How to Win Public Transport Tenders: Lessons From the Bid Table

I spent 10+ years as a commercial lead in public-transport technology, across payments, ticketing, MaaS and DRT, and across bus, rail and ferry. I have also lost plenty and I know for sure that the losses taught me more than the wins ever did. That’s a true statement, we learn more from our losses than our wins (Samuel Smiles…)

Here is the uncomfortable truth I took from all of it! The win is so often decided before the writing starts. I see teams reacting and moving when a tender drops. The best are proactive and work before the tender document is a reality.

economAIcs is the AI bid-intelligence platform for public transport tendering. I founded it because of everything in this article. But this piece is not a product pitch. It is what I wish someone had told me at bid number ten instead of bid number 300.

Where are public transport tenders actually won and lost?

Before a response is written, in the engaging, room reading and self preparation most teams skip.

Watch how a typical bid team spends its time. Something like 80% of the effort goes into the writing: drafting, formatting, wordsmithing method statements late into the night. Almost nothing goes into reading the authority, or if it does, it’s done by a Business Development Team, with insight that doesn’t always flow between departments.

When a tender landed on my desk, I read it in a fixed order. First, the evaluation criteria and weightings, because that is the actual brief. Not the background section, not the glossy strategy preamble. The scoring table tells you precisely what this authority values and in what proportion. Second, the authority’s recent awards, because past behaviour tells you how they really score, whatever the document says. Third, my own track record against that profile, because the right past win is the strongest evidence you can put in front of a panel. Only then did I think about words.

Do the reading properly and the writing becomes the easy part. Skip it and no amount of polished prose will save you.

How do contracting authorities actually score bids?

On evidence against published criteria, marked by a panel whose behaviour is more predictable than most bidders think.

Evaluators are people with a scoring matrix and a stack of responses to get through. I do wonder how much AI is being used nowadays on the buying side of the table however?

The bids that score well make the marker’s job easy: they answer the question that was asked, in the structure it was asked in, with evidence attached to every claim.

That last part is where most responses fall down, ‘We are committed to service excellence’ scores nothing. A sentence like ‘on a comparable network contract, punctuality moved from 89% to 96% in eighteen months’ scores, because an evaluator can defend that mark to a moderation panel. One condition, the contract and the numbers have to be real and yours. Adjectives are free and every one of your competitors is using the same ones. Evidence is scarce, and it wins.

And here’s a note from me for free. General AI tools are amazing at creating fancy adjectives, but are rarely tuned to build answers from your deep evidence base. So whilst they write faster, they aren’t writing better.

And evaluation is getting more professional, not less. The Procurement Act 2023, in force since February 2025, has pushed transparency and justification further up the agenda³. The direction of travel is clear, panels that document their reasoning and expect suppliers to prove what they claim.

Why do good teams lose bids they should win?

In my experience there are three ways and I have the scars from them.

They start too late. By the time a tender appears on a portal, the winning bidder has often been preparing for months. Contract end dates are public. Re-tenders are largely predictable. If the first time you think about a contract is the day the notice is published, you are already behind someone.

They lose it internally. I lost a bid in Ireland that still stings. We had the better solution. Sales and the technical lead spent the bid window arguing over price and specification, with neither of us having the one thing that could settle the debate…data. The authorities never saw our best answer. They saw a compromise that nobody internally really believed in. That bid taught me that historical analysis and deep data are not nice to have in bidding. They are how you stop losing to your own organisation.

They cannot find their own winning answers. The worst bid I ever ran, we lost on a question we could have aced. We had won an almost identical contract eighteen months earlier. Nobody could find the answer in time. It was sitting in a folder on a laptop that had left the building with the person who wrote it. That is the real cost in transport bidding, processes are long, contracts are longer, the speed of personnel change can outpace the speed of procurement. The wins walk out the door when senior people leave and the next bid starts from memory instead of evidence.

How big is the UK transport tendering market?

Bigger than most people inside it realise. Operators are competing for well over £5bn of contestable work a year and that is a conservative figure.

English local bus operations alone turn over £6.6bn a year¹. Councils spent £2.3bn on home-to-school transport in 2023-24². Add rail operating contracts and wider local-authority transport on top and the annual prize comfortably clears £5bn.

Now put that number next to how the work is actually chased. When I moved deeper into this world I assumed there would be serious systems underneath it: deep analysis, structured data, proper tooling. What I found was Excel spreadsheets and people trying to remember what they wrote on a similar bid three years ago. Billions in contract value, decided by the sharpest commercial operators I have ever worked with, running on templates and memory. Smart people, lazy tools. That is the gap.

(For how franchising is redrawing this map, see my piece on bus franchising.)

How is AI changing public transport tendering?

More carefully than the hype suggests, because in transport procurement the audit is the job.

I can tell you a story about that. I was on a golf course and a transport director told me he had nearly submitted a bid that would have lost his company £500k. A general AI tool had invented an average driver salary when asked. It was 7% off the truth. He caught it 24 hours before submission. Most people would not have. AI in bidding without grounding is a £500k mistake waiting to happen.

So the interesting shift is not AI that writes your responses faster. Writing was never the bottleneck. The shift that matters is AI doing the intelligence work: watching the market so you see the winnable tenders early, modelling how each authority behaves and scores, and holding your own bid history so the next response starts with years of structured, findable evidence instead of a frantic search. I call that agentic bid intelligence, and it is the idea the Brain is built on: your own AI, calibrated to your wins, your evaluators, your blind spots, with every output source cited.

That last clause is the entry ticket, not a feature. In a market where evaluators buy on evidence, a confident wrong answer does not save you time. It loses the contract and the trust behind it. Source cited or it does not ship. If you are weighing up tools in this space, I have written a separate field guide to choosing AI bid software for transport that covers exactly what to ask.

For me, AI has to be vertically powerful, not horizontally generic. Transport is a niche industry, nuanced, delicate, intricate. The AI tool you use to bid has to be trained on the transport market. Has to be tuned to the policies and processes that govern the transport market. Has to be skilled in connecting the minute dots across a large authority that happen in the world of transportation.

What would I tell a bid team starting Monday?

Do the reading before the writing, every time. Know the contract end dates in your target geography and start shaping months before the notice. Settle internal arguments with data, not seniority, because the authority can smell a compromise. Guard your bid history like it is revenue, because it is: your past wins are the strongest evidence you own, and they are walking out of the building faster than you think. And if you use AI anywhere near a submission, hold it to the same standard an evaluator will hold you to. Receipts or nothing.

None of this is complicated. Almost nobody does it, because doing it manually is brutal. That is the problem I am building economAIcs to end. YourTransportBrain. The memory your entire bidding operation runs on.

If you bid for public transport contracts and any of this stings, we should talk.