A procurement newsletter landed this morning with a good piece in it. The Purchasing & Procurement Center published Finance Didn’t Cut Your Savings. Five Leaks Did. — a careful walk through where reported procurement savings disappear between the handshake and the accounts.
The framing is honest and the arithmetic is real. Report four million in savings. On a good year, eighty-five to ninety percent survives to the number finance will defend. On a bad year, the same four million falls to around fifteen percent. Six hundred thousand dollars. Same team, same work, a number four times smaller.
Five leaks explain the spread. Cost avoidance, which finance cannot book. Price and contract leakage, where the negotiated rate is not the rate paid. Timing, where the saving is real but lands in a later period. Budget capture, where the money is never lifted out of the line. And baseline integrity — the reference price every saving is measured against.
It is a useful piece and I would recommend it. I also think one of those five is in the wrong list, and that noticing which one is the whole point.
One of these is not like the others
Four of the five are things that happen to a saving after it exists.
Avoidance exists and cannot be booked. The rate exists and is not applied. The saving exists and arrives in a later period. The saving exists and the budget holder spends it. In every case there is a real number and something happens to it on the way to the accounts.
Baseline integrity is not in that class. It is not something that happens to a saving. It is the question of whether the saving was ever a number at all.
The author sees this and says so without quite naming it. Baseline integrity is described as the leak that decides all the others. A weak baseline, it says, does not just shrink your savings — it removes your ability to track or defend any of them. And then, revealingly, the third item on the list is introduced as timing, the one that isn’t really a leak.
Two of the five are flagged inside the article as not fitting the frame. That is what it looks like when a structure cannot hold the thing it has found.
What a baseline actually is
The baseline is the point the measurement starts from.
Every saving is a difference between two numbers. If the first number is estimated, reconstructed after the fact, or set by the party reporting the result, then the difference is not a measurement. It is an assertion with arithmetic attached.
That is not leakage. It is the instrument being unsound before anything is measured. Everything downstream inherits the defect — including, as the article correctly notes, cost avoidance, which requires proving what the price was going to be.
So the honest count is four leaks and one measurement problem. And the measurement problem determines whether the other four are worth discussing.
The part that answers itself
Now read the prescriptions for the two largest items.
For budget capture: partner with finance on capture, agree what comes out, track it together.
For baseline integrity: agree the baseline with finance before the saving is booked, so it cannot be reopened.
Both remedies require a working relationship with finance as a precondition. Both are correct. And both assume the availability of the one thing the twenty-year record says is missing.
Twenty years of the same finding
In May 2006, Purchasing Magazine published How to Speak Like a CFO. Its opening line: finance executives in corporate America simply do not believe that purchasing departments are bringing in the savings they claim.
In January 2008 I wrote about that article here, alongside a December 2007 piece reporting that procurement executives still could not get onto the CFO’s agenda. That same year I published a white paper citing an Aberdeen study: of 11.9 percent in average identified savings presented by purchasing, 3.2 percent was actually booked by finance. A difference of 73 percent between identification and realization.
In 2019, a CPO Viewpoint study of more than two hundred companies found 66 percent of finance professionals describing procurement as a hindrance to business objectives, with only a third believing procurement helps with cost saving.
In September 2025, a Fine Tune and CFO Leadership survey of over a hundred CFOs found 29 percent of those at large companies had high confidence in procurement’s ability to deliver measurable P&L-aligned savings. Thirty-eight percent overall believed procurement savings actually reach the bottom line.
This morning: eighty-five percent on a good year, fifteen percent on a bad one.
Twenty years. Different surveys, different decades, different vocabulary. The number has not moved.
What Virginia showed in 2014
In 2014 the Commonwealth of Virginia asked me to present at the Public Procurement Forum on bridging the disconnect between finance and purchasing. Virginia’s eVA program was already a success — it had been since the early 2000s, and I had been covering it since 2007. This was leadership identifying an area for improvement inside something that was working.
Before the session started I asked everyone to change tables. Finance people had been sitting with finance people, procurement with procurement. I mixed them.
Many of them were meeting face to face for the first time. Same organization. Some for years.
