Pegasystems CFO Ken Stillwell told investors at a Canaccord fireside chat that the company is expanding its AI capabilities and resetting its sales approach after first-half 2026 annual contract value growth fell short of expectations. The workflow software maker is counting on improved pipeline building and new AI tools to recover momentum in the second half.
Stillwell said Pegasystems serves large enterprises that need to manage structured, regulated workflows - dispute management, loan origination and similar processes - that commercial off-the-shelf software can't readily handle. Customers use Pega's platform to configure applications rather than building custom software, which he said can create technical debt and complicate change management. He counted Salesforce, Microsoft and Adobe as competitors in certain use cases, along with internally developed software.
New AI tools target faster app development
Pega Blueprint, the company's design tool, is meant to reduce the lengthy front-end process of helping customers identify and visualize applications they want to modernize. Users define their industry, problem, personas and potential integrations to produce a visual representation of a proposed workflow application. In some cases over the past 12 months, Stillwell said, customers using Blueprint created a use-case template and configured an application for deployment in fewer than 90 days.
The company recently made generally available Pega Infinity 2026, which includes Pega Infinity Studio. That product lets customers import a Blueprint design into a development environment and use AI to help complete the application. Stillwell said the company had given a beta version to roughly 20 customers for about three months before release and used their feedback to refine the product.
Customers are also interested in connecting AI models to Blueprint, he said. Pega supports connections to different models through MCP connections and offers native models within Blueprint. Building workflows through prompts and discussion is a newer experience for many Pega customers, who historically used drag-and-drop development methods.
Fixed AI pricing and the ACV miss
Pega charges customers a fixed AI-enabled price for a unit of work rather than billing per token, Stillwell said. The pricing model is intended to give customers cost certainty while putting responsibility for managing token consumption on Pega. The company's architecture helps manage that risk by determining where AI is needed and selecting an appropriate model for each task - not every activity requires a frontier model, he said, citing automated customer-service call wrap-up as work that could use a less resource-intensive model.
"Our job is to help our clients to only use AI when it is needed to be used, and then when it is used, to use the right model," Stillwell said.
Stillwell described first-half ACV growth as "unimpressive" and disappointing. He attributed the performance to management complacency after a strong start to 2025, insufficient pipeline-building late last year and a slower-than-needed shift in the sales organization from a "farmer" mentality to a more proactive "hunter" approach. He also said enterprise buyers were distracted by AI during the first half as vendors broadly promoted AI products. The company saw in March and April that its sales activity measures were not progressing sufficiently.
Pipeline recovery and cash flow outlook
Pega's pipeline entering the second half was significantly higher than a year earlier and exceeded the level needed to meet its back-half growth target, Stillwell said. Many financial-services customers that had been focused on AI governance and compliance have since established guardrails, control processes, model choices and, in some cases, token-spending budgets. Sales activity measures improved "dramatically" over the prior six weeks, he said.
Given lower bookings in the second quarter and its normal seasonal pattern, the third quarter could produce slightly negative cash flow, Stillwell said, while the fourth is expected to be strong. If the company doesn't recapture its first-half ACV growth shortfall, its full-year cash flow could face pressure. He characterized full-year cash flow as likely to be relatively flat year over year. Management maintained its framework for achieving more than $700 million in free cash flow by 2028, which depends on double-digit ACV growth in 2027 and 2028 plus operating leverage.
On AI agents, Stillwell said Pega's workflow technology can provide structure and governance around how agents execute tasks, particularly in regulated activities where processes must be completed in a prescribed sequence. He said AI is useful for testing, data analysis, extraction, analytics and tactical coding, but companies remain cautious about allowing agents to generate code or take actions beyond what humans can effectively supervise.
"Left unstructured and uncontrolled, AI will do varied things," Stillwell said. "Some good, some very bad."
Why this matters for sales professionals
For sales teams, the core lesson is that a strong AI product doesn't automatically translate into bookings if pipeline generation lags. Pega's own management acknowledged its sales organization was slower than it should have been to shift from maintaining existing accounts to actively hunting new buyers. The company's recovery plan - building pipeline early, monitoring sales-activity metrics weekly and responding to customers' delayed AI adoption decisions - is a working example of how enterprise sales leaders can adapt when buyers pause purchasing decisions during rapid AI transitions. Pega also signals that sellers who want to win AI-era deals should be ready to talk about AI's cost structure, their AI adoption curve, and why your app is also expected to be popular for routine tasks. For an overview of how AI is reshaping sales work, see AI for Sales. For more on how leaders are approaching sales strategy in this period, see AI for Executives & Strategy.
Your membership also unlocks:
