AI Agents Are Poised To Take Over Software Buying, But Are We Ready?

Let AI brokers do all of the scoping?

getty

Not too way back, the way in which we purchased software program was, let’s consider, Netflixed. As an alternative of receiving disks from a supplier to load and carry on a shelf, it was transmitted and up to date over the wire, a part of a cloud-type supply mechanism. Now, the trade goes by way of one other transition, past Software program as a Service.

“AI software program is now writing extra software program,” tech entrepreneur and investor Anthony Pompliano just lately identified on an X/Twitter post. “Humanoid robots are going to ultimately manufacture and assemble extra humanoid robots. This concept of exponential productiveness is one thing now we have by no means actually seen in our lifetime. The influence might be a lot bigger than all of us anticipate.”

So, will software program even be deciding, buying, and putting in new software program as properly? That day is already right here, and may assist ship a few of that exponential productiveness, in addition to change the composition of an trade.

AI brokers are beginning to dimension up software program specs and purchases for us, as defined in a McKinsey report issued in October, suggesting that conventional SaaS licensing fashions might go the way in which of set up disks. Per-seat pricing fashions are being supplanted by agent-to-agent interactions that base software program worth on outcome- or usage-based fashions.

“We’re getting into a brand new period the place AI brokers turn out to be the first customers of SaaS,” in accordance with Vara Kumar Namburu, co-founder and head of R&D and options at Whatfix, which helps enterprises handle software program lifecycles. “AI brokers gained’t exchange SaaS or people; they’ll redefine each—automating the routine, amplifying decision-making, and guaranteeing simplicity and management stay on the heart of digital work.”

AI brokers will “autonomously execute workflows, interpret intent, and collaborate throughout methods to ship outcomes. Very like self-driving automobiles now share the identical roads as human drivers, we’ll quickly see AI brokers and people sharing the identical digital interfaces, working aspect by aspect,” Namburu believes.

This transformation is reshaping how software program is purchased and valued, he continued. “The per-seat mannequin – constructed for a human-user world – will regularly give solution to usage-based pricing that mirrors actual exercise and measurable worth creation.”

This rising strategy is inverting the economics of software program buying, “Conventional SaaS is dear to construct however has near-zero marginal prices, whereas AI is cheap to develop however incurs excessive, variable operational prices,” mentioned Michael Mansard, principal director at Zuora Subscribed Institute, a assume tank for the subscription economic system. “These shifts problem seat-based or feature-based fashions, since they fail when worth is tied to what an AI agent accomplishes, not how many individuals log in.”

Inside such a framework, vendor revenues are derived from “outcome-based fashions, comparable to paying for a proportion of financial savings or good points,” Mansard added. “Final result-based provides a fairer mannequin, aligning incentives and transferring efficiency danger to the software program supplier. For instance, Zendesk fees its customer support agentic AI $1.50 per precise case decision.”

Presently, fewer than 10% of AI providers are monetized this fashion, Mansard continued, “given how complicated it’s to each show and function. For the foreseeable future, we’ll see a spectrum of fashions, comparable to by exercise – AI work hours — or by output – AI deliverables.”

The ramifications of post-SaaS go properly past pricing, mentioned Dor Sasson, CEO and cofounder of Stigg.io, a software program billing service. “It’s a permissioning downside. Within the SaaS world, people purchased entry: seats, credit, outcomes. Within the agentic world, software program buys autonomy. AI brokers don’t care about seat limits or utilization tiers; they care in regards to the license to behave, to learn knowledge, execute workflows, and make choices inside methods.”

For that purpose, per-seat and even per-usage pricing gained’t work on this new period. “AI brokers don’t enroll. They transact,” mentioned Sasson. “The alternative isn’t simply one other monetization hack; it’s a brand new contract between distributors and software program itself – agentic licensing.”

Nonetheless, agentic licensing gained’t be showing in a single day, warned Bryan Murphy, CEO of Smartling, a content material translation and globalization service. “The transition is messier than the frameworks recommend. The true problem isn’t the pricing mannequin itself, it is that the majority SaaS firms bolted AI onto merchandise designed for a unique period.”

Murphy added that he’s been by way of sufficient know-how waves – from on-premise to cloud to SaaS – “to acknowledge when the trade is hitting an inflection level, and this one’s totally different. Most enterprises are caught on this awkward center floor: they know per-seat pricing does not work when brokers are doing the work people used to do, however switching to consumption or outcome-based fashions requires rethinking your whole price construction and ROI calculations.”

Getting ready for the post-SaaS agentic period “means rearchitecting your platform from the bottom up, retiring legacy options that don’t make sense anymore, and accepting that your product roadmap from 12 months in the past is now irrelevant,” Murphy defined. “For patrons, the near-term actuality means evaluating distributors on whether or not they’ve genuinely rebuilt for an agentic world or simply repackaged legacy instruments with a chatbot interface.”

To get a way of vendor preparation, it’s vital to ask distributors “how their pricing aligns with the precise worth you’re getting,” Murphy suggested. “If you happen to’re paying per seat however brokers are dealing with 60% of the amount, one thing’s damaged. Push for transparency on how consumption is measured and what you are really paying for.”

The winners on this transition “might be prospects who demand outcomes-based contracts and urge distributors to show ROI in concrete phrases, not theoretical productiveness good points,” mentioned Murphy.