Find out whether recruiting providers creates enough completed work to recover its cost. Separate ready supply from paid activation, and test a fixed cohort before funding the next one.
A service marketplace recruits twelve specialists. Eight finish qualification. Four complete their first jobs. The founder reports a low cost per signup, but has not answered the investment question: how much contribution has that recruiting effort actually returned?
Marketplace supply acquisition cost measures what your business spends to bring providers into its network. For a launch decision, define the endpoint explicitly: an application, a qualified ready provider and a provider completing a first accepted, paid job are different outcomes. Then compare the original cohort’s cumulative platform contribution with its acquisition and activation cost.
Founder verdict: Conditional. Recruit more when a defined supply gap serves real buyer needs, qualified providers accept the economics, and observed contribution supports another bounded investment. Stop scaling the channel when it mainly produces profiles, duplicates existing capacity or depends on hypothetical lifetime value to excuse current losses.
Choose the supply investment you are testing
This is not a general onboarding checklist or a formula for setting your commission. It is a decision about funding the next provider cohort through one acquisition channel. Keep the service, qualification standard and cost boundary stable enough to interpret the result.
In Sharetribe’s initial-supply guide, Studiotime founder Mike Williams describes segmenting suppliers, testing outreach, helping with listing creation and delivering value to early participants. Those are useful acquisition activities. Each also consumes time or money that a founder should record rather than treating assistance as free.
Constructed scenario: a network connects small-business teams with independent spreadsheet-cleanup specialists. Customers supply an authorized, non-sensitive workbook and buy a bounded cleanup against an agreed checklist. The offer excludes accounting judgments, financial advice, production-system access and guarantees about business results. The service, demand, prices and outcomes below are illustrative assumptions—not a customer story.
| Element | Definition to test | What could invalidate it? |
|---|---|---|
| Buyer | A team lead with a real workbook problem, permitted inputs and a named approver. | Interest never becomes an accepted paid order. |
| Offer | One disclosed workbook-cleanup unit with input limits, exclusions and acceptance evidence. | Most requests require unpriced bespoke investigation. |
| Provider | A specialist with reviewed samples, suitable competence, accepted terms and confirmed capacity. | People join but cannot or will not accept the actual work. |
| Acquisition | One bounded direct-outreach campaign to a defined provider segment. | Another batch adds names without useful coverage. |
| Money and repeat | A disclosed transaction fee funds coordination; providers retain agreed proceeds. Later jobs may recover recruiting cost. | Repeat transactions fail to appear or require losses to maintain. |
Test demand separately through a recorded buyer channel. Ask about a current eligible workbook, show the actual scope and price, and record paid commitments and reasons for declining. More provider signups cannot create that demand. The marketplace liquidity scorecard addresses whether requests reach acceptable fulfillment; this article asks whether the provider investment is paying back.
Define readiness and activation before calculating cost
Use a short funnel whose stages have evidence, not optimistic labels:
- Contacted: an identifiable prospect received the recruitment offer through the recorded channel.
- Applied: the prospect submitted the required information and consented to the relevant process.
- Qualified: a reviewer checked relevant competence, samples and applicable participation requirements.
- Ready: the qualified provider accepted the service boundaries and compensation and confirmed usable availability.
- Activated: the provider completed a first qualifying job that was accepted and paid under the defined reporting convention.
Activation is a chosen measurement convention here, not a universal industry definition. Retain completion, acceptance and payment dates separately. If payment is pending, label it pending; do not silently move it into the paid-activation count.
Andreessen Horowitz’s marketplace metrics framework emphasizes successful matching and examining unsuccessful outcomes. Apply that discipline to supply: a reviewed profile is not evidence that a provider has found worthwhile work.
Check whether the provider wants the real offer
Use a permitted sample workbook and a compensated skills exercise where appropriate. Ask the specialist to identify excluded work, estimate effort and state acceptable proceeds. Confirm available windows and what preparation is included. Qualification should not be relaxed to improve activation statistics.
For an illustrative $200 job, a 20% platform fee leaves $160 to the provider before the provider’s costs. At an assumed three hours valued at $40, labor consumes $120, leaving $40 before other costs. Five hours consumes $200. A completed job can therefore activate the provider while giving them a reason never to return.
Ask whether they would accept another comparable assignment after seeing actual effort. Do not promise guaranteed work, earnings or preferential matching merely to get an application.
Build one cost pool, including people who never activate
Fix an entry rule before recruitment begins—for example, providers first contacted through this campaign during one stated week. Resolve duplicate identities and prior participants through review. Do not classify the same person as a new acquisition every time they change an email address.
