Agency Research Margin Calculator: AI-Moderated vs Traditional

An agency’s research margin is determined by the full cost of delivering the promised engagement. Interview price is one input. This worked model uses explicit assumptions so you can replace them with your own rates and hours.

Download the editable project budget. It includes sample choice, platform and recruitment costs, researcher effort, contingency, and contribution margin. Its volumes, hours, and client fee are illustrative inputs to replace. Use the agency research workflow to define delivery scope and the complete agency research guide to connect the budget to recurring client work.

What belongs in the project cost?

Use five cost lines: platform interviews, recruitment and incentives, researcher labor, delivery and project management, and overhead. Avoid counting panel recruitment twice when it is included in an all-in interview price. Include buyer-arranged incentives when using your own sample.

For User Intuition Starter voice, the platform price is $30 per quality interview using your sample. Standard panel recruitment brings the total to $60 per quality interview. Specialty audiences need a separate quote. Starter has no monthly fee; other modalities and plans have different terms.

A worked 30-interview project

The following is an illustrative planning model, not a recommended client fee or an observed agency margin. Assume a $6,000 client fee, 12 hours of design, review, and delivery at a loaded labor cost of $150 per hour, and overhead equal to 15% of direct costs. Assume no buyer-arranged incentives in the BYO column; add them if applicable.

The sample source changes contribution by $1,035 in this model because the overhead assumption also scales with direct costs. If your overhead is fixed per engagement, use a fixed line instead. Keep sales costs, taxes, and company-level fixed costs separate or include them explicitly; do not call a contribution estimate net profit.

How sensitive is the margin to review time?

Six additional hours at the assumed $150 hourly cost add $900 in labor and $135 in modeled overhead. Contribution falls to $1,860 with your sample or $825 with the standard panel. The corresponding margins are 31.00% and 13.75%.

That sensitivity is the reason to inspect real outputs before setting a fee. Interview automation may reduce fieldwork labor while leaving a difficult synthesis problem. Budget for contradictory evidence, stakeholder revisions, and a follow-up study when the initial evidence is insufficient.

How should agencies compare traditional and AI-moderated delivery?

Use the same research objective, audience, evidence standard, and final deliverable. Enter actual quotes for recruitment and human moderation alongside measured internal review time. Do not assume that human moderation costs disappear from every project: sensitive conversations or exploratory work may remain researcher-led.

Compare total dollars, calendar time, and research suitability together. An apparently cheaper method is a poor substitution if it cannot answer the question. A faster fieldwork stage also does not prove the client recommendation will be ready sooner.

What should a recurring-program model include?

Model each wave’s sample, new recruiting costs, interview usage, review hours, and client service. Add any subscription commitment and unused capacity. Subscription terms are available on the pricing page. Compare actual expected usage and current plan terms, not only the headline per-credit rate.

For the agency’s own retainers, define included work and how additional waves or revision rounds are priced. The pricing-model guide covers project, retainer, and subscription choices. The capacity guide helps keep sales promises tied to available review time.

Build the model from actual cost ownership

Assign each cost to the party that pays it. A client may supply the participant list while the agency pays incentives. A platform may include standard recruiting while a specialist recruiter handles one difficult segment. The agency may use an employee for analysis and a contractor for presentation work. Record these arrangements explicitly before adding the numbers. Otherwise an apparently complete total can omit costs that sit between organizations.

Use loaded labor cost for internal profitability analysis rather than confusing the employee’s cost with the agency’s client billing rate. If your model uses a billing rate as a proxy, label that choice so the result is interpreted consistently. A project can cover direct cash costs and still consume scarce senior time that would have earned more on another engagement. Track that opportunity separately rather than hiding it inside an arbitrary margin target.

Choose a consistent treatment of overhead. A percentage of direct costs is simple for a worked example, but your agency may allocate fixed project management time or use a company-level recovery rate. Apply the same method to alternatives and state what it includes. Changing the allocation rule halfway through a comparison can create a margin improvement that comes entirely from accounting choices.

Separate interview usage from project scope

Interview count is an input to the platform bill. Scope is the work the client expects. Adding ten interviews to an existing audience can be different from adding a new market, a new stakeholder group, or a second decision question. The latter changes guide design, recruitment, analysis, and presentation work even when the additional interview count is small.

Create a change log with the original scope, the requested addition, the cost effect, and the person who approved it. Update the model before committing to the change. If a client asks for another segment after fieldwork has begun, evaluate whether the original guide and recruitment plan are suitable. Extra interviews that do not support the revised question can increase costs without increasing usable evidence.

Make quality assumptions explicit as well. User Intuition bills for interviews that pass its quality checks, but a billable interview is not a guarantee that the full study answers the client’s decision question. Researchers may still need additional evidence or another method. Reserve review time and decide how follow-up work is scoped, rather than treating quality-only billing as a substitute for research design.

Test a downside case before setting the fee

Use at least three scenarios: the planned engagement, a recruiting complication, and a heavier review requirement. Keep the changes specific. In the recruiting case, substitute a specialist quote for the affected segment. In the review case, add the senior hours and revision rounds your team believes are plausible. Avoid multiplying every cost by the same arbitrary percentage when you know which assumptions are uncertain.

For each scenario, calculate the total delivery cost, contribution dollars, and contribution margin. Then ask whether the team can still deliver the agreed scope at that fee. A negative scenario does not automatically mean the project should be rejected. It may mean the scope needs a contingency, a staged decision, or a separately priced option. The purpose is to identify that choice before the engagement starts.

Model timing alongside margin. A project can be profitable while creating a cash gap between paying for fieldwork and receiving the client payment. Use the actual payment arrangements in your operational plan. This guide’s contribution examples do not include financing effects or determine how revenue should be recognized. Keep the model tied to the commercial terms your agency actually uses.

Review the estimate after the engagement

Compare every major assumption with the completed project: quality interviews, sample source, recruiting charges, design hours, review hours, and revisions. Distinguish an estimating error from a deliberate scope change. If the client requested extra work that was approved and billed, it should not appear as unexplained delivery inefficiency. If review took longer because the guide missed a learning goal, record the methodological cause.

Use a short post-project note to improve the next estimate. Identify one change to the brief, sample plan, or review process that would have reduced avoidable work. Do not assume that faster delivery on a familiar topic will transfer to a new category. Keep estimates specific enough to reflect the differences between repeat tracking, exploratory work, and a complex strategic question.

At the portfolio level, track both demand and available researcher time. Lower platform costs create an opportunity only if the agency can win suitable projects and deliver them well. Additional interview capacity does not create client demand or unlimited synthesis capacity. Scale the number of commitments after the team has demonstrated that its estimates hold across more than one project.

Keep client pricing independent of platform prices

The agency decides what to charge for the responsibility and work it undertakes. A low platform bill does not establish a high defensible client fee, and a high platform bill does not establish that the engagement is valuable. Use the model to understand your economics, then make the client proposal clear about scope, evidence, delivery, and the decision the work is meant to inform.

Inspect the workflow before your next client brief

Hear actual calls and see the sample presentation from a 43-participant Walmart shopper example. The two call recordings, transcripts, and 15-slide readout let you assess probing depth, traceability, and how study limits are communicated. The presentation download is a PDF.

Use the agency research page for client-delivery workflows, the consulting page for engagement research, or the insights-team page for internal research programs. Bring one real brief to the evaluation. Agree the decision, sample requirements, guide, cost basis, and review standard before launching.

Use the agency research pricing models to choose how to charge for the work, and the proposal template to make deliverables and exclusions explicit.