Direct answer: give every completed PPV campaign a stable ID and record the asset version, offer, eligible audience rule, excluded prior buyers, sent count, delivered count, unlock count, observed revenue, creator workload, evaluation window and evidence grade. Calculate rates only from matching denominators. Compare campaigns only when their audience, window and accounting rules are sufficiently aligned.
This tracker records completed sends. It does not plan a PPV, recommend a price, write a message, decide whether to resend or provide revenue benchmarks. Its job is to stop campaign memory from collapsing into “that one did well.”
Use the PPV message planner before the send. Transfer its Campaign ID, audience rule, asset version and planned conditions into the tracker rather than recreating them after results are known.
Set the campaign unit
One row should represent one offer, asset version, audience rule, send event and evaluation window. If price, asset, audience or message changes materially, open a new Campaign ID. Do not combine several sends into one result simply because they occurred during the same week.
Record page/account, time zone, sent timestamp and evaluation cutoff. A campaign with a 24-hour observation window should not be ranked against one observed for a month without a clearly aligned snapshot.
Keep a parent Experiment ID when several rows are intentional variations. The parent groups the work; each campaign row preserves the actual conditions.
Freeze audience eligibility and exclusions
Write the audience rule as an observable state. Record the eligible count before sending, the selected count, and exclusions such as previous purchasers, invalid state, creator pause or another documented rule. Do not infer audience personality from spending alone.
Keep prior-buyer exclusion evidence attached to the campaign record. A duplicate asset name or missing purchase history can make the result unusable even when platform totals appear complete.
Compare audience overlap between campaigns. A repeat result from largely the same recipients answers a different question from a result among new eligible subscribers.
Identify the exact asset and offer
Record Asset ID, edit/version, duration or set size where relevant, preview version, description version and creator-defined category. Store a reference, not sensitive media, in the tracker.
Record the displayed offer and currency exactly as shown. Keep discounts, bundles or access conditions visible. The tracker does not decide whether the offer was optimal; it preserves what the audience actually encountered.
Connect the offer to the offer ladder worksheet only if the creator has a documented ladder. Do not retroactively label a campaign to make its place in a strategy look intentional.
Separate sent, delivered, opened and unlocked states
Use the exact states available from the creator's current source. Sent is not delivered, delivered is not opened, and opened is not unlocked. When a state is unavailable, write unavailable rather than estimating it from another total.
For each rate, show the formula. Delivery rate equals delivered divided by sent when both are valid. Unlock rate can use delivered or another declared eligible exposure denominator, but the denominator must be named beside the result and held consistent in comparisons.
Check current platform definitions at OnlyFans Terms and within the live analytics surface. Third-party labels can drift from what the platform currently reports.
Record revenue without double counting
Capture observed gross unlock revenue tied to the Campaign ID and evaluation window, any available platform deductions as separately sourced fields, refunds or reversals visible during the window, and the selected net definition. Never combine gross campaign revenue with a net comparison row.
Revenue per delivered record equals supported campaign revenue divided by delivered count. Revenue per unlock equals supported campaign revenue divided by unlocks. When the denominator is zero, display not calculable rather than zero.
If several campaigns can claim the same purchase, mark attribution unresolved. Do not allocate the value to whichever campaign has the strongest headline.
Count creator workload
Record planning, asset preparation, audience preparation, review, sending, response handling and reporting minutes. Use actual time where practical and flag estimates. Two campaigns with similar revenue can create very different workloads.
Keep reusable asset-production cost separate from campaign-specific effort unless the creator applies a consistent allocation rule. The PPV break-even calculator owns forward-looking cost coverage; this tracker records what happened.
Add workload notes for unusually high support or clarification demand. That experience evidence can matter even when unlock totals look acceptable.
Copy the PPV campaign tracker workflow
Artifact: copy this header row: Campaign ID | parent test | account | asset ID | asset version | offer | currency | audience rule | eligible count | prior-buyer exclusions | other exclusions | selected count | sent timestamp | time zone | sent | delivered | opened | unlocked | revenue definition | observed revenue | reversals | evaluation cutoff | planning minutes | preparation minutes | response minutes | reporting minutes | evidence grade | confounders | reviewer | decision handoff.
Add derived fields: delivery fraction | unlock fraction and named denominator | revenue per delivered | revenue per unlock | total creator minutes | minutes per unlock. Store the numerator and denominator beside every calculated rate.
A fictional record might show 120 selected, 116 delivered, 9 unlocked and an observation window ending after seven days. Those values demonstrate calculations only. They are not a target, benchmark or prediction for another creator.
Grade evidence before comparing rows
Use direct when the campaign, recipient state, purchase record and window align. Use directional when relevant movement appears but attribution or audience overlap remains uncertain. Use unavailable when a required denominator or source cannot be reconstructed.
Filter comparisons to matching definitions. If one campaign counts sent recipients and another counts delivered recipients, show separate columns or recompute from raw counts. Do not rank incompatible rates.
The UK Government Service Manual's guidance on setting service performance metrics supports defining measures around an actual service purpose rather than selecting isolated vanity totals.
Close each campaign with a factual review
Summarise conditions, observed counts, workload, evidence grade, unresolved issues and the next decision owner. The next step may be archive, investigate, design a new test or consider a new send; the tracker itself does not authorise that decision.
Lock the completed row. Corrections should preserve previous value, new value, reason, source and timestamp. This prevents historical performance from moving silently when reporting definitions change.
Audit a sample against source records. Check Campaign IDs, asset versions, prior-buyer exclusions, denominators, windows and revenue definitions. A sophisticated dashboard cannot rescue inconsistent inputs.
Compare campaigns with a matching-rules view
Create a filtered view that includes only rows with the same audience definition, revenue definition, evaluation window and unlock denominator. Display the exact comparison rule above the table. This is stricter than placing every campaign on one chart, but it prevents incompatible rows from producing a false ranking.
When a rule differs, do not discard the campaign. Keep it in the source tracker and assign a comparison-group label. The row can support its own factual review or join future campaigns that use the same definitions.
For asset-version tests, require the same audience preparation and window unless the variation intentionally changes them. If several conditions differ, describe the comparison as exploratory rather than attributing the result