Why Clients Now Expect Faster Project Delivery

Jeanelle Johnson runs PwC's Washington DC practice, more than 2,200 consulting and accounting staff who report to her.

When she spoke with Business Insider this month, she said clients now expect faster project delivery than they did a year ago, and by a specific margin. Work her team once quoted at eight to ten weeks now gets held against half that.

"Something that we might have said takes eight to 10 weeks, the expectation is probably half of that," Johnson said. "The shelf life is half."

She was careful about what that means. Clients aren't asking for a full transformation squeezed into a month. They want pilots and proofs of concept fast, because AI tools have set a new baseline for how quickly a first version should exist.

Asked whether the compressed timelines were straining her teams, Johnson didn't dodge it. "We're working on it," she said. That's a rare admission from a managing partner.

The New Baseline For Faster Project Delivery

Johnson's clients aren't comparing her firm's speed to a competitor's. They're comparing it to what AI tools already let them do themselves.

Once someone has watched a chatbot draft a first-pass analysis in minutes, an eight-week engagement for similar work reads as slow by default, whatever the quality difference actually is. That comparison is unfair in plenty of cases. It's also not going away.

The Firms Already Built For the New Speed

The 2026 Professional Services Maturity Benchmark, co-published by Rocketlane and SPI Research, surveyed 509 professional services organizations representing over 245,000 employees.

It put numbers behind the gap between firms using AI widely in service execution and firms that aren't.

The same report found generative AI touched 27.1% of professional services projects in 2025, up from 19.3% the year before. That's a fast climb for an industry that's historically slow to change how it works.

Firms hitting those on-time numbers typically rebuilt the delivery process around AI, rather than adding a chatbot on top of the one they already had. That's close to how our AI and automation work with clients actually starts: fixing the workflow, then automating it.

Speed That Breaks Things Isn't the Win It Looks Like

Photo by Javier Piva Flos on Pexels

Google's DORA team found the same acceleration on the software side, with a catch worth sitting with.

Its September 2025 State of DevOps findings showed AI coding tools no longer dragging down delivery throughput, reversing a 1.5% slowdown per 25% rise in AI adoption that the 2024 survey had recorded.

Stability didn't follow the same curve. The 2024 data showed a 7.2% drop in release stability for every 25% increase in AI adoption, and the newer survey found stability still falling, just measured differently.

DORA lead Nathen Harvey put it plainly: "It's no surprise that we see throughput starting to inch up first before instability goes down."

Cutting a timeline in half by skipping review or testing gets a faster first delivery and a rougher second one. The firms in the Rocketlane data posting both higher on-time rates and higher margins got there by changing the process, not by removing the checks that used to slow it down.

If your business quotes timelines the way PwC does, most teams can already deliver faster once.

The harder test is whether the pipeline behind that speed holds up on the tenth project, which is the kind of build-out we handle inside apps and SaaS engagements when a client's internal tooling can't keep pace with what they're now promising their own customers.

Firms still treating AI spend as a side experiment rather than a delivery-process rebuild are worth weighing against what OpenAI's spending means for AI vendor strategy before committing budget the same way.

Cover photo by wal_172619 on Pixabay