Follow ZDNET: Add us as a preferred source on Google.
ZDNET’s key takeaways
- Software pricing is moving to outcome-based models.
- Users and vendors need to agree on success metrics.
- The nature of software engineering work is evolving.
In the year ahead, your relationship with your software vendors may change radically, perhaps even a greater shift than the switch from disks to Software as a Service. You may start paying only for the actual results the software delivers, versus simply paying a monthly charge that you pay even if the application sits on a shelf.
Also: 6 essential rules for unleashing AI on your software development process – and the No. 1 risk
Of course, paying for results requires consistent, agreed-upon metrics to determine what exactly is being advanced, and this will be the challenge for users and their vendors. For example, for users of Zendesk’s solutions, the business model defines success by “automated resolutions; when AI fully resolves a customer issue end-to-end, without human intervention,” Chris Donato, president and chief revenue officer at Zendesk, told ZDNET. “It’s a measurable, accountable way to tie pricing directly to results.”
A couple of months ago, we discussed the changes in the way software is being purchased, based on a McKinsey analysis that predicted that per-seat software licenses may soon be obsolete. Consumption- and outcome-based pricing models would be the basis for software charges, and much of the purchasing would be performed by AI agents.
This has far-reaching implications for the software market of 2026, as explained in a recent analysis from West Monroe, also forecasting the end of per-seat licensing. “AI is rewriting the economics of the entire software industry,” the report’s authors explained. At least 12% to 15% of enterprise IT budgets now go to AI, for one, and the market is likely to increasingly favor AI-native service providers.
“Those who build AI capabilities that deepen customer experiences will boost renewal rates and margins,” the study’s authors predicted.
Also: Will AI replace all software? Why GPT-5 emboldens the doomsayers
Another trend likely to be seen at both vendor organizations and their customer sites is leaner engineer teams, they add. At this point, 80% of engineers need to upskill for AI-driven roles over the coming year. “AI is reshaping how software gets built, and by whom. From automating code generation and testing to accelerating release cycles, AI is collapsing the traditional product and software development lifecycles.”
In this next phase, there will be a push to re-engineer the software development and deployment process, the West Monroe analysts added. “Rethinking processes, metrics, and training to support AI-first workflows that balance speed with governance. As AI amplifies human capability, leaner engineering teams will deliver more output with less overhead.”
While there may be leaner engineering teams, this doesn’t mean AI will be replacing human talent anytime soon, Donato emphasized. “Outcome-based models align incentives, reward efficiency, and give customers clearer ROI than ever before. When done right, this shift doesn’t replace people, it elevates them. AI handles the routine, while human teams focus on deeper customer engagement and innovation. That’s the future of software value, and it’s already here.”
This means moving beyond simply sending a check to a vendor every month or quarter. “It is breaking the old assumption that software costs are predictable and fixed,” said Ed Barrow, CEO and co-founder at Cloud Capital. “As spend starts to move with activity levels and AI-driven usage, companies need new ways to plan, forecast, and stay accountable in real time.”
Also: Will AI replace software engineers? It depends on who you ask
The change will theoretically bring about closer collaboration between finance, product, and engineering teams operating together, “since usage and cost are now connected at every level,” Barrow explained. “Many existing systems cannot keep up with that pace of change.” It means stronger connections between financial data and product data, he added.
It’s already happening on the ground for many digital-native companies. “AI is redefining how software delivers value and how customers should pay for it,” said Donato. “Seat-based pricing made sense when humans were the primary users. But as AI agents handle more of the work, outcomes, not access, are becoming the clearer, more results-driven way to measure value.”
One company, Cozmo AI, has been working with an insurance company with such an outcome-based revenue model.
“The SaaS model was built for humans using tools to improve productivity,” said Alok Kumar, CEO and co-founder at Cozmo AI. “They don’t pay per user anymore, you pay per outcome — per claim closed, per premium renewed, per payment recovered. And the performance of AI will be judged by KPIs such as accuracy and conversion, which is the same way we evaluate people.”
Here are ways to prepare for this emerging world of software pricing:
1. Seek strategic partnerships – not just vendor relationships
In the AI era, vendors need to serve as partners in continuous value creation. “Seek providers who can serve as AI enablement partners that emphasize knowledge transfer, joint innovation efforts, and shared investment in AI solutions,” the West Monroe team recommended.
2. Seek visibility and control
“Look for providers who make AI performance and usage transparent,” said Donato. Also, look for vendors that will “provide spend forecasting and alerts, and help you benchmark success. These are not one-time contracts, they’re living partnerships built on shared data and trust.”
3. Renegotiate outsourcing contracts
When it comes to outsourcing, money is flowing right out the door without accountability for the AI era. Many of today’s contracts “are based on labor hours and headcount, creating misaligned incentives as AI boosts productivity,” according to the West Monroe team. “Design contracts with shared savings models to ensure both parties benefit from AI-driven efficiency improvements and foster stronger partnerships.”
4. Build strong AI fluency across your engineering organization
Track metrics such as “the percentage and quality of code developed using AI tools,” the West Monroe analysts also suggested. “Encourage hands-on experimentation to boost confidence and foster grassroots adoption since engineers often approach tools differently. Standardizing AI tool usage requires significant training and ongoing enablement efforts.”
Comments are closed, but trackbacks and pingbacks are open.