Pega Targets $2bn Revenue Milestone as It Bets on 'Predictable AI' - CEO Alan Trefler Slams 'Delusional' Multi-Agent Approach

Why It Matters

The shift toward Blueprint‑based AI underscores enterprise demand for audit‑able, repeatable processes, reshaping competitive dynamics in the SaaS and AI markets.

Key Takeaways

• Pega Cloud ACV grew 33% YoY, driving revenue outlook
• Blueprint AI creates design‑time workflows for predictable runtime
• CEO warns against deploying thousands of autonomous AI agents
• Free cash flow rose 45% to $491 million, backlog hit $2 billion
• Guiding $2 billion revenue by 2026, adding $1 billion buyback

Pulse Analysis

Pega’s latest earnings reveal a rare combination of strong top‑line growth and disciplined cash generation. The 33% year‑over‑year acceleration in Pega Cloud ACV, coupled with a 45% jump in free cash flow to $491 million, pushed the contract backlog past $2 billion and set the stage for a $2 billion revenue target in 2026. Investors, however, remain cautious, as the share price fell 5% after hours, reflecting lingering worries about slower overall revenue momentum and shrinking net income.

At the heart of Pega’s strategy is its Blueprint AI design agent, which embeds generative AI into the workflow‑design phase rather than allowing autonomous agents to reason at runtime. Trefler argues that this model delivers the predictability and auditability required in regulated industries, where consistent outcomes are non‑negotiable. By contrast, competitors that push large fleets of AI agents risk erratic responses and governance challenges, a risk Trefler labeled as "delusional" for enterprises that need reliable process execution.

The broader market implication is a potential re‑alignment of AI investments toward hybrid solutions that balance innovation with operational stability. As SaaS companies grapple with the so‑called "SaaS‑pocalypse," those that embed change‑ready architectures—like Pega’s Blueprint—are better positioned to weather competitive pressure from pure‑play AI vendors. If the industry continues to prioritize reliability over sheer flexibility, Pega’s bet on predictable AI could set a new benchmark for enterprise software providers aiming for sustainable growth.

Pega targets $2bn revenue milestone as it bets on 'predictable AI' - CEO Alan Trefler slams 'delusional' multi-agent approach

Image taken by diginomica author

Pegasystems beat its annual guidance with Pega Cloud annual contract value (ACV) accelerating 33 % year‑over‑year as reported (28 % in constant currency), positioning the company to cross the $2 billion revenue threshold in 2026. But investors, as they tend to be, weren’t entirely impressed, as shares fell 5 % in after‑hours trading. The cause was likely reflecting concerns about revenue‑growth deceleration and declining net income year‑over‑year, despite the positive earnings.

The Q4 2025 earnings call saw the company doubling down on an AI strategy that puts workflow automation ahead of prompt engineering or over‑reliance on frontier models, with CEO Alan Trefler delivering some of his sharpest criticism yet of competitors’ multi‑agent approaches. Trefler thinks that releasing agents into the wild, where prompt instructions and runtime reasoning are applied on the fly, is a fundamentally misguided approach in the enterprise, with Pega favoring its reliable, predictable, ‘Blueprint’ approach.

Agent proliferation

During the analyst Q&A, Trefler was as animated as ever and said that enterprises deploying thousands of individual AI agents are heading for trouble. When asked about Pega’s role as enterprises start live‑deploying agents, he didn’t hold back:

“I don’t expect customers to be installing tens of thousands of agents. I think the people who want to install tens of thousands of agents are delusional. We went through a parallel environment years ago around interfaces. People talked about microservices, and the question was, how many of these microservices connect your enterprise? The reality is, the people who put too many of them found that they went out of control.”

The dig at competitors became more pointed when he added:

“I think having an agent control tower to control your agents tells you something about the architecture, which is not a good thing.”

While Trefler didn’t name names on the call, the implication is clear: ServiceNow’s governance and AI‑tower approach. He believes vendors positioning around orchestrating large numbers of agents with control‑tower capabilities are revealing architectural limitations rather than strengths.

Trefler’s core argument centers on predictability versus runtime reasoning. He explained that competitors are “taking generative AI models and using them at runtime” – meaning the model reasons from scratch every time a user engages with the system. The problem, according to Trefler, is reliability:

“Our competition rethinks the problem from scratch over and over again. And the slightly frightening thing is, the models don’t always come up with the same answers, even in situations and regulated industries where coming up with the same answer is not just important, it is imperative.”

