What is the basic procurement savings formula?

Subtract the new unit price from the agreed baseline price, then multiply the difference by the relevant quantity. Use planned quantity for the forecast and supported actual quantity for realized price savings. Deduct documented one-time implementation costs to reach a net figure.

What is a valid procurement savings baseline?

One that is supported by evidence and comparable in scope. A prior paid price taken from invoices is the strongest ordinary case. A current contract rate, an approved budget, a verified market benchmark or a should-cost model can all be valid where the reason is documented and Finance accepts the source.

What is the difference between forecast and realized savings?

Forecast savings apply the price difference to planned quantity before anything is bought. Realized price savings apply the same difference to the quantity actually purchased. Keep both. The gap between them explains how well the initiative was estimated.

What is the difference between hard savings and cost avoidance?

Hard savings compare a supported prior or current price with a lower new price, so spend on the same activity falls. Cost avoidance compares a credible future cost with the outcome that was negotiated instead. Avoidance rarely releases cash from a budget line, so it is normally reported separately.

Should lower purchasing volume count as savings?

No. Buying fewer units reduces spend without proving that procurement improved the unit price. Report the quantity change as a volume effect in its own field, beside the savings figures rather than inside them.

When should implementation costs be deducted?

Deduct supported one-time costs required to obtain the benefit, such as qualification, tooling, migration or transition, whenever the agreed methodology requires a net figure. Material recurring costs need their own explicit treatment rather than a silent deduction.

What evidence does Finance need?

Typically the baseline source, proof that the comparison is like for like, the new commercial terms, planned and actual quantities, effective dates, implementation costs, the treatment of FX and inflation, and a named reviewer with an approval status, date and comment.

Can AI calculate or approve procurement savings?

AI can help read documents, suggest how a line should be classified, or find likely matches between a tracker and a spend export. Deterministic rules should perform the calculation so the result is reproducible, and a person should approve the value that gets reported.

How often should procurement savings be reviewed?

Monthly review works for most teams, because it keeps actual quantity close to the period it belongs to. Initiatives with long implementation periods also need a review at the point the new terms take effect, so the forecast is not reported as realized before anything has been bought.

How do I track this in Excel?

Give each initiative one row and each value its own column, including baseline method, forecast, realized, implementation costs, net realized, cost avoidance, Finance-approved value and evidence status. The free ProcSave tracker template is set up this way, and the free calculator runs the same formulas for a single initiative.

When does a spreadsheet stop being enough?

Usually when actual spend has to be matched to initiatives across thousands of transactions, when evidence needs to be attached and found again months later, or when several people edit the same file and the change history matters. Those are workflow problems rather than formula problems.

ProcSave reference guide

Procurement savings methodology

Q: What is a procurement savings methodology?

It is the set of rules an organization uses to choose a baseline, calculate expected and realized savings, separate savings from other changes in spend, define the evidence required, and decide which values Finance will approve for reporting. There is no single universal treatment, so each organization agrees its own policy with Finance.

A practical method for deciding what counts as a saving, choosing a defensible baseline, separating price from volume, calculating forecast and realized value, and recording what Finance approved.

Use the free calculator Download the free tracker

Last updated 19 August 2026

What is a procurement savings methodology?

A procurement savings methodology is the set of rules an organization uses to choose a baseline, calculate expected and realized savings, separate savings from other changes in spend, define the evidence required, and decide which values Finance will approve for reporting.

There is no single universal accounting treatment for procurement savings. Public bodies publish criteria, and professional bodies publish guidance, but the treatment that applies to your reporting is the one your organization agrees with Finance and writes down.

The distinctions that decide every savings number

Most disagreements about a savings figure are really disagreements about which of these values somebody is quoting. Keeping them as separate fields, rather than one number that changes meaning, is the whole method.

