Compare scope before comparing fees
AI Readiness Assessment Pricing
Compare AI readiness quotations through evidence scope, effort, Arabic evaluation, internal time, exclusions and change assumptions using clearly illustrative SAR examples.
AI readiness assessment pricing should be tied to the decision being supported, the evidence that must be examined and the work required to produce a defensible conclusion. A low fee can be poor value if it excludes the source review or Arabic evaluation the customer needs. A larger fee is not automatically justified by a longer report. Compare the scope, assumptions, deliverables and customer effort before comparing totals.
This guide is for Saudi buyers preparing a readiness brief or evaluating quotations. It uses an illustrative multi-brand furniture business assessing an internal furniture-care assistant. All rates, fees and effort figures below are invented planning examples, not OpenForAI prices, market averages or quotations. A real price requires an agreed scope and current commercial terms. The examples show how to make that discussion specific.
Price the decision and evidence boundary
The furniture business wants to know whether a controlled trial can prepare supported internal care guidance from approved brand manuals. The assistant would not invent treatment instructions, make warranty decisions or send advice directly to customers. The assessment must examine source authority, applicable product versions, language requirements and the people who will review proposed guidance. Those requirements define the work more clearly than the phrase “AI assessment”.
State how many workflows, source families, business units and languages are included. Identify whether the assessor will inspect actual material, interview owners, review access arrangements or only facilitate a workshop. These methods produce different evidence. A quote should make the difference visible so the customer does not buy a discussion while expecting a substantiated operational conclusion.
Use assessment preparation to make the request concrete before seeking prices. A short evidence index and known limitations can reduce uncertainty in the estimate. Preparation does not guarantee a lower fee, but it helps bidders explain what they are pricing and which assumptions could change the effort.
Understand the work that creates assessment cost
The number of documents matters less than the difficulty of interpreting them for the proposed decision. Twelve consistent manuals with clear owners may be simpler than three contradictory sources spanning product variants and uncertain reuse rights. A count of files alone can therefore conceal effort. Ask the assessor which differences drive investigation, review and explanation.
| Factor | Lower-complexity condition | Condition requiring more investigation |
|---|---|---|
| Workflow boundary | One internal drafting task | Several tasks with different owners and consequences |
| Source meaning | Clear versions and product applicability | Conflicting manuals, unclear identity or missing definitions |
| Language | One agreed source and output language | Arabic and English acceptance with material source conflicts |
| Access | Authorised review route already available | Multiple environments or unresolved permission questions |
| Decision ownership | Available owners with defined authority | Unresolved responsibilities across brands or units |
| Deliverable | Bounded decision and evidence record | Additional executive, technical or procurement outputs |
The GAO Cost Estimating and Assessment Guide describes estimating practices including scope, work breakdown, assumptions and sensitivity analysis. It is a methodological reference, not a Saudi fee schedule. The worked comparison below applies those basic estimating ideas to a fictional readiness purchase.
Request a work breakdown that explains the total
A useful quotation can group effort into scoping, evidence review, owner discussions, language evaluation, conclusion preparation and customer review. The customer does not need to control every hour, but should understand what each activity contributes. Ask which output demonstrates completion and what happens if the expected evidence is unavailable.
For an illustrative eight-day assessment, the work could comprise one day for scope and current-workflow review, two for source sampling, one for access and owner questions, two for bilingual evaluation preparation, one for the decision pack and one for factual review and handover. Eight effort-days do not necessarily mean eight consecutive calendar days. Availability, review intervals and parallel work affect elapsed time.
Keep the breakdown proportionate. A very small engagement may use a simpler structure, while several independent workflows may need separate work packages. The purpose is to expose assumptions and dependencies, not to create an elaborate accounting exercise. If a quote contains a large undifferentiated “strategy” line, ask what decision and evidence it is expected to produce.
Normalise quotations before choosing the cheapest
Assume two fictional quotations for the furniture-care assessment. Quote A includes eight effort-days at an illustrative blended rate of three thousand riyals per day, producing a fee of twenty-four thousand riyals. Quote B includes six effort-days at three thousand two hundred riyals per day, producing nineteen thousand two hundred riyals. These are arithmetic examples only, with applicable taxes, travel and third-party charges excluded.