Instead, Pega uses what it calls its Blueprint AI design agent to create workflows at design time, which then run predictably at runtime. The distinction is that AI creativity happens during the design phase – where humans validate and approve – whilst execution follows predetermined workflows. As Trefler put it:

“Businesses have processes that run them, and they want those processes to be respected. They want architectures that will be able to do reliable, repeatable — our favourite word — predictable things.”

The numbers behind the strategy

Pega’s positioning comes alongside a solid set of earnings and clear financial execution. Total ACV grew 17 % year‑over‑year as reported, beating guidance. But the real story is in the cloud acceleration – Pega Cloud ACV’s 33 % growth represented a significant pickup from last year’s 18 % reported growth rate.

COO and CFO Ken Stillwell attributed the performance to three factors:

1. Blueprint moving from “a promising experiment in 2024 to a fundamental change in how we sell in 2025.”

2. The strongest global sales execution the company has had.

3. Increasing client demand for “predictable AI agents integrated into proven enterprise workflows.”

Free cash flow was up 45 % year‑over‑year to $491 million, exceeding guidance by $51 million. The company’s contractually committed backlog crossed $2 billion for the first time, growing 28 % year‑over‑year as reported. Pega Cloud backlog now represents 74 % of total backlog, up from previous periods.

Perhaps most telling for the Blueprint strategy’s impact was how Trefler highlighted that new sales staff are now productive “in a month or less” compared to the five to six months it used to take. He said:

“A lot of that is that Blueprint just makes it so easy for them to get it and for them to explain it to their clients.”

For 2026, Pega is guiding to total revenue of $2 billion (approximately 15 % growth), 15 % ACV growth in constant currency, and $575 million in free cash flow (17 % increase over 2025). The company’s board also authorized an additional $1 billion in share‑buyback capacity, reflecting confidence in cash‑flow durability.

The SaaS survival argument

Trefler also weighed in on what he called the “SaaS‑pocalypse” – the recent market punishment of software companies. When asked which characteristics would determine software‑company survival, Trefler argued the “death of SaaS may be somewhat exaggerated” but also acknowledged pressures.

He argued that software companies that are “basically glorified spreadsheets with limited functionality” are vulnerable because “you can do all sorts of magical things in Claude or even Copilot that enable you to go after those types of applications.” Companies building systems that need to adapt and change will continue to succeed. Trefler said:

“But the thing that I would say is most important is: Is the system built for change? Because the problem with these code‑based systems that are attacking the SaaS world is they don’t have a particularly visible architecture. They’re just kind of written. And, you know, going after somebody else’s 3,000 modules of code is incredibly daunting and very, very difficult to do correctly.

In our world, because you can see the Blueprint, we have so much scaffolding and infrastructure. We have the idea of a case. We have the idea of stages. We have the idea of steps. We have the idea of service levels, of personas. Our systems are built around the business entities of an organization. And because they’re built that way, it’s possible to navigate them and, as a result, possible to change them.

Businesses that require change, I think, are going to be the ones that are going to be most interested in a technology like ours. And the businesses where you just write something that’s going to sit on the shelf for two, three years or months—those are the ones that I think are going to be most vulnerable.”

My take

Trefler is always entertaining. He’s a fantastic CEO to listen to and engage with – because whether you agree with him or not, he has a strong position and can argue it effectively. On Blueprint, the positioning is grounded in a very clear enterprise reality – businesses do need predictable, auditable processes, particularly in regulated industries. I’ve been pointing this out a lot myself lately. The Blueprint approach of using AI creativity at design time whilst maintaining workflow predictability at runtime is clever, and it tackles the problem of how to get the benefits of AI without the risks of agents going rogue in production.

That said, there’s a risk that Pega’s stance will appear a touch conservative. As consumers become more comfortable with prompt‑based problem solving through tools like ChatGPT and Claude, employee expectations may shift. We’ve seen this pattern before with mobile and SaaS – consumer adoption pushes into enterprise expectations. If competitors successfully deploy thousands of agents with acceptable reliability, Pega’s workflow‑first approach might look like it’s optimizing for yesterday’s enterprise rules and concerns. I remember having a conversation with the CEO 10 years ago about Pega’s cloud strategy and Trefler was also conservative then about how many companies would actually want to move off‑premise. Some of this may be cautiously meeting customers where they are, rather than running ahead – but Blueprint feels more like a long‑term architectural decision.

What’s certain is that Pega is betting its $2 billion milestone on enterprises prioritizing reliability over flexibility in their AI deployments. Given current buyer nervousness around AI adoption, that might be exactly the right call for 2026. Whether it’s the right bet for 2027 and beyond remains an open question.