What each procurement savings term means, how it is determined, and what it does not prove
Term	What it means	How it is calculated or determined	What it does not prove
Baseline	The agreed comparison point the new price or rate is measured against.	Chosen from a prior paid price, a current contract rate, an approved budget, a market benchmark or a should-cost model, and documented.	That the comparison is like for like, or that Finance accepts the source.
Forecast savings	The benefit the initiative expects before anything has been bought.	Unit price difference multiplied by planned or approved quantity.	That the new terms were used, or that anything was actually saved.
Realized price savings	The unit price benefit applied to what was actually purchased.	Unit price difference multiplied by supported actual quantity.	That implementation costs have been considered, or that Finance accepts the baseline.
Net realized savings	What remains once supported one-time implementation costs are deducted.	Realized price savings minus documented one-time costs.	That every recurring cost or accounting effect has been accounted for.
Finance-approved savings	The value Finance accepts for reporting.	A decision, recorded by a named reviewer with a date and a comment.	That it should overwrite the forecast or the calculated result. It sits alongside them.
Cost avoidance	The difference between a credible future cost and the outcome that was negotiated instead.	Supported future price or rate minus the agreed price, applied to relevant quantity.	That cash was released. It usually does not reduce a budget line, so it is reported separately.
Volume effect	The part of a spend change caused by buying more or fewer units.	Quantity difference multiplied by the new unit price.	Anything about procurement performance. It is reported beside savings, never inside them.
Implementation costs	One-time costs required to obtain the benefit, such as qualification, tooling, migration or transition.	Documented actual or committed costs attributable to the initiative.	That recurring costs have been handled. Those need their own treatment.
Savings realization rate	How much of the forecast was ultimately accepted for reporting.	Finance-approved savings divided by forecast savings.	That the forecast was wrong. A low rate can mean weak evidence, changed volume or a stricter policy.

The ProcSave method, step by step

This is the working method ProcSave recommends. It is not an accounting standard and it does not replace your own policy. Where your organization has agreed something different with Finance, follow that and keep the sequence.

Define the initiative and scope
Write down what is being bought, for which entities and sites, over which period, and who owns the result. An initiative that cannot be described in two sentences usually contains two initiatives, and they will need separate baselines.
Watch for. Scope that grows quietly during the year, so the final claim covers spend the original case never considered.
Choose a supported baseline
Pick the comparison point and record where it came from. A prior paid price taken from invoices is the strongest ordinary case, because it is what the organization demonstrably paid. Anything else needs a reason.
Watch for. A baseline chosen because it produces a better number rather than because it is the best supported comparison available.
Make the comparison genuinely comparable
Check specification, service level, quantity units, delivery terms, payment terms, geography and contract length. If any of them moved, either normalize the comparison or say plainly that the figure covers a changed scope.
Watch for. A unit that quietly changed, such as a price per litre compared with a price per drum, which can move a result by an order of magnitude.
Separate price, volume, scope, FX and inflation
Spend can fall for reasons that have nothing to do with procurement. Split the change into the part explained by unit price, the part explained by quantity, the part explained by a change in what was bought, and the parts explained by exchange rates or indexation. Report each separately.
Watch for. A total spend reduction presented as a savings figure when most of it is a volume or currency effect.
Calculate the forecast
Multiply the unit price difference by planned or approved quantity for the reporting period. This is the number the business case is judged on later, so record it once and leave it alone.
Watch for. Annualizing a full-year benefit and reporting it during a partial year in which it has not yet occurred.
Recalculate using actual activity
Once purchases have happened, apply the same unit price difference to the supported actual quantity. Keep the forecast next to it. The gap between them is information about the plan, not an error to be tidied away.
Watch for. Overwriting the forecast with the realized figure, which destroys the only evidence of how well the initiative was estimated.
Deduct supported implementation costs
Subtract documented one-time costs needed to obtain the benefit. Recurring costs are a separate question and need an explicit treatment agreed with Finance rather than a silent deduction.
Watch for. Qualification, tooling, migration and change costs that are real but sit in another budget, so nobody nets them off.
Record cost avoidance separately
Where the comparison is a credible future cost rather than a price already paid, record it in its own field with its own evidence. It can be a genuine and valuable result. It is usually not cash the budget releases.
Watch for. An avoidance figure added into a hard savings total, which is the fastest way for a whole report to lose credibility.
Attach the evidence
Link the baseline source, the new commercial terms, and the transaction data behind the actual quantity. A number a reviewer cannot trace to a document is an assertion, however carefully it was calculated.
Watch for. Evidence that lived in somebody’s inbox and left with them.
Record the Finance decision without overwriting the calculation
Store the approved value as its own field, with the reviewer, the date and the comment. If Finance adjusts or rejects a figure, keep the original calculation and the decision side by side so the trail explains itself.
Watch for. A rejected claim that quietly disappears from the tracker, leaving no record that it was ever made.