Include the attributable work of researching prospects, outreach, follow-up, screening, sample assessment, setup assistance and activation support. Count founder labor at an explicit planning cost even when no wage is paid. Record cash spending separately so contribution economics and cash timing are not confused.
Allocate shared costs consistently and disclose the basis. If a campaign also recruits providers for another service, do not charge its entire cost to both cohorts. Keep one-time platform development visible but separate from the repeatable recruiting-cost pool; exclude it only with an explicit explanation.
Then report several denominators without pretending they answer the same question:
- Cost per applicant: defined cohort cost ÷ applicants.
- Cost per ready provider: that same cost ÷ providers meeting the readiness rule.
- Cost per activated provider: that same cost ÷ providers reaching the first-job rule by the cutoff.
If the denominator is zero, report no observed activations and the full spend—not a zero acquisition cost. Later activations can change the measured ratio, but they do not erase the earlier result or the time spent waiting.
Calculate observed payback without inventing a lifetime
Sharetribe’s product-market-fit guide connects acquisition and service economics with repeated marketplace activity. For an early cohort, however, a long-run lifetime estimate may be unsupported. Begin with contribution actually observed by a stated elapsed-age cutoff.
Use a consistent job contribution measure: platform fees retained, less attributable delivery coordination, transaction costs, recoveries and buyer-acquisition allocation. Exclude the supply-acquisition pool from that per-job figure because you will deduct it once at cohort level. The marketplace take-rate worksheet helps distinguish customer service value, retained fees and provider proceeds.
Cohort recovery balance = cumulative qualifying job contribution − original supply acquisition and activation cost. Payback has been observed when this balance first reaches nonnegative under your stated measure. It is not net profit, a lifetime-value estimate or proof that a later refund cannot reverse the result.
A cohort that looks cheap at signup but has not paid back
Assume the campaign contacts twelve providers, all of whom apply. Eight become ready and four activate by a common 60-day age. Recruiting, qualification and activation support cost $1,200 in total. All figures are illustrative planning inputs, not market benchmarks.
| Measure | Calculation | Result |
|---|---|---|
| Per applicant | $1,200 ÷ 12 | $100 |
| Per ready provider | $1,200 ÷ 8 | $150 |
| Per activated provider | $1,200 ÷ 4 | $300 |
| Observed contribution | 20 qualifying jobs × $25 | $500 |
| Recovery balance | $500 − $1,200 | −$700 at day 60 |
The assumed $25 contribution could consist of a $40 platform fee less $8 of delivery coordination, $3 of other transaction-variable costs and $4 of buyer-acquisition allocation. Provider proceeds are already outside the $40 fee; do not subtract them again. Replace allowances with reconciled actual costs and account for failed buyer requests as well as completed jobs.
At a constant $25 contribution, recovering $1,200 requires 48 qualifying jobs in total. With 20 observed, another 28 would be needed under those unchanged assumptions. That is a break-even requirement, not a prediction that providers will obtain those jobs or that the contribution will remain constant.
Do not remove the eight providers who never activated from the cost pool. Do not count the same transaction twice because both a buyer cohort and a provider cohort refer to it. You may analyze the transaction through both views, but reconcile to one business total. Define allocation explicitly if a job uses multiple providers.
Copy a fixed-age provider-cohort ledger
Use one row per original provider and a linked transaction/cost register. A spreadsheet is sufficient for a small pilot; the useful asset is a reconciled definition.
| Record group | Retain | Review condition |
|---|---|---|
| Identity and entry | Provider reference, first-contact date, channel, cohort, existing/new status and offer version. | Review duplicates; keep the original entry rule. |
| Stage evidence | Application, qualification decision, terms acceptance, availability confirmation and first qualifying job dates. | Missing evidence remains pending, not retrospectively invented. |
| Cost | Cash, labor minutes/rate, shared-cost allocation, setup assistance and recurring support boundary. | Count each cost once; retain rejected and non-activated providers. |
| Contribution | Unique job ID, provider, retained fee, buyer-acquisition allocation, delivery costs, credits and recovery adjustments. | Reconcile to the transaction register; no GMV-as-profit shortcut. |
| Age and decision | Common elapsed-age cutoff, total eligible providers, activators, contribution, recovery balance and open obligations. | Immature cohorts remain pending; reviewer chooses continue, rework or stop. |
Compare cohorts at the same age since entry, not simply on the same calendar day. If the newest provider has only ten days of observation, do not present the entire batch as a completed 60-day cohort. State the unresolved observation period.
Keep inactivity reasons distinct: no suitable buyer demand, unavailable dates, rejected compensation, missing requirements, unresolved inputs or provider withdrawal. A channel can recruit excellent specialists into a service with insufficient demand. Blaming onboarding for every zero-job record leads to the wrong spending decision.
