You won it: the first ninety days of a public contract

Winning is the halfway point. The first three months decide whether the contract makes money, whether the relationship works, and — now that performance on larger contracts is published — what the rest of the market will read about you later.

Settle the administration while everyone is still pleased

There is a short window at the start of a contract when goodwill is high and nobody is annoyed about anything yet. Use it for the dull things that become disputes later:

• Invoicing — the exact entity, purchase order arrangements, the portal, the reference, and what evidence must accompany a claim.
• Named contacts on both sides, with deputies, and an agreed escalation route.
• Reporting — what, in what format, how often, to whom, starting when.
• Access and dependencies — sites, systems, security clearances, information you need from them and by when.

Write it down and send it. A short mobilisation note that says "this is what we agreed" is worth more later than any number of remembered conversations.

Agree how the KPIs will actually be measured

The indicators are in the contract; the measurement usually is not. Establish the data source, who produces the figure, when it is reviewed and what happens if you disagree with it. Where a measure depends on something the buyer controls — access, sign-off, information — record that dependency now rather than arguing about it at the first review.

Then keep your own record in parallel. Not to fight with, but because a rating you can evidence is a rating you can correct.

Deliver the boring promises first

Bids contain commitments that are easy to make and easy to forget: a named manager, a mobilisation meeting, a social value pledge, a review cadence, a system to be implemented. Buyers remember them, and the ones that quietly do not happen shape the relationship more than anything you deliver well.

Extract every commitment from your own bid into a checklist with owners and dates. It is a slightly uncomfortable exercise and it prevents the most common form of contract disappointment.

Expect the first invoice to be rejected

It very often is, for a technicality — wrong entity, missing purchase order, absent evidence. Send it early enough that a rejection still leaves time, and treat the first cycle as a test of the process rather than a payment. Once one invoice has gone through cleanly, the rest usually follow.

Start thinking about the re-tender in month three

The contract has an end date and it will be advertised again. The evidence you will want then — performance data, satisfaction, savings delivered, social value actually provided — is far easier to collect as you go than to reconstruct in the final quarter. Note the likely re-tender date now, and keep a running file of what you would put in the bid.

Get the contract manager relationship right early

The person who manages the contract day to day is rarely the person who ran the procurement, and they inherited you rather than chose you. That relationship determines how problems are handled for the whole term, and it is established in the first few weeks.

Meet in person if you can. Ask what went wrong with the previous supplier and what they most want to be different — the answer is usually specific, rarely in the specification, and tells you what "good" means to the person who will rate you. Then do that thing visibly in the first month.

Handover from an incumbent is a risk, not a formality

Where you are replacing another supplier, assume the handover will be less complete than planned. Outgoing suppliers are rarely enthusiastic and often no longer resourced for it. Identify early what you actually need to receive — asset registers, historical data, keys and access, system credentials, outstanding works, staff information where employment rights transfer — and put dates against each, with the buyer copied in.

Where staff transfer with the contract, take advice early. Employment obligations on transfer are a specialist area, the timescales for consultation are real, and getting them wrong is expensive in a way that dwarfs the contract margin.

Watch the first variation

Requirements move. Someone will ask for something outside the specification, usually informally and usually early, and how you handle the first one sets the pattern for the rest of the term. Do the reasonable thing, and record it: what was asked, what you did, and whether it sits inside the contract or outside it.

Suppliers who never say no end up delivering an unpriced contract; suppliers who refuse everything become difficult to work with. The workable position is to be generous and to keep a written record, so that when the cumulative change is material there is a factual basis for the conversation rather than two conflicting memories.