A price needs its scope

A fee becomes useful when you can see what it buys. Comparing two numbers before comparing their scope can turn a cheaper headline into the wrong conclusion.

This note explains a practical comparison method. It draws on three primary methodological sources and uses a fictional example. It reports no observed market prices, supplier ranking or client findings.

What the established methods support

The U.S. Government Accountability Office describes normalization as preparing cost data for meaningful comparison, with attention to units and work content. That supports an essential discipline: understand what an amount represents before calculating with it. GAO's context is program cost estimation, which differs from comparing suppliers' selling prices. GAO Cost Estimating and Assessment Guide, chapter 9, pages 93-94.

Federal procurement guidance supplies a related price-analysis principle. A historical price may be unsuitable when circumstances differ materially; terms, quantities and economic conditions matter. We use this as methodological precedent, not a claim that federal purchasing rules govern every commercial comparison. FAR 15.404-1(b)(2)(ii).

For industries outside wholesale and retail trade, BLS defines the prices used in its Producer Price Index with reference to the product or service, buyer, transaction terms and timing. It generally seeks actual transaction prices rather than list prices. Its statistical purpose differs from an individual buying decision, but the distinction reinforces why a published starting price and a realized transaction should retain different labels. BLS PPI Handbook of Methods, Prices.

These sources support the underlying comparison discipline. They do not validate a proprietary Permeate scoring system or establish any market's typical fee. The workflow below is our practical synthesis.

Write the question before collecting the numbers

Are we comparing offers for a defined assignment, describing advertised prices or preparing a budget scenario? Keep those questions separate. For an offer comparison, write the required work, locations, deliverables, timing and acceptance conditions first.

Use two tests. Scope fit: would this offer satisfy the assignment? Price basis: do we know what the buyer would pay under the stated conditions? A complete number can fail the first test. An appropriate service can fail the second.

Preserve the record behind each amount

We recommend retaining these fields with every observation:

An empty cell stays unknown. In this workflow, it never silently becomes zero or “included.” Reposted copies of the same offer retain one observation identity.

A fictional comparison

Illustration only. Every offer and amount below is invented. These are not market benchmarks or recommended fees. All amounts are USD.

The assignment requires two site reviews, one consolidated report and one follow-up meeting. For this illustration, deliverable standards and timing are identical, the additions complete the required scope, and there are no other charges.

Offer A has the lower base fee. Its defined-scope total is $4,800 + $1,600 + $400 = $6,800, which is $800 above Offer B. This establishes only the arithmetic within this invented scenario. It does not show which real supplier offers better value.

Remove the second-site price from A and the comparable total becomes unknown: $5,200 plus an unpriced second site. The missing number is the next research question.

There is still useful conditional information. With every other example assumption unchanged, A would cost less if its second-site charge were below $800, tie at $800 and cost more above $800. This threshold is a calculation, not evidence that any supplier would offer that price.

Normalize only what the evidence permits

Keep the quoted number beside the calculated total. Identify whether each adjustment is a sourced charge or an analyst's scenario. Do not describe a scenario as a quotation.

If one offer promises a materially different standard of work, matching site counts does not make the services equivalent. Record the mismatch. A technical review or supplier clarification may be needed before any price ranking is useful.

Likewise, a per-site amount can hide a minimum charge or unusual site complexity. We would use it to investigate a difference, not to erase that difference. Unknown scope cannot be repaired by a tidier denominator.

Stop at the conclusion the record supports

Our proposed reporting states are:

• Comparable for this assignment: required scope and material charges are sufficiently established for the stated comparison.
• Conditional comparison: the conclusion depends on visible, decision-relevant assumptions.
• Unresolved comparison: missing scope or price information prevents the claimed conclusion.

These are working labels, not externally certified categories. If two reviewers disagree about a material inclusion, preserve the disagreement and its consequence until the underlying evidence resolves it.

A collection of public prices should not become “the market rate” merely because it fits in a table. For any future market-price study, we would state the target market, collection window, selection process, distinct observations, exclusions and limitations. This note has no such dataset.

Keep the qualification beside the figure

Our publication practice is to keep labels such as “fictional,” “starting price” and “calculated scenario” close to amounts in tables and summaries. That gives humans and machines context to inspect. Whether an AI system preserves those qualifications is a separate empirical question; this note does not test it.

How this note was researched

On September 15, 2026, we conducted a targeted review of primary institutional guidance on cost normalization, price comparison and transaction-price definitions. We inspected the cited GAO, Acquisition.gov and BLS material. Selection was purposive; this was not a systematic review or an exhaustive search of research literature. The institutions address different tasks, and their agreement on a principle is not a controlled test of this workflow.

The contribution here is a transparent synthesis and a fully specified fictional example. The next evidentiary step would be to apply a declared protocol to observed records and report both the comparisons that survive and those that remain unresolved.