Principles this method holds to

Lower total spend does not automatically mean procurement created savings.
Lower purchasing volume should not be counted as a price saving.
An opening supplier quote is not automatically a valid historical baseline.
Forecast, calculated realization and Finance approval stay as three separate values.
A rejected or adjusted Finance decision stays visible in the audit trail.
Cost avoidance is never mixed into hard savings without disclosure.
AI can help read evidence or suggest a classification. Deterministic rules should do the arithmetic and people should approve what is reported.

Choose a baseline

Answer a few questions about the evidence you have. Nothing you select leaves your browser, and the helper never asks for a number, a supplier or a company name.

Which baseline should I use?

Question 1 of about 3

This helper needs JavaScript. It asks up to five questions about which evidence you have, then suggests a baseline type. The same guidance is written out under the baseline hierarchy, which lists each baseline in order of how easily it survives review.

Is there a comparable historical paid price or contract rate for the same thing?

Comparable means the same specification, the same unit of measure and broadly the same commercial terms.

Yes, invoices or paid transactions show what we paidYes, there is a current contract rate, but little or no purchase historyNo comparable internal history existsA first purchase, a new category, or a supplier we have not bought from before.

A more convenient comparison is not a more credible one

Roughly in order of how easily each comparison survives review. This is ProcSave’s working order, not an accounting standard. The right choice depends on your scope, your policy and the evidence you can actually produce, and a lower tier with strong documentation beats a higher tier you cannot support.

Comparable actual historical paid price
What the organization demonstrably paid for the same thing.
Needs. Invoices or paid transaction data covering a representative period.
Current supported contract rate
A committed rate, even where little or nothing has yet been bought against it.
Needs. The executed contract or rate card, with effective dates.
Approved budget or business-case baseline
A figure the organization already agreed to spend.
Needs. The approved budget line or business case, and who approved it.
Verified market benchmark or should-cost model
Useful for a first purchase where no internal history exists.
Needs. The source, its date, its scope, and the assumptions behind any model.
Supplier opening proposal
Weakest, because the supplier chose the number. Treat as cost avoidance at best, and only with support.
Needs. The written proposal, plus evidence the price was credible rather than an anchor.

Three worked examples

All figures are illustrative and chosen to be easy to follow. The arithmetic is the same arithmetic the free calculator runs.

Example A · illustrative

A straightforward price reduction

Previous supported unit price: $100
New unit price: $92
Planned quantity: 10,000
Actual quantity: 8,000
Supported implementation costs: $10,000

Forecast savings: ($100 - $92) x 10,000 = $80,000
Realized price savings: ($100 - $92) x 8,000 = $64,000
Net realized savings: $64,000 - $10,000 = $54,000
Volume effect: (10,000 - 8,000) x $92 = $184,000

The business bought 2,000 fewer units, so spend fell by a further $184,000. That is a volume effect and it belongs in its own field. Presenting it as additional price savings would claim credit for a decision procurement did not make.

Example B · illustrative

A documented increase that was negotiated down

Current price paid: $50
Supplier renewal notice, documented: $60
Negotiated renewal price: $54
Annual quantity: 20,000

Cost avoidance: ($60 - $54) x 20,000 = $120,000
Hard savings against price paid: ($50 - $54) x 20,000 = -$80,000

Both numbers are true and they point in opposite directions. The team avoided $120,000 of a proposed increase, and the organization will still pay $80,000 more than last year. Reporting only the first is how a savings report stops matching the budget. Keep avoidance in its own line unless Finance has approved a different treatment.

