Can a Small Market-Monitoring Service Work? Test the Deliverable First

A small market-monitoring service is viable only when a defined buyer repeatedly accepts a narrow deliverable that helps make a decision, and the work to deliver, correct and support it can be covered. Before buying proxy capacity or building alerts, validate that chain manually.

This is a feasibility method for an agency or new operator. It is not a demand forecast, income claim, IPHTML performance claim or price list.

1. Define one decision-ready service

Start with one buyer role, one recurring decision and one bounded report. Example: a category manager receives a weekday exception list for 20 approved competitor listings in one market and decides whether to start an internal review before 11:00 local time. The service supplies cited evidence and an exception label; it does not automatically change prices.

Do not buy a proxy when a manual check, buyer feed, official API or licensed dataset already passes the actual decision need. The source-selection guide covers the rights, coverage, freshness and provenance gates that still apply to a service.

2. Turn curiosity into evidence, not a promise

Ask a prospective design partner about the last time they made the decision: what triggered it, what evidence they used, who checked it, what arrived late or was distrusted, and what action followed. Show a blank sample report rather than promise coverage or accuracy. Record the result using a commitment ladder: curiosity (problem conversation), design partner (agrees scope and feedback), paid-pilot commitment (signs or agrees a fixed paid scope), payment received (verify it in a business record), then renewal payment received after observed delivery. A signed scope is not payment; each level is different evidence.

GOV.UK’s discovery guidance recommends learning who users are, what they do now and what barriers they face before planning a service; its research planning guidance treats assumptions as research questions. A public discussion about dashboards lacking actionable guidance is only a question worth testing, not market-size or willingness-to-pay evidence.

3. Deliver the pilot manually and record acceptance

Use a short fixed pilot: for example five delivery days, 20 approved records, one market and one report time. Manually verify matching, required fields and stated context. Record source time, report time, correction, buyer decision and minutes spent. An unpaid pilot must be labelled unpaid; its success does not prove a paid sale.

report_id, source_reference, observed_at, decision_deadline,
exception_rule, reviewer, accepted, rework_reason, action_taken

Predefine stop rules: missing permission, no owner acting, repeated unresolved mismatch, late delivery or effort above the cap. The representative trial guide can later define a fair authorized technical sample; it cannot validate commercial demand.

4. Contract the boundary and change cost

Write sources and permitted method, identifiers and fields, market context, delivery cadence, correction window, report owner, support boundary and retention. A new source, market, login-dependent field, backfill, report format or faster deadline is a new estimate, not invisible included work.

Proxy routing does not supply permission, matching, validation or support. On refusal, challenge or rate limit, pause and use an approved route; do not rotate identities to bypass it.

5. Test per-client economics and service capacity

Hypothetical example only. For one design partner, monthly costs are research/customer coordination 6 hours × $40 = $240; monitoring/validation 8 × $35 = $280; delivery/review 4 × $35 = $140; support/rework allowance 2 × $35 = $70; assumed source/network allowance $90. Recurring cost is $820. A hypothetical $1,100 monthly fee leaves $280 before tax, sales, overhead, refunds and omitted costs.

recurring cost = 240 + 280 + 140 + 70 + 90 = $820
break-even recurring fee = $820
hypothetical contribution = $1,100 − $820 = $280

Assume separate setup: 12 hours × $40 = $480. With all setup in month one, cost is $1,300 and the same hypothetical fee yields -$200. Two extra rework hours make recurring cost $890; a $60 larger source/network allowance makes it $880. These are sensitivity calculations, not a quote or forecast.

Download the bilingual service-feasibility worksheet. Leave the actual/planned input and calculated columns blank until the pilot; it does not execute formulas. Enter each hours and rate pair, calculate each cost, then follow the printed formulas for recurring contribution, month-one result, sensitivities and capacity. The example column is hypothetical only.

Capacity must also be explicit: if one client consumes 20 monthly research/customer-coordination, monitoring/validation, delivery/review and support/rework hours, a 40-hour service allowance supports at most two such clients before sales, administration, contingencies or change requests. Replace every input with observed pilot time. The internal cost model estimates observation work; this model adds acceptance, delivery, support, rework and scope control.

6. Decide whether a proxy belongs in the service

Only after a pilot demonstrates a recurring decision, accepted delivery and an eligible source route should an operator test whether authorized location-specific collection is needed. Bring approved sources, regions, measured transfer, frequency, session requirement, deadline, validator and stop conditions.

If those requirements need a residential route, review residential proxies and contact the team with the authorized workload and measured pilot needs. Confirm suitability, current terms and limits before committing. If a manual or first-party route works, keep it.

Sources checked October 20, 2026: linked GOV.UK Service Manual pages and a public discussion. The worksheet, contract boundary and arithmetic are editorial methods with hypothetical values. No customer, performance, price, income or market-demand result is claimed.