The apparent saving in B is four thousand eight hundred riyals. But B excludes the two days of bilingual evaluation preparation included in A. If B's provider offers the same additional work at its stated illustrative rate, the added amount is six thousand four hundred riyals. Its normalised fee becomes twenty-five thousand six hundred riyals, which is one thousand six hundred riyals more than A.
| Item | Quote A | Quote B |
|---|---|---|
| Initially quoted effort | 8 days | 6 days |
| Assumed daily rate | SAR 3,000 | SAR 3,200 |
| Initial fee | SAR 24,000 | SAR 19,200 |
| Missing bilingual work added for comparison | Included | 2 days × SAR 3,200 = SAR 6,400 |
| Fee for the aligned illustrative scope | SAR 24,000 | SAR 25,600 |
The comparison does not establish equal quality. After aligning scope, assess the proposed method, relevant competence, evidence access and acceptance terms. A bidder may use a different valid approach. Ask them to explain how it supports the same decision rather than insisting that every provider use identical staffing or hours.
Include the customer's work in the budget
An external fee is only part of the resources needed. Source owners must explain material, operational staff may walk through the workflow and reviewers must check the findings. If those people are unavailable, the engagement can stall or produce a narrower conclusion. A proposal that assumes unlimited customer availability may be less practical than one with an explicit participation plan.
Suppose two source owners each contribute six hours and four reviewers each contribute two hours. That is twelve plus eight, or twenty customer-hours. At an invented internal planning cost of one hundred and eighty riyals per hour, the allocation is three thousand six hundred riyals. This is a planning measure, not a claim about wages or an additional cash invoice.
Combined with illustrative Quote A, the external fee and internal time allocation total twenty-seven thousand six hundred riyals before excluded charges and any additional work. Keep the categories separate in the budget. Internal time may already be salaried, but it still competes with other work. Use ownership and adoption readiness to confirm that the required participation is realistic.
Choose a pricing model that matches uncertainty
A fixed fee can be useful when the decision, evidence boundary and deliverables are sufficiently clear. Its value depends on explicit assumptions and a workable change process. Confirm what happens if a source is missing, a workshop must be rescheduled or the customer adds another workflow. A fixed headline price is not a complete scope agreement.
Time-based pricing can suit exploratory work where effort is uncertain, but it needs a ceiling, reporting rhythm and stop decision. Ask what evidence will be available at each review point and who can authorise further effort. A capped discovery phase may help define a later fixed scope. The cap should not silently expand simply because the investigation found more interesting questions.
A staged engagement can separate initial scoping from deeper evidence review. Ensure each stage has a usable output and a decision about continuation. The customer should not have to purchase the entire programme to discover that the first stage never defined the intended workflow. Select the model for its fit to uncertainty and decision needs, not because one commercial label is universally superior.
Define what is excluded before discussing discounts
Readiness assessment, remediation, prototype development, production integration and ongoing operation are different kinds of work. A quotation may include limited technical exploration without including a deployable system. Ask for the boundary in terms of deliverables and authority: what will be examined, what will be built, what will be changed and what will remain a recommendation.
For the furniture business, reviewing manual conflicts is not the same as rewriting every manual or securing new content rights. Preparing bilingual acceptance cases is not the same as proving production answer quality. Identifying an access gap is not the same as implementing the access design. These distinctions help the buyer budget the next step without treating a readiness fee as a full solution price.
Use readiness versus proof of concept to agree the experiment boundary. Ask separately about taxes where applicable, travel, translation, specialist review, software usage and follow-up revisions. Do not assume these are included or excluded based on a short marketing description; they should be clear in the current quotation.
Make uncertainty visible through a sensitivity example
A useful estimate explains what could change it. Suppose the original scope covers a defined pair of brands and the customer later adds two more. If the fictional estimate assumes three-quarters of a day of additional source review for each new brand, the added effort is one and a half days. At the illustrative three-thousand-riyal rate, that adds four thousand five hundred riyals to external effort cost.
