Most Companies Set Goals for Purchases and Never Check Them

← Back · ACT II · Chapter 9: The Value Control Loop

Value management has four steps. The first is defining value at purchase, in a business case, with numbers attached. The second is demanding evidence during selection: checking references and testing the vendor’s claims. Both happen inside the deal, while people are paying attention to it.

The third step is going back to that definition on a planned date and comparing it with what actually happened, and the fourth is correcting what the comparison shows. Neither has a natural moment once the deal closes. Unless a contract or a review schedule gives them a date, they wait for something to break loudly. When nothing breaks, the outcome can still move and nobody reopens anything.

A dashboard does not close the loop either, because a report that no one has to act on changes nothing. The loop closes when the comparison is written into the process with a date, and when a correction follows from what it shows: renegotiating terms, fixing a gap, retiring a tool, or revising the definition because the business moved.

There is a simple way to test whether the loop is real: name the last correction it produced. If no one can, the loop is open, and everything downstream of it is running on trust. A claim that someone owns the result is worth what the record behind it shows, which I wrote about in the case of executive sign-off on AI purchases.

A vendor that delivers gains from this too. A dated comparison is how its results get into the buyer’s own record before the renewal conversation starts.

One place the loop is left open is deployment, where production is mistaken for capability.

The United Kingdom now has a statutory instance of a loop left open, where the rating is published for the next buyer and the register that would carry it there is still to be built: UK Buyers Now Rate Suppliers Publicly, With No Way to Compare.

See also Procurement Prices Every Concession and Never Checks the Result, on the concession-pricing apparatus and the verification line it leaves out.

United States federal buying has a written review that comes back at least once a year and still does not ask what the purchase produced: Federal Contract Reviews Will Ask Six Questions, None on Results.

A federal modernization fund shows the same loop left open at the funding gate, where a savings estimate was required for approval and nothing in the repayment terms depended on it afterward: A Federal IT Fund Required Savings Estimates but Not Savings.

An Australian agency held a post-implementation workshop after a register rebuild that went live four years late, and the four priority actions it identified were all about how the next project should be run: A Post-Implementation Review That Never Asked If the System Worked.

Keep reading: ACT II · Chapter 9: The Value Control Loop