Back to Perspective

The bid-writing AI race is solving the wrong problem

AI can accelerate proposal production. Its greater opportunity is to improve the evidence, decisions and commercial thinking that determine whether a pursuit deserves investment.

AI is changing how quickly teams can research, analyse and draft a response. For organisations pursuing public-sector business, that opens a larger question about how the whole pursuit operation should work.

Costly failures can begin before the response is written. An opportunity enters the pipeline because the revenue is attractive. Optimism hardens into commitment. Buyer and incumbent knowledge is incomplete. Evidence gaps emerge late. By submission, the team has committed substantial effort and has little room to change its position.

Start with the decision to compete

A pursuit draws on people, management attention, commercial expertise and capital. Each commitment should have a clear rationale.

What makes the opportunity attractive? What evidence supports the organisation’s ability to win and deliver? How strong is the incumbent? Which capabilities, relationships or proof points need work? What would justify the next investment?

Qualification should bring those questions together and make the assumptions visible. As evidence develops, the team should be able to revisit its position and explain what has changed. That decision shapes both the pursuit itself and the effort the organisation can afford to commit elsewhere.

Create advantage before the tender

Advantage develops through account knowledge, framework choices, early market engagement, evidence readiness, stakeholder understanding, partner strategy and commercial positioning.

A team that understands the buyer’s priorities and has prepared relevant evidence starts with more room to shape a convincing offer. Research and engagement should help the team make a decision, develop an advantage or prepare evidence it will need later.

That thinking carries through the lifecycle. Identify relevant opportunities, assess their viability, develop the path to win and connect the solution, delivery and commercial position before committing it to a response.

Use AI to challenge confidence

Ask AI to test whether confidence is justified.

What does the market evidence actually say? Which buyer needs can we trace to a source? Which proof points survive scrutiny? Where does the commercial position depend on a partner, a delivery assumption or an uncertain cost?

Specialist AI agents can assemble evidence, compare sources, identify contradictions and prepare analysis for review. Experienced people bring context and judgement. They challenge the outputs, resolve uncertainty and decide how to proceed. Calculations and decision rules should be visible and reproducible so the team can understand how it reached an assessment.

Retain the reasoning behind the outcome

Every pursuit creates knowledge about the buyer, the market and the organisation’s own ability to compete. Some of the most valuable learning sits in the reasoning. What did the team believe at qualification? Which assumptions changed? Where was the evidence weak? Why did the price move? What did the evaluator’s feedback reveal?

Keeping that trail lets teams examine an outcome against the decisions that produced it. An approved lesson should inform a method, an evidence requirement, a review question or a future pursuit decision.

Design the operation around that work

The opportunity for AI spans the whole pursuit lifecycle. For growing organisations, it can make specialist work more accessible. For established suppliers, it creates a way to examine how effort is allocated and where experienced people can contribute most.

This is the operating model Vela is designed to support. Specialist AI and experienced people contribute to the same pursuit, with clear responsibilities and human control. The question for leaders is how to organise that capability around the decisions that matter, so each pursuit starts with better evidence and each outcome improves the next.