On this page
- Research question and scope
- Why an unstated assumption becomes a dispute
- Defining an assumption statement
- Building the quote sample
- Coding assumptions by category
- Separating assumptions from exclusions
- Comparing quotes with change orders
- Interpreting the assumption rate
- Limitations and threats to validity
- Reporting and next action
- Sources
Research question and scope
A website quote usually arrives as a total, a list of deliverables, and a timeline. The total depends on assumptions the client may never see, such as how many templates exist, who supplies copy, and whether the current CMS can be reused. This study asks a narrow question: do written quotes state the assumptions behind each price, and when they do not, which assumptions later appear as change orders? The scope is the quote history from one outsourced website relationship over two years.
The unit of analysis is one price line tied to one deliverable, phase, or service block. A line may be a fixed fee, a rate card entry, or an allowance. The method does not judge whether the price was fair. It judges whether the record shows the conditions that would make the price move, because those conditions are what the client needs to compare bids and plan a budget.
Why an unstated assumption becomes a dispute
An assumption is a fact the estimator treats as true to produce the number. The client assumes the quote covers responsive layouts, while the estimator priced desktop pages and a separate mobile pass. The team assumes the client will supply final copy, while the client expects the agency to write it. Both positions can be reasonable, and both become expensive when the gap appears after work starts.
Change orders are where the gap surfaces. A client receives a request for additional hours and reads it as a pricing failure, while the team reads it as work outside the agreed basis. The argument is rarely about the extra hours themselves. It is about whether the original number ever included the condition now being priced. A quote that names its assumptions turns that argument into a comparison against a written record.
Defining an assumption statement
An assumption statement has three parts: the condition, the source, and the effect on price or schedule. The condition is a fact about the site, the client, or the environment. The source explains how the estimator knew or why the fact is being assumed. The effect states what happens if the condition proves false. A line that says only "copy supplied by client" is a note, not a full assumption, because it does not say what the team will do if copy arrives late.
Code a statement as complete only when all three parts are present in the quote or in an attached schedule. Code it as partial when the condition is named but the effect is missing. Code it as absent when the client must infer the condition from the deliverables. This three-level scale is easier to apply consistently than a simple yes-or-no count, and it shows where a small edit to the template would improve the record.
Building the quote sample
Start with every quote version issued during the observation period, not only the version that was signed. Keep the original documents unchanged and work from copies with a stable quote identifier. Record the issue date, the version number, the total, the currency, the proposed start date, and the named contact on each side. A quote that was revised three times contributes three records, because each revision may add or drop assumptions.
Exclude internal estimates that were never sent to the client, since the research question concerns the client-facing record. Exclude generic marketing collateral that lists capabilities without a price. Include the proposal, the pricing schedule, and any annex that the client received in the same package. If the package references a separate assumptions document, include that document and link it to the quote version, because the assumption may live outside the main price table.
Coding assumptions by category
Use a small set of categories that match how website work actually changes: content, templates, integrations, access, approvals, technology, and timeline. Content covers who writes, edits, and approves copy and images. Templates cover the number and complexity of page layouts. Integrations cover payment, CRM, booking, and analytics connections. Access covers accounts, hosting, and third-party systems the team needs. Write one sentence of definition for each category before coding starts.
Approvals cover review rounds and named decision-makers. Technology covers the CMS, framework, and hosting stack. Timeline covers dependencies such as legal review, translation, or a client holiday. Add an "other" category only after two reviewers agree that a statement fits nowhere else, and record the wording that forced the addition. Keep the category list stable across the sample so the counts remain comparable between quarters.
Separating assumptions from exclusions
An exclusion names work the price does not cover, while an assumption names a condition the price depends on. The difference matters because an exclusion is a boundary and an assumption is a risk. A quote that excludes "content migration" tells the client what is not included. A quote that assumes "the client will migrate content using the existing export" tells the client what must be true for the price to hold.
Code both, but keep the counts separate. A quote can have a clear exclusion list and still fail the assumption test, because the client never learns which conditions would trigger a new price. A quote can also state assumptions without listing exclusions, which leaves the client unsure where the work stops. The strongest records pair each assumption with the exclusion or change-order path that applies when it fails.
Comparing quotes with change orders
Match each change order to the quote version and assumption that predicted it. Record the amount, the reason, the date requested, and the date approved. A change order that traces to a named assumption is evidence that the assumption was material. A change order with no matching assumption is a gap in the quote, even when the extra work was legitimate and fairly priced.
Calculate two counts: the share of assumptions that later produced a change order, and the share of change orders that trace to a written assumption. The first shows which assumptions deserve more attention during scoping. The second shows how much of the final cost was foreseeable from the quote. Report both with the number of quotes and change orders beside them, because a percentage from six documents can mislead.
Interpreting the assumption rate
A high rate of complete assumption statements is not automatically good. A quote can list twenty assumptions and still bury the one that matters, such as an uncapped number of design rounds. Read the statements for materiality, not only for presence. The useful question is whether the assumptions that changed the price were visible before the client signed. A short quote with three material assumptions can be stronger than a long schedule of minor ones.
A low rate is not automatically bad either. A small, well-understood template build may rest on conditions both sides already know. The risk rises when the site is old, the client has several stakeholders, or the work crosses teams. Compare the assumption rate with the change-order rate and with the number of quote revisions. A pattern of few assumptions and many change orders points to a quoting habit that needs repair.
Limitations and threats to validity
The sample comes from one relationship, so the result describes that relationship rather than the market. A different client, sector, or contract size may produce a different mix. The coding also depends on the quality of the quote documents, and a team may discuss assumptions in meetings without writing them down. Treat the count as a property of the written record, not as proof that the conversation was poor.
Estimators may also write assumptions defensively after a dispute, which inflates the rate in later quotes. Note the date of any template change and compare the periods before and after it. Keep the coding sheet and the definitions with the report so another reviewer can reproduce the result. External references can frame why written change control matters, but they do not supply a local assumption rate.
Reporting and next action
Summarize the result in one page with the complete, partial, and absent counts, the top three assumption categories, and the change orders that trace to a written assumption. Name one assumption that should be added to the quote template and one exclusion that needs clearer wording. Assign an owner and a date for the template edit. Send the summary to the client alongside the next quote so the change is visible.
Repeat the audit after two quoting cycles and compare the same categories. A rising complete rate is useful only when the quote mix and the observation window stayed comparable. Keep the old and new templates side by side so the effect of the edit can be read. The goal is a quote that lets the client see which conditions carry the price and which ones will require a new decision.
Sources
The sources below frame written change control, contract modification, accessibility risk, and performance work. They do not provide a local assumption rate, and each was checked on November 11, 2026. Use them as context for why a price should name its conditions, and keep the local coding sheet separate from these references. Read them as background for the audit design rather than as evidence about this relationship.
- Federal Acquisition Regulation Part 43, Contract Modifications
- NIST SP 800-40 Rev. 4, Guide to Enterprise Patch Management Planning
- Web Content Accessibility Guidelines 2.2
- W3C: Evaluating Web Accessibility Overview
- web.dev: Core Web Vitals
- HTTP Archive: Page Weight
- CISA Known Exploited Vulnerabilities Catalog
Turn the guidance into a reviewable handoff
Record the owner, input, acceptance condition, evidence, and escalation point before production begins. That small contract makes the work easier to inspect and approve.
Plan a website production handoff