a higher tier, or metered? If it is metered, ask what the bill looks like if usage doubles, and whether a cap exists.
Before paying for a premium tier, check whether your own staff use the AI features. Among the 380 vendors asked, 48% report that fewer than half of eligible users actively use their AI features.
At renewal, note that 45% of vendors in the survey report downward pricing pressure from AI. For software you sell, ask whether you can see which customers use AI and what it costs you to serve them.
A steady price on a variable cost does not remove the risk. It only decides who holds it, and for how long.
Produced by the WebPulse Newsroom with AI assistance from the original reporting credited below, and checked against that source by our editorial review. How we use AI.
Original reporting: Revenera.
