You have been handed a scope and a budget, and the budget cannot buy the scope. What do you put in front of the customer?

Size it before you negotiate it

You cannot have this conversation from an impression. The first move is a bottom-up sizing you would defend line by line, because everything afterwards — what to cut, whether to bid, what to say to the sales lead — depends on knowing the size of the gap rather than sensing it.

Take a customer with a stated budget of £400,000 for a claims-intake replacement, and a scope covering intake, document ingestion, an exceptions workbench, integration with the existing core system, test automation and deployment.

Bottom-up sizing, from the workstreams as scoped

  intake service and API                     120 days
  document ingestion                          95 days
  exceptions workbench                       110 days
  core-system integration                     85 days
  test automation and UAT support             90 days
  deployment, environments, IaC               45 days
  architecture and delivery management        65 days
                                            ---------
  effort                                     610 days

  blended day rate across the role mix        £820
  delivery cost         610 x 820      =    £500,200
  contingency, 12% weighted by risk    =     £60,000
  total cost                           =    £560,200
  price at a 25% gross margin  560,200 / 0.75 = £747,000

  stated budget                              £400,000
  gap                                        £347,000

The gap is not a discount. Meeting £400,000 at the same margin means a delivery cost of £300,000, which at £820 a day buys about 366 days — roughly sixty per cent of the effort the scope requires. That single figure is what makes the rest of the conversation possible, because you are no longer arguing about whether the price is high; you are telling the customer they have funded sixty per cent of what they described.

Cut scope, not quality

The 366 days you do have must buy something that works in production. So the reduction is made by removing whole outcomes, not by thinning every line, and there is a specific set of costs that cannot be reduced without the delivery becoming unsafe: security work, test automation for whatever you do build, and a deployable pipeline with environments. A phase one delivered by hand onto an unmanaged environment is not cheaper, it is a liability transferred to whoever operates it.

What 366 days buys, chosen as a coherent phase one

  intake service and API                     120 days
  document ingestion, single source only      60 days
  core-system integration, read plus create   70 days
  test automation and UAT support             55 days
  deployment, environments, IaC               45 days
  architecture and delivery management        40 days
                                            ---------
                                             390 days

  390 days is 24 over the 366 the budget funds. Recovered by
  dropping the create side of the core-system integration to
  phase two, leaving read-only at 50 days, and holding
  architecture and delivery management to 36.
  Revised: 366 days, cost £300,120, price £400,160,
  rounded to the £400,000 the customer has.

  Deferred to phase two, indicatively 244 days: the exceptions
  workbench, the second and third document sources, and the
  create side of the core-system integration. At the same rate
  with contingency restored, about £299,000.

Notice what £300,120 of cost has absorbed. In the full sizing that pot was delivery plus a 12 per cent contingency; here every pound of it is delivery, so phase one carries no contingency at all. That is not an oversight to be discovered in month four, it is the actual price of the budget and it goes in front of the customer in those words: at £400,000 there is no allowance for anything you have not already told us. Carry the 12 per cent instead and £400,000 funds £267,900 of delivery, roughly 326 days, which is forty days less again and takes another outcome out of phase one. Both positions are defensible; only the undisclosed one is not.

What you deliberately did not do is keep the exceptions workbench at half the effort. A workbench built in fifty-five days rather than a hundred and ten is not a smaller workbench, it is an unfinished one, and it will be the thing the customer points at in month nine.

Say what is lost in their metric

Feature lists do not communicate the consequence, so the deferral has to be expressed against the number the business case was approved on. If the case rests on same-day acknowledgement of claims and thirty per cent of claims require manual enrichment, then phase one delivers same-day acknowledgement for the seventy per cent that arrive complete, and the handling cost of the remaining thirty per cent is unchanged. That is the honest statement: phase one buys the headline metric and not the efficiency saving.

Putting it that way does something a scope table cannot. It tells the sponsor exactly which half of their business case survives, which is the sentence they need in order to go and ask for the rest of the money — and asking for phase two money against a working phase one is a far easier internal argument than asking for the full amount now.

When the gap is unbridgeable

Sometimes there is no coherent phase one. If a regulatory date requires the whole thing live, or the integration cannot be split, or the minimum viable delivery still costs twice the budget, then the honest finding is that the opportunity is not fundable as described, and saying so early is the valuable act rather than the defeatist one. That finding is worth more to your own organisation than a bid, because it is the difference between spending twenty presales days and spending them on something winnable.

There is one thing worse than declining, and it is the strategy of bidding the full scope at the reduced number with the intention of recovering the difference through change control. It reads as competitive, it wins, and then every month of delivery is a negotiation with a customer who believes they bought something they did not. The margin never recovers, the reference is lost, and the people who priced it are usually not the people who pay for it.

Why the arithmetic goes in front of the customer

The reflex is to hide the sizing and present only the option, on the grounds that showing your cost base weakens you in negotiation. In practice the opposite happens with a serious buyer: a price with no visible basis invites a percentage to be taken off it, whereas a price traceable to effort and a rate card can only be reduced by removing something, and now the customer is choosing what to remove. That reframing is the entire objective. You cannot be negotiated down from a number the buyer can see is made of work, but you can be negotiated down from a number that is just a number.

Being handed an impossible budget is not a pricing problem to be solved quietly; it is a scope decision the customer has not been given the information to make, and giving them that information is the job.