The dialogue that started in that room continued well past the session. The Commonwealth flew a flag over the State Capitol in Richmond in recognition of the contribution, and Director Robert E. Gleason wrote that the presentation had been helpful to an extraordinary number of the Commonwealth’s procurement staff. It is the honor I value most, and it was not for building anything.
The lesson underneath it is the one that matters here. You cannot agree a baseline with a department you have never spoken to. The literacy gap sits downstream of the relationship gap, and the relationship gap sits downstream of nobody ever having established a shared starting point.
How you find the question everyone is missing
This is a small example, and small examples are the useful ones because the method is visible in them.
Start from the failure, not the plan. The failure here is that reported savings do not survive to the P&L, at a rate that has not improved in twenty years.
Take the best available explanation and check whether it fits. Five leaks is a good explanation. It accounts for most of the spread.
Then look for the item the frame has to keep apologizing for. Two of the five arrive with a note explaining why they belong on the list. That is where the frame is under strain.
Ask what class each item actually belongs to. Four are events. One is a measurement condition. They are not the same kind of thing and cannot be fixed by the same kind of action.
Then follow the remedy back to what it assumes. The two largest items are both fixed by agreeing something with finance. Ask whether that is available. The record says it usually is not.
That sequence takes about twenty minutes and it does not require a framework. It requires being willing to leave the list you were handed.
The same condition, in three places
In 1998 I was brought into a defence maintenance operation running at 51 percent next-day delivery. Every measure on the board was accurate. The question that mattered was what time of day the technicians submitted their orders — because a next-day requirement is a duration, and a duration has a start point nobody had established. The answer was four o’clock in the afternoon.
In 1990 a very large software company recognized license revenue at contract signing rather than at delivery. Nobody had settled which point a sale is measured from. The restatement followed.
This morning: a saving is a difference from a prior price, and nobody has established what the prior price was.
Three domains. Defence logistics, revenue recognition, savings reporting. One condition running through all of them — the starting point of the measurement is never established, and everything built on it inherits the defect.
That is not a procurement problem. It is where procurement happens to be standing when it shows up.
What to do about it
The article’s advice on baselines is right and worth repeating: capture the prior price in writing at the moment you negotiate, use an independent external reference where no true prior price exists, and agree it with finance before the saving is booked.
I would add one step before all three.
Find out whether your finance counterpart has ever sat at a table with you without an agenda. If the answer is no, the baseline conversation has not failed yet — it has not started. And no amount of tooling, dashboard integration or AI orchestration will start it for you.
Virginia worked out how to do that in 2014. It took rearranging the seating.
Truth Is Believing. Accuracy Is Knowing. Outcome Is Proof.™
-30-
CASE STUDY — When Competent Accounting Builds the Blindness
From “The Map You Design and the Map You Trace,” Procurement Insights, 29 July 2026.
The 1998 trace did not stop at four o’clock. It kept going, and it left the operation’s four walls entirely — into finance, the last place anyone mapping a procurement process would ever look.
The purchase orders captured the cost of the part only. The suppliers’ courier charges arrived separately, on the invoices. Finance could not reconcile the two — the true landed cost of a part was never in one place — and worse, the markup the contract allowed was calculated on the part cost alone. The shipping the PO never captured was quietly eating the margin. On transactions the program was booking as revenue, it was losing money.
The detail that makes it unforgettable is why no one had caught it: finance had done nothing wrong. They had separated part cost and shipping into two line items, each managed and each reconciling cleanly on its own. The loss existed only in the relationship between the two lines — and no one was looking at the two lines together. Competent accounting had built the blindness.
Read that against the five leaks.
There is no leak in it. Nothing escaped between the handshake and the accounts. Both line items were accurate, both were reconciled, both were managed by people doing their jobs correctly. A leak audit would have found nothing, because there was nothing leaking.
The loss lived in a relationship between two measurements that nobody had established as a single measurement. The landed cost of a part was never a number anywhere in the system — the same way, in the newsletter’s fifth item, the prior price is never a number anywhere in the system.
⚠ And note where it was found. Not in procurement. In finance — reached only because the chain kept going after the four o’clock answer, and crossed a boundary nobody had authorized it to cross.
That is the argument of this post in one case. The baseline is not a leak to be plugged. It is a measurement that was never constituted, and the place it goes missing is usually the gap between two functions that are each performing correctly.
