Analysis · No. 14

Before the tender is written: where suppliers can still shape a contract

Public buyers asked suppliers for views before writing a tender 12,617 times in 21 months. About a third of the time they had not yet settled what to buy. Those are the contracts a small company can still shape. They are also slower, and fewer of them are ever tendered.

Seam analysis by Matt Bevington (how to cite) · first published 17 November 2026 · Find a Tender notices, 24 February 2025 to 16 November 2026 · How we did this

12,617procurements that began with a market engagement notice since the Procurement Act came into force
1 in 3were worded as exploratory: the buyer was asking suppliers what is possible
53%of exploratory ones changed their estimated value by more than a tenth before the tender. For defined ones, 40%
33%of exploratory ones had reached a tender. For defined ones, 52%

Three kinds of notice

Before a public body writes a tender, it can publish a notice asking suppliers to talk to it first. The law calls this preliminary market engagement. How the notice is worded tells you how much is still open.

We read the title and full description of every one of these notices for set phrases, and sorted them into three kinds. Seam now shows the result as a tag on each engagement notice in Contracts for you, on Home.

TagWhat the buyer is sayingTypical wordingShare of notices
ExploratoryWe have not settled what to buy. Tell us what is possible.“request for information”, “soft market testing”, “understand the range of solutions”, “feasibility”, “help shape the specification”31%
ConsultativeWe know roughly what we want. Tell us what you think, or whether you are interested.“questionnaire”, “seeking feedback”, “market capacity”, “engagement event”12%
DefinedWe know what we will buy, and a tender is coming.“intends to procure”, “is seeking to appoint”, “re-tender”, “forthcoming”19%
No tagThe notice describes the contract and uses none of these phrases.39%

The shares are of the 8,684 procurements whose engagement notice came out on or before 16 May 2026. Those are at least six months old, so each has had time to reach a tender. The rest of this piece follows them.

Exploratory means the contract is still moving

If a buyer is still deciding, the contract it finally tenders should look different from the one it first described. It does.

A notice can give an estimated value at the engagement stage and again at the tender. Where both were given, the figure changed by more than a tenth in 53% of exploratory cases, against 40% of defined ones. It went up about three times as often as down.

How often the estimated value changed before the tender

Share of procurements whose tender gave a value more than 10% above or below the engagement notice's. Only those giving a value at both stages.

  • Exploratory53%
  • Consultative47%
  • Defined40%
  • No tag40%

The name of the contract changed more often too: in 61% of exploratory cases, against 52% of defined ones. A changed title usually means a changed scope.

Buyers also chose a more open-ended way to run the tender. 31% of the tenders that followed an exploratory notice used the competitive flexible procedure, a route the buyer designs itself and which can include rounds of discussion with bidders. For defined notices it was 24%.

None of this proves that suppliers’ replies caused the changes. It does show that an exploratory notice marks a requirement that had not set, which is the point at which a reply can count for most.

The price is time, and a real chance of nothing

A buyer that is unsure what to buy is also unsure whether to buy. Exploratory engagement is the slowest route to a contract and the least certain.

How many had reached a tender

Share of procurements with a tender notice under the same reference by 16 November 2026. Engagement notices published on or before 16 May 2026.

  • Exploratory33%
  • Consultative42%
  • Defined52%
  • No tag49%
  • Fewer are tendered. 33% of exploratory exercises had reached a tender, against 52% of defined ones. 22% had been awarded, against 36%.
  • They take longer. The typical exploratory exercise took 126 days from engagement notice to tender. A defined one took 87: almost six weeks less.
  • The dates slip. When the buyer gave a date for the tender, the typical exploratory tender came out 39 days after it. The typical defined one was 17 days late.
  • Often there is no date at all. 44% of exploratory notices gave no expected tender date, against 36% of defined ones.

One caution. “No tender” here means no tender under the same procurement reference. A buyer that talks to the market and then buys through a framework, a pre-agreed list of suppliers, leaves no trail we can follow. Some of the exercises that appear to end in nothing ended in a purchase by another route.

Where buyers are least sure

Software stands apart. Almost half of software and IT engagement exercises were exploratory, far more than in any other field.

