How many research requests can one offshore analyst handle each week?
A measurement protocol for estimating weekly research capacity from accepted deliverables, review time, and observed rework.

Question and scope
This protocol estimates how many research requests one offshore analyst can complete each week at a defined quality standard. The scope is one analyst working on a stable set of request types for a single GTM team. It excludes weeks with onboarding, extended leave, or a new system rollout. The question matters because a capacity number built from optimistic assumptions becomes a delivery promise the analyst cannot keep.
The estimate uses records the team already has: request intake, start and finish times, review notes, and rework history. It does not require a new survey or a time-tracking product. The goal is a range that a manager can defend when planning the next month. A single precise number would hide the variation that matters most in a small research function.
Define the unit of work
Define one unit as a request that ends in an accepted deliverable. A request that is cancelled before work begins is not a completed unit. A request that is reopened after acceptance becomes a new unit if the decision changed. An informal question answered in a chat message is not a unit unless it produces a documented deliverable that the team uses.
Classify request types before counting. Common types include a source note, a competitor update, a market summary, and a buyer interview synthesis. Each type has a different effort profile, so the capacity estimate should report the mix. An analyst who handles twenty short source notes has not shown the same capacity as one who handles twenty interview syntheses. The mix belongs beside the number.
Measure demand and incoming volume
Count the requests that arrived each week and the date each one entered the queue. Separate requests that were accepted from requests that were deferred or declined. A high incoming volume does not prove that the analyst was at capacity if most requests were out of scope. Record the intake source as well, because requests that arrive through private messages often bypass the queue and distort the count.
Look for bursts and recurring peaks. A launch week may produce five requests in two days, while a quiet week produces one. Report the median and the range, not only the average. A capacity plan built on a quiet month will fail in a launch month. The demand record should show whether the team has a steady flow or a pattern of spikes.
Measure productive hours
Start with the contracted hours and subtract the time that is not available for research. That includes approved leave, internal meetings the analyst must attend, and recurring administrative tasks. Do not subtract every interruption, because a small allowance for breaks and context switching is part of a normal workday. The result is the productive hours available for request work.
Record how the productive time was actually used across request types. If the analyst spends a third of the week on one recurring report, that time is not available for new requests. A capacity estimate that ignores recurring work will overstate the room for new demand. Show the recurring load separately so the manager can see the fixed part of the week.
Count review and waiting time
Research work often waits for a reviewer, a source owner, or a decision maker. Measure the wait separately from the analyst's working time. A request that takes two hours of analyst effort but waits three days for approval has a long cycle time and a short effort time. Both numbers matter, and confusing them leads to the wrong fix.
Count the reviewer's time as part of the delivery capacity when the reviewer is a fixed constraint. If every deliverable needs a manager review, the manager's available review slots limit weekly output. Record the number of reviews completed and the queue at the end of the week. The smaller of analyst capacity and review capacity is the real operating limit.
Adjust for quality and rework
Define the quality bar before measuring capacity. A deliverable might need a complete source log, a clear separation of fact and inference, and a reviewer sign-off. Sample completed work against that bar and record the share that needed correction. A capacity estimate that counts every first draft as finished will overstate what the team can actually rely on.
Add the observed correction time to the average effort per request. Use the actual rework rate rather than a target. If the rate is high because briefs are unclear, fix the brief before using the number to judge the analyst. The quality adjustment is a planning input, not a performance penalty. It shows the review and training time the work genuinely requires.
Use at least four weeks of data
Pull at least four consecutive weeks of history, or the full history if the engagement is shorter. Include weeks with a typical mix and note any unusual week in the report. Export the request records rather than relying on a weekly summary that may have already smoothed the data. Keep the raw export with the calculation so another person can repeat it.
Check the records for missing start times, duplicate entries, and requests that were closed without a deliverable. Correct what can be corrected and label what cannot. A small sample with a clear limitation is more useful than a large sample with silent gaps. The report should say how many records were excluded and why. Keep the correction notes with the sample so the next reviewer can see which gaps were data problems and which were recording problems.
Report a range, not a promise
Calculate the weekly completed units for each week in the sample. Report the median and the range from the slowest to the fastest typical week. Add the request mix and the review constraint beside the result. A range tells the manager what to expect across normal variation. A single average hides the weeks when the queue or the review load changed.
State the assumptions that produced the range. Include the quality bar, the productive hours, the recurring load, and the rework adjustment. If the assumptions change, the range changes. Present the estimate as a planning input for the next month, not as a permanent standard. The team should revisit it when the request mix or the review process changes.
Limits and sources of error
Time records are often incomplete, especially for short tasks and interruptions. The analyst may round hours, and the reviewer may not record review time at all. A capacity estimate built on incomplete records is still an estimate. Label it as such and avoid presenting it as a measured guarantee. The direction of the error matters, so note whether the missing time is likely to raise or lower the result.
The work mix can change after the estimate, and a new campaign or launch can create demand the history does not show. The cited public sources describe occupational outlook and business survey practice. They do not provide a capacity benchmark for this analyst or this team. Use the result as a local planning input and revisit it when the work changes.
Review cadence and next step
Review the estimate monthly while the scope is stable and after any material change in volume, access, or quality. Compare the observed completions with the plan and record the variance. If the variance persists, adjust the scope, the review process, or the support hours. Do not ask the analyst to absorb a structural mismatch through longer weeks.
Use the result to choose one change. If review is the constraint, add reviewer capacity or batch the reviews. If the brief is the constraint, fix the intake fields. If the mix is the constraint, rebalance the request types. Write the decision and the next review date beside the capacity range so the team can see whether the change helped.