Run a capped recruitment experiment
- Hypothesis: one outreach channel can recruit willing, qualified spreadsheet specialists whose completed work recovers the defined recruiting cost within the founder’s chosen observation horizon.
- Test: cap one cohort at twelve providers and $1,200 of combined cash and valued labor. Confirm actual buyer opportunities before expanding ready supply. Observe everyone to the same 60-day age; do not promise every recruit work.
- Signals: stage transitions, reasons for non-activation, paid accepted jobs, provider effort and repeat willingness, actual contribution, recovery balance and outstanding obligations.
- Illustrative gates: require all serving providers to pass qualification and accept actual terms; require nonnegative cohort recovery by day 60, no concealed unresolved obligations and a useful explanation for every non-activation. A founder may choose a different horizon and loss limit prospectively; these are not industry benchmarks.
- Decision: continue a limited next cohort only when supply remains necessary and the gates pass. Rework a named bottleneck when the cause is supported. Stop buying more supply through the channel if attainable work cannot support its cost or providers reject the realized economics.
The worked example fails the recovery gate: it is $700 short. That does not prove the provider relationship has no future value. It means payback has not yet been observed under the chosen rule. Any extended observation or further investment needs an explicit new limit, not a rewritten history.
Build This Business in Workhint
Workhint’s Workflow Engine describes configurable request steps, assignments, role-routed approvals and required details or files. The Workhint solution for founders describes customer/provider onboarding and separate roles around marketplace requests. Those capabilities fit the evidence handoffs in a reviewed provider-to-work pilot.
Neither page establishes automatic acquisition-cost attribution, identity deduplication, cohort calculations or payback forecasting. Keep the analytical ledger and financial execution in separately verified processes. Treat this copy-ready prompt as desired configuration requirements to test, not a claim that every requested behavior is prebuilt:
Build an operator-reviewed marketplace for small-business teams buying bounded spreadsheet cleanup from qualified independent specialists. Give customers, providers, a qualification reviewer, an operations owner and a commercial reviewer appropriate separate roles and access. Capture permitted non-sensitive inputs, workbook limits, requested outcome, exclusions, acceptance checklist, delivery window, customer price and agreed provider proceeds.
Capture provider source/channel, first contact, original cohort reference, relevant samples, competence review, accepted terms and confirmed availability. Use Prospect, Applied, Qualification Review, Details Needed, Ready, Paused and Withdrawn as proposed provider states. Record first qualifying completion, acceptance and externally confirmed payment as separate evidence for human activation review. Do not automatically treat Ready as activated.
For customer work use Requested, Inputs Needed, Scope Review, Provider Acceptance, In Progress, Delivery Review, Correction Needed and Closed. Match manually by competence and capacity. Require accepted scope and provider agreement before commitment; require completion evidence and customer review before closure. Route missing inputs to customers, declines and rescheduling to operations, and cancellations or disputed delivery to a named reviewer. Keep provider qualification separate from the pressure to meet a recruitment target.
Retain unique provider and job references, cohort/cost-record references, operator minutes, completion evidence, exceptions and approved financial adjustments. Validate the record arrangement and permissions with test cases. Administrators review duplicates and missing evidence manually. Collection, provider payments, refunds and accounting remain in separately verified processes with reviewed references. Calculate fixed-age acquisition cost, activation and cumulative contribution externally; the commercial reviewer records the next-cohort decision. The outcome is traceable evidence for a recruitment investment decision, not automated profitability proof.
Scale useful supply, not a cheaper signup count
Launch one service and one recruitment channel with explicit spend and customer-commitment caps. Before increasing either, have a non-founder reviewer reproduce the cohort calculation and explain the non-activation cases. Check whether the next cohort addresses an actual skill or availability gap rather than competing for the same scarce jobs.
More repeatable qualification may reduce effort, but verify the reduction in later cohorts rather than crediting promised automation today. The broader service-marketplace operating model connects demand, supply and fulfillment. This investment gate keeps their combined economics visible when growth is tempting.
FAQ
Is cost per provider signup a useful metric?
It can diagnose an acquisition step, but it cannot establish usable supply or payback. Show the stage definition, full cost boundary and later outcomes alongside it.
Should providers with no jobs be removed from the cohort?
No. Their recruiting costs are part of the original investment. Retain them and investigate why they did not activate.
What is a good supply acquisition payback period?
There is no universal horizon for this decision. Choose a prospective limit consistent with your cash capacity, service frequency and evidence. State when the cohort is too young to evaluate.
Does positive cohort recovery mean the marketplace is profitable?
No. It only covers the costs included in that measure. Fixed overhead, other acquisition activity, platform development, taxes and unresolved or later adjustments can change the business result.

Leave a Reply