Everyone has AI now. Only the best are winning more because of it.

Sixty-two percent of teams now use AI to generate RFP answers.¹ Submission volumes have climbed to 166 per year.¹ Response windows have halved.² Win rates have fallen.¹ Everyone added more intelligence to the process. The process responded by demanding even more.

The Floor Keeps Rising

The boardroom assumption is simple enough: AI makes responses faster, so we need fewer people. The data disagrees. Enterprise RFPs now contain 200 to 500 questions, up dramatically from five years ago.² Bandwidth is the number one challenge for bid teams for the first time ever.¹ And despite a 16-point surge in AI adoption in a single year, win rates have dropped.

AI didn’t reduce the demand for intelligence in bidding. It increased it significantly. When AI started responding, procurement raised their expectations in lockstep. More questions. Tighter deadlines. Deeper compliance. More suppliers crossing the threshold and competing to win. Every efficiency gain from the chatbot wave was almost instantly absorbed by a system that expanded to consume it.

Running Faster, Standing Still

Biologists call this the Red Queen Effect. You run faster just to stay in the same place. Nearly every team has ChatGPT/Claude/CoPilot producing the same polished prose. When everyone sounds credible, credibility becomes the baseline, not the differentiator.

Go/no-go discipline has dropped eight points in a year.¹ Teams are chasing volume, and many haven’t thought about going deeper. You need to go fast, sure, but Jevons paradox demands you reinvest your “savings” to continue to compete.

Where the Intelligence Moves

We have seen this before. When Excel launched, it automated the manual work that sustained hundreds of thousands of bookkeepers. Entire floors of clerks were eventually replaced by one person with a spreadsheet – but the floor didn’t empty. It expanded.

The U.S. alone today employs over 1.6 million bookkeeping clerks³ and 1.6 million accountants on top of that.³ Automation eliminated manual tasks but unlocked entirely new categories of financial work: modelling, forecasting, strategic advisory, financial analysis. Companies ran exponentially more numbers and needed far more people to interpret them. The demand for financial intelligence grew faster than automation could absorb it. Excel grew accounting services; it didn’t eliminate them.

Bidding is following the same path. AI generates a compliant first draft. That instantly becomes table stakes. The intelligence that wins sits elsewhere: curating knowledge so AI output is accurate and differentiated, coordinating experts so the right insight reaches the right question, and governing content so AI amplifies truth rather than hallucination.

The bid team is becoming the organisation’s intelligence layer. Not writing answers, but ensuring collective knowledge is structured, current, and accessible enough for AI to use well. Leveraging AI not just to draft, but to push the technology to win.

The Next Big Thing

My response to Jevons paradox is not to resist it, but to reinvest deliberately. Treat the time AI gives you back as capital. Allocate it upstream into capture strategy, client intelligence, and knowledge curation. The teams that spend their AI dividend on submitting more average bids will burn out. The teams that spend it on thinking harder about fewer bids will win disproportionately.

The good news is that most of your competitors haven’t figured this out yet. They are still measuring AI success by how fast they can produce a first draft. That gives you a window. Use it to get serious about your knowledge infrastructure, upskill your team from writers to intelligence architects, and start treating every bid as a chance to build cumulative organisational advantage. The window will not stay open forever.

¹ Loopio, RFP Response Trends & Benchmarks (2026)² Arphie, RFP in IT: 2026 Data (2026)³ U.S. Bureau of Labor Statistics, Occupational Outlook Handbook (2024)