The revised external fee would be twenty-eight thousand five hundred riyals if that change and rate were agreed. This does not establish the actual effort for any brand. The example isolates one assumption so the customer can see its effect. Additional owner time or language work may also change and should be estimated separately rather than assumed to remain constant.
Ask for sensitivities around the uncertainties that matter: source conflicts, access delays, additional workflows or specialist review. Avoid adding an unexplained contingency percentage merely to make the estimate look prudent. A reserve should correspond to identifiable uncertainty and a rule for authorising its use. The customer should know which work is committed and which is only a possible next decision.
Agree how scope changes will be recorded
A change request should identify the new question, its reason, the affected evidence and the proposed effect on effort, timing and deliverables. For the furniture case, adding customer-facing advice changes more than output volume: it changes the audience, consequences and acceptance needs. Treat it as a new scope decision rather than a small wording adjustment.
Also distinguish a genuine addition from correcting an agreed deliverable that was incomplete. Review the proposal and acceptance terms rather than assuming every revision is billable or every new request is included. Clear documentation helps both parties resolve the difference. The assessment buyer should be able to see what changed before authorising additional spend.
Preserve the original estimate and the accepted revision. A final invoice that is understandable only through informal conversations is difficult to review. The commercial record should show the decision that created each material change, while avoiding unnecessary administrative detail for minor clarification. The goal is a transparent relationship between scope and price.
Evaluate deliverable quality independently of the fee
Ask how a material conclusion will be traced to evidence. The decision pack should show the workflow, examined sources, limitations, findings and conditions for the next step. A long presentation without source-level reasoning may be poor value, while a concise pack can be useful if its conclusions are supportable and the operational owner can act on them.
For the furniture-care assistant, a useful finding might identify that guidance for one product variant cannot be transferred to another without an authoritative source decision. The buyer should be able to see the affected material and the required owner action. A generic recommendation to improve data quality would not provide the same decision value.
Use the readiness decision pack to define acceptance and assessment failure diagnostics to identify warning signs in a proposed method. Do not require a positive readiness outcome as proof of value. A well-supported decision to narrow or pause a trial can be the most useful result of the engagement.
Compare assessor independence and implementation incentives
An assessor may also offer remediation or implementation. That can create useful continuity, but the buyer should understand how recommendations and subsequent commercial proposals will be separated. Ask whether the findings can be used by another delivery team, whether source evidence is available to the customer and how alternative next steps will be considered.
A useful assessment should allow a proportionate outcome, including a smaller trial, a non-AI process improvement or no immediate implementation. It should not require a particular product purchase to make the conclusion intelligible. If a proprietary scoring method is used, ask for enough explanation to understand the findings and their effect on the decision.
Review relevant competence for the actual task, including Arabic evaluation and source interpretation where required. Do not infer expertise from a broad AI label or from the size of a provider. A lower-cost specialist and a larger team may each be appropriate in different scopes. The evidence needed to judge fit should be specific to the work being purchased.
Connect payment milestones to reviewable work
Discuss commercial milestones in terms of deliverables the customer can inspect. A scope confirmation should identify the decision and evidence boundary. An evidence-review milestone should show what was examined and which questions remain open. A final handover should contain the supported conclusion and the agreed next actions. The actual payment arrangement belongs in the engagement's commercial terms; these examples describe what makes progress understandable.
Keep acceptance of the assessment separate from permission to launch the AI workflow. The furniture business could accept a complete, well-supported report that recommends delaying a customer-facing trial. Paying for the assessment does not approve the operational change. Conversely, an enthusiastic sponsor's wish to proceed does not establish that an incomplete deliverable has met its agreed acceptance criteria.
Clarify the opportunity to correct factual misunderstandings. If the assessor has described the wrong manual revision, the customer should have a defined way to provide the correct reference and see how it affects the conclusion. This is different from requiring the assessor to remove an inconvenient finding without evidence. A review window should support accuracy while preserving the independence of the analysis.
Also ask what the customer retains after the engagement. A usable evidence index, decision rationale and list of unresolved conditions can reduce the cost of future handovers. Access to a polished presentation alone may leave another team reconstructing the same investigation. Compare the practical reusability of the work alongside the fee, while respecting legitimate confidentiality and ownership arrangements stated in the agreement.