Example C · illustrative

A specification that changed underneath the comparison

Old service, 4-hour response, unit price: $1,000
New service, next-business-day response, unit price: $850
Annual quantity: 500
Priced value of the reduced service level: $120 per unit

Unadjusted comparison: ($1,000 - $850) x 500 = $75,000
Normalized for the scope change: ($1,000 - $850 - $120) x 500 = $15,000

Most of the apparent saving was the business accepting a slower service, not a better price for the same thing. Normalizing needs a supported value for what changed. Where no such value exists, report the figure with the scope change stated plainly rather than quietly comparing two different services.

The free procurement savings calculator runs this arithmetic on your own figures and shows every substitution.

What should not automatically be counted as procurement savings?

None of these are always wrong. Each one needs a stated reason and an agreed treatment before it belongs in a savings total.

01
Buying fewer units
Spend falls, but the unit price did not improve. Report it as a volume effect.
02
Delaying a purchase
The cost usually arrives in the next period. Deferral is a timing change, not a saving.
03
Removing required scope
Paying less for less is a business decision. Price it and disclose it.
04
Comparing against an unsupported list price
A price nobody was going to pay is not a baseline. Use what was paid or what was committed.
05
Annualizing before implementation
A full-year forecast reported as realized during a partial year overstates the current period.
06
Ignoring transition or implementation costs
Qualification, tooling and migration costs reduce the net result even when they sit in another budget.
07
Counting the same saving twice
A benefit that appears under a category initiative and again under a supplier programme is one benefit.
08
Treating a forecast as realized
Until purchases happen at the new terms, the forecast is an expectation.
09
Treating a calculated result as Finance-approved
The formula produces a result. Approval is a separate decision by a named person.
10
Mixing cost avoidance into hard savings
Combine them only where Finance has approved that treatment, and disclose it in the report.

Evidence checklist

What a reviewer typically needs in order to accept a figure without going back to the team. Print it and work through it before submitting a claim.

Baseline and comparability

Baseline source, named and attached
Comparable specification and scope confirmed
Previous invoice, contract, purchase order or rate card
New commercial terms

Quantities and timing

Planned quantity for the period
Actual quantity or transaction evidence
Effective and implementation dates
FX and inflation treatment recorded

Costs and assumptions

Implementation and recurring costs documented
Cost-avoidance assumptions written down
Calculation version recorded

Ownership and decision

Procurement owner named
Finance reviewer named
Approval status, date and comments
Change history preserved

Five values, kept side by side

Each stage is a separate stored value, not an edit to the one before it. The reason is simple: once a later number overwrites an earlier one, nobody can answer why the two differ, and that question is the one a reviewer always asks.

Procurement forecast
Procurement
The expected benefit at the point the initiative was approved.
System-calculated realized value
Deterministic rules
The same formula applied to supported actual activity, with no judgement applied.
Procurement-submitted value
Procurement
What the team puts forward, with any adjustment explained in writing.
Finance-approved value
Finance
What Finance accepts for reporting, recorded with a reviewer, a date and a comment.
Finance-adjusted or rejected value
Finance
Where the decision differs from the submission, both remain visible and the reason is stored with them.

A monthly reporting shape that answers the usual questions

An illustrative row layout. The point is the columns: each of them is a field somebody will ask about, and none of them can be reconstructed later if the tracker only stored one savings number.