What was being boughtEngagement exercisesExploratoryReached a tenderAwards with a start-up or scale-up among the winners
Construction and property1,48327%45%12% (41 of 348)
Health and social care services1,35030%48%14% (56 of 409)
Software and IT1,19249%27%28% (50 of 181)
Professional and business services1,18227%47%10% (35 of 350)
Facilities, waste and environment78726%51%8% (21 of 252)
Medical and lab supplies46429%48%34% (45 of 132)
Transport and vehicles45829%46%21% (26 of 125)
Repairs and maintenance34317%47%13% (13 of 101)
Education and training34225%50%15% (17 of 116)
Defence, security and telecoms equipment30036%32%11% (5 of 46)
Energy and utilities26228%47%16% (12 of 76)
Everything else52133%45%24% (34 of 143)

Software also has the lowest share reaching a tender, at 27%. Part of that is real: buyers explore and decide not to proceed. Part is likely to be the framework effect described above: a good deal of public software is bought through pre-agreed supplier lists, and that would not show up here.

Defence and security equipment comes second for exploratory wording, at 36%, and is the other field where under a third reached a tender.

Shaping a tender is not the same as winning it

Start-ups and scale-ups win contracts that began with exploratory engagement at about the same rate as any other. The wording of the notice does not change who wins. What is being bought does.

Awards with a start-up or scale-up among the winners

Share of awarded procurements where at least one winner was under ten years old and passed Seam's start-up test. Only awards where a winner could be matched to Companies House.

  • Exploratory18%
  • Consultative17%
  • Defined15%
  • No tag14%

The gap between exploratory and defined is small enough to be chance.

The spread by field is much wider. A start-up or scale-up was among the winners in 34% of medical and lab supply awards and 28% of software awards, against 8% in facilities and 11% in defence equipment.

So an exploratory notice is an opening to influence what is asked for. It is not a shortcut to the contract. A company that helps shape a requirement still has to bid for it against everyone else, and may find the tender goes to a supplier list it is not on.

How to use the tags

  • Exploratory: reply if you want to shape it. This is where your view of what is possible can change what gets bought. Go in expecting a wait of four months or more, and a real chance that no tender follows. Ask the buyer which route it expects to buy through, so you can get onto the right supplier list in time.
  • Consultative: reply if you plan to bid. The buyer is testing its approach and who is interested. A short, practical answer puts you on its radar and can still change how the contract is packaged.
  • Defined: prepare the bid. The requirement is set and the tender is more likely to arrive, and sooner. Use the engagement to ask questions and meet the buyer.

Seam tags every live engagement notice that matches what you sell, with the reply date. Get your briefing

How we did this

Sources and method

The notices. We read every notice published on Find a Tender from 24 February 2025, when the Procurement Act 2023 came into force, to 16 November 2026. We kept each preliminary market engagement notice (notice type UK2) and every later notice carrying the same procurement reference, which is how a tender or an award is tied back to the engagement that came before it. That gave 12,617 procurements. The outcome figures use the 8,684 whose engagement notice is at least six months old.

The wording. Each notice’s title and full description were checked against three lists of phrases. An exploratory phrase wins over a defined one, and a defined one over a consultative one. Stock wording that almost every notice carries, such as “this does not commit the buyer to procure”, decides nothing. This is a phrase list, not a person reading each notice, so some are wrongly sorted. A buyer that titles every notice “request for information” is tagged exploratory each time.

What moved. The value comparison uses only procurements that gave an estimated value at both the engagement and tender stages. Titles were compared after removing words such as “market engagement” and “RFI”.

Start-ups and scale-ups. Winners were judged by the same test as the rest of Seam’s analysis: a private limited company in a trading industry, under ten years old on the day of the award, and not a large corporate, an investor, a listed company or an older business under a new registration. The full test. Winners were matched by name to the Companies House register of 1 November 2026. Of 9,254 winners named, 7,755 could be matched and judged, and 643 were start-ups or scale-ups.

What this cannot show. It measures what was tendered and who won, not who influenced the specification. It cannot see a purchase made under a different reference after the engagement. And 21 months is a short record: some exercises counted as having no tender will still produce one.

← All analysis