Which is also why the remedy is not a better audit. Two functions each reconciling cleanly will never find it, no matter how carefully either one looks. Somebody has to be looking at both at once — and in most organizations, nobody has been asked to.
Hansen Models™ · Procurement Insights
Related
The Fifth Leak Is Not a Leak
Posted on September 10, 2026
0
A procurement newsletter landed this morning with a good piece in it. The Purchasing & Procurement Center published Finance Didn’t Cut Your Savings. Five Leaks Did. — a careful walk through where reported procurement savings disappear between the handshake and the accounts.
The framing is honest and the arithmetic is real. Report four million in savings. On a good year, eighty-five to ninety percent survives to the number finance will defend. On a bad year, the same four million falls to around fifteen percent. Six hundred thousand dollars. Same team, same work, a number four times smaller.
Five leaks explain the spread. Cost avoidance, which finance cannot book. Price and contract leakage, where the negotiated rate is not the rate paid. Timing, where the saving is real but lands in a later period. Budget capture, where the money is never lifted out of the line. And baseline integrity — the reference price every saving is measured against.
It is a useful piece and I would recommend it. I also think one of those five is in the wrong list, and that noticing which one is the whole point.
One of these is not like the others
Four of the five are things that happen to a saving after it exists.
Avoidance exists and cannot be booked. The rate exists and is not applied. The saving exists and arrives in a later period. The saving exists and the budget holder spends it. In every case there is a real number and something happens to it on the way to the accounts.
Baseline integrity is not in that class. It is not something that happens to a saving. It is the question of whether the saving was ever a number at all.
The author sees this and says so without quite naming it. Baseline integrity is described as the leak that decides all the others. A weak baseline, it says, does not just shrink your savings — it removes your ability to track or defend any of them. And then, revealingly, the third item on the list is introduced as timing, the one that isn’t really a leak.
Two of the five are flagged inside the article as not fitting the frame. That is what it looks like when a structure cannot hold the thing it has found.
What a baseline actually is
The baseline is the point the measurement starts from.
Every saving is a difference between two numbers. If the first number is estimated, reconstructed after the fact, or set by the party reporting the result, then the difference is not a measurement. It is an assertion with arithmetic attached.
That is not leakage. It is the instrument being unsound before anything is measured. Everything downstream inherits the defect — including, as the article correctly notes, cost avoidance, which requires proving what the price was going to be.
So the honest count is four leaks and one measurement problem. And the measurement problem determines whether the other four are worth discussing.
The part that answers itself
Now read the prescriptions for the two largest items.
For budget capture: partner with finance on capture, agree what comes out, track it together.
For baseline integrity: agree the baseline with finance before the saving is booked, so it cannot be reopened.
Both remedies require a working relationship with finance as a precondition. Both are correct. And both assume the availability of the one thing the twenty-year record says is missing.
Twenty years of the same finding
In May 2006, Purchasing Magazine published How to Speak Like a CFO. Its opening line: finance executives in corporate America simply do not believe that purchasing departments are bringing in the savings they claim.
In January 2008 I wrote about that article here, alongside a December 2007 piece reporting that procurement executives still could not get onto the CFO’s agenda. That same year I published a white paper citing an Aberdeen study: of 11.9 percent in average identified savings presented by purchasing, 3.2 percent was actually booked by finance. A difference of 73 percent between identification and realization.
In 2019, a CPO Viewpoint study of more than two hundred companies found 66 percent of finance professionals describing procurement as a hindrance to business objectives, with only a third believing procurement helps with cost saving.
In September 2025, a Fine Tune and CFO Leadership survey of over a hundred CFOs found 29 percent of those at large companies had high confidence in procurement’s ability to deliver measurable P&L-aligned savings. Thirty-eight percent overall believed procurement savings actually reach the bottom line.
This morning: eighty-five percent on a good year, fifteen percent on a bad one.
Twenty years. Different surveys, different decades, different vocabulary. The number has not moved.
What Virginia showed in 2014
In 2014 the Commonwealth of Virginia asked me to present at the Public Procurement Forum on bridging the disconnect between finance and purchasing. Virginia’s eVA program was already a success — it had been since the early 2000s, and I had been covering it since 2007. This was leadership identifying an area for improvement inside something that was working.