An illustrative monthly procurement savings report with one row per initiative
Initiative	Owner	Baseline	Forecast	Realized price savings	Implementation costs	Net realized	Cost avoidance	Finance-approved	Variance	Evidence status	Approval status
Packaging resin	A. Okafor	Prior paid price	$80,000	$64,000	$10,000	$54,000	-	$54,000	-$26,000	Attached	Approved
Facilities cleaning	M. Lindqvist	Approved budget	$45,000	$45,000	$0	$45,000	-	$38,000	-$7,000	Partial	Adjusted
Software renewal	R. Baptiste	Supplier renewal notice	-	-	$0	-	$120,000	$120,000	-	Attached	Approved, avoidance
Logistics lanes	S. Haddad	Competitive bid benchmark	$120,000	$96,000	$18,000	$78,000	-	-	-	Not reviewed	Pending

Illustrative figures. The free Excel tracker template is laid out with these columns already.

To see these columns as a portfolio rather than a list, the interactive savings dashboard runs the same structure over a sample register.

Procurement savings questions

Sources and further reading

These are published, primary sources. They set out how specific public bodies define and validate savings. They are not a universal accounting standard, and none of them is a ProcSave document. Where this guide states a working method, that is ProcSave’s recommendation and it is labelled as such.

NHS England
Value and savings methodology
A published organizational methodology covering how value and savings are defined, categorized and reported across NHS commercial work.
Scope. Applies to NHS commercial activity.
UK National Audit Office
Progress with VFM savings and lessons for cost reduction programmes (HC 291, 2010 to 2011)
Sets out the criteria reported savings had to meet: properly calculated against a robust cost baseline, net of costs, new to the period, cash releasing, realised by the point of reporting, sustainable, and scored only once. The principles behind those criteria are why this guide keeps forecast, realized and approved values apart.
Scope. Criteria agreed with HM Treasury for UK central government reporting.
UK National Audit Office
A short guide to structured cost reduction (June 2010)
Argues that the cost of achieving a saving, including staff time and infrastructure, has to be taken into account to establish the true reduction.
Scope. Guidance for UK public bodies.
European Union
Directive 2014/24/EU on public procurement
Makes the most economically advantageous tender the overriding award concept, and allows the cost criterion to be a cost-effectiveness approach such as life-cycle costing rather than price alone. A useful counterweight to judging a category on its headline savings number.
Scope. Applies to public contracts across EU member states.

Procurement savings methodology

What is a procurement savings methodology?

The distinctions that decide every savings number

The ProcSave method, step by step

Define the initiative and scope

Choose a supported baseline

Make the comparison genuinely comparable

Separate price, volume, scope, FX and inflation

Calculate the forecast

Recalculate using actual activity

Deduct supported implementation costs

Record cost avoidance separately

Attach the evidence

Record the Finance decision without overwriting the calculation

Principles this method holds to

Choose a baseline

Which baseline should I use?

A more convenient comparison is not a more credible one

Comparable actual historical paid price

Current supported contract rate

Approved budget or business-case baseline

Verified market benchmark or should-cost model

Supplier opening proposal

Three worked examples

A straightforward price reduction

A documented increase that was negotiated down

A specification that changed underneath the comparison

What should not automatically be counted as procurement savings?

Buying fewer units

Delaying a purchase

Removing required scope

Comparing against an unsupported list price

Annualizing before implementation

Ignoring transition or implementation costs

Counting the same saving twice

Treating a forecast as realized

Treating a calculated result as Finance-approved

Mixing cost avoidance into hard savings

Evidence checklist

Baseline and comparability

Quantities and timing

Costs and assumptions

Ownership and decision

Five values, kept side by side

Procurement forecast

System-calculated realized value

Procurement-submitted value

Finance-approved value

Finance-adjusted or rejected value

A monthly reporting shape that answers the usual questions

Procurement savings questions

What is a procurement savings methodology?

What is the basic procurement savings formula?

What is a valid procurement savings baseline?

What is the difference between forecast and realized savings?

What is the difference between hard savings and cost avoidance?

Should lower purchasing volume count as savings?

When should implementation costs be deducted?

What evidence does Finance need?

Can AI calculate or approve procurement savings?

How often should procurement savings be reviewed?

How do I track this in Excel?

When does a spreadsheet stop being enough?

Sources and further reading

Turn the methodology into a repeatable process