These questions help procurement judge value without pretending to predict the financial return of an untested AI system. The immediate purchase is a clearer, evidenced decision. Benefits from a later implementation require their own scope, baseline and evaluation.
Request a quotation that can be accepted and governed
Provide the decision, workflows, source families, languages, owners, known access limits and intended deliverable. Ask the provider to state its method, effort assumptions, fee model, exclusions, customer participation and change process. Request a clear description of the evidence that will be available at handover. These elements make quotations comparable without forcing artificial uniformity.
Use the illustrative calculations as a way to inspect arithmetic and scope, not as a target price to impose on a real provider. Current rates and terms must be confirmed for the actual engagement. The most useful commercial question is what uncertainty the proposed spend will resolve and whether the resulting decision is worth that work.
For a scoped discussion, use Enterprise Data & AI Readiness and send a readiness assessment enquiry. Bring the intended next step and a concise source index. The team can then discuss what needs to be examined and the conditions required for a concrete quotation.
FAQ
Your questions, answered.
How should the assessment fee change if two more brands are added after scope approval?
Treat the extra brands as a scope-change decision, identifying their sources, review needs and affected deliverables. In this article’s invented example, two additional brands require an assumed 0.75 review day each: 1.5 days × SAR 3,000 = SAR 4,500, taking the external fee from SAR 24,000 to SAR 28,500 if the change is approved. These figures are illustrative, not tariffs or market estimates. Assess source-owner time and additional language work separately. Retain the original quotation and written revision with its authorisation; do not turn the 0.75-day assumption into a standard allowance for every brand.
Why can a cheaper readiness quote become more expensive after comparison?
The initial scope may exclude work another quote includes, such as Arabic evaluation, source inspection or a decision-review session. Normalise the scope before comparing totals. In the fictional example, adding two missing bilingual days changes Quote B from SAR 19,200 to SAR 25,600. The arithmetic illustrates a method, not actual pricing or equal quality. Then assess competence, method and acceptance terms separately.
Should we choose a fixed fee or time-based assessment?
Choose according to scope clarity and uncertainty. A fixed fee needs clear assumptions, deliverables and change handling. Time-based work needs a cap, progress evidence and a decision before further effort. A staged approach can use initial discovery to define a later scope. None is universally best. Ask how the model protects a usable decision when sources are missing or the customer changes the task.
Are remediation and proof-of-concept development included in assessment pricing?
Only if the current quotation explicitly includes them. Assessing source conflicts, defining acceptance cases and identifying an access gap are different from correcting every source, building a prototype or implementing production controls. Describe each deliverable and what will actually be changed. Agree separate scope where needed so the customer can budget the next step without assuming that a readiness fee covers the whole AI solution.
How should we budget our internal effort for an assessment?
Identify source-owner interviews, workflow walkthroughs, evidence preparation and review of findings. Estimate the time actually required and confirm availability. Keep internal planning cost separate from external invoices and avoid presenting an assumed hourly allocation as a wage benchmark. Internal time may already be salaried, but it competes with other work and can constrain the assessment if it has not been reserved.
Does eight days of assessment effort mean delivery in eight calendar days?
No. Effort and elapsed time differ. Access arrangements, owner availability, customer review and work performed in parallel affect the schedule. Ask the provider to distinguish working effort from calendar assumptions and dependencies. A useful quotation explains which customer decisions or materials are needed to maintain the proposed timing and what happens if those prerequisites are delayed.
Should we pay only when the assessment concludes that we are ready?
A positive conclusion is not the appropriate sole measure of value. A defensible finding that a workflow should pause or narrow can prevent an unjustified investment. Agree acceptance around the decision, evidence, limitations and actionable conditions. The commercial terms should be reviewed for the actual engagement, but they should not encourage unsupported optimism merely to produce an acceptable readiness result.
What information makes a readiness quotation reviewable?
Provide the next-step decision, workflow count, source families, languages, business units, available owners and known access limits. Request the method, effort assumptions, deliverables, exclusions, customer participation, fee model and change process. Ask how findings will be traced to evidence and how the decision pack will be accepted. This creates a concrete basis for comparing offers beyond a headline fee or report length.