Before the session started I asked everyone to change tables. Finance people had been sitting with finance people, procurement with procurement. I mixed them.
Many of them were meeting face to face for the first time. Same organization. Some for years.
The dialogue that started in that room continued well past the session. The Commonwealth flew a flag over the State Capitol in Richmond in recognition of the contribution, and Director Robert E. Gleason wrote that the presentation had been helpful to an extraordinary number of the Commonwealth’s procurement staff. It is the honor I value most, and it was not for building anything.
The lesson underneath it is the one that matters here. You cannot agree a baseline with a department you have never spoken to. The literacy gap sits downstream of the relationship gap, and the relationship gap sits downstream of nobody ever having established a shared starting point.
How you find the question everyone is missing
This is a small example, and small examples are the useful ones because the method is visible in them.
Start from the failure, not the plan. The failure here is that reported savings do not survive to the P&L, at a rate that has not improved in twenty years.
Take the best available explanation and check whether it fits. Five leaks is a good explanation. It accounts for most of the spread.
Then look for the item the frame has to keep apologizing for. Two of the five arrive with a note explaining why they belong on the list. That is where the frame is under strain.
Ask what class each item actually belongs to. Four are events. One is a measurement condition. They are not the same kind of thing and cannot be fixed by the same kind of action.
Then follow the remedy back to what it assumes. The two largest items are both fixed by agreeing something with finance. Ask whether that is available. The record says it usually is not.
That sequence takes about twenty minutes and it does not require a framework. It requires being willing to leave the list you were handed.
The same condition, in three places
In 1998 I was brought into a defence maintenance operation running at 51 percent next-day delivery. Every measure on the board was accurate. The question that mattered was what time of day the technicians submitted their orders — because a next-day requirement is a duration, and a duration has a start point nobody had established. The answer was four o’clock in the afternoon.
In 1990 a very large software company recognized license revenue at contract signing rather than at delivery. Nobody had settled which point a sale is measured from. The restatement followed.
This morning: a saving is a difference from a prior price, and nobody has established what the prior price was.
Three domains. Defence logistics, revenue recognition, savings reporting. One condition running through all of them — the starting point of the measurement is never established, and everything built on it inherits the defect.
That is not a procurement problem. It is where procurement happens to be standing when it shows up.
What to do about it
The article’s advice on baselines is right and worth repeating: capture the prior price in writing at the moment you negotiate, use an independent external reference where no true prior price exists, and agree it with finance before the saving is booked.
I would add one step before all three.
Find out whether your finance counterpart has ever sat at a table with you without an agenda. If the answer is no, the baseline conversation has not failed yet — it has not started. And no amount of tooling, dashboard integration or AI orchestration will start it for you.
Virginia worked out how to do that in 2014. It took rearranging the seating.
Truth Is Believing. Accuracy Is Knowing. Outcome Is Proof.™
-30-
CASE STUDY — When Competent Accounting Builds the Blindness
From “The Map You Design and the Map You Trace,” Procurement Insights, 29 July 2026.
The 1998 trace did not stop at four o’clock. It kept going, and it left the operation’s four walls entirely — into finance, the last place anyone mapping a procurement process would ever look.
Read that against the five leaks.
There is no leak in it. Nothing escaped between the handshake and the accounts. Both line items were accurate, both were reconciled, both were managed by people doing their jobs correctly. A leak audit would have found nothing, because there was nothing leaking.
The loss lived in a relationship between two measurements that nobody had established as a single measurement. The landed cost of a part was never a number anywhere in the system — the same way, in the newsletter’s fifth item, the prior price is never a number anywhere in the system.
⚠ And note where it was found. Not in procurement. In finance — reached only because the chain kept going after the four o’clock answer, and crossed a boundary nobody had authorized it to cross.
That is the argument of this post in one case. The baseline is not a leak to be plugged. It is a measurement that was never constituted, and the place it goes missing is usually the gap between two functions that are each performing correctly.
Which is also why the remedy is not a better audit. Two functions each reconciling cleanly will never find it, no matter how carefully either one looks. Somebody has to be looking at both at once — and in most organizations, nobody has been asked to.
Hansen Models™ · Procurement Insights
Share this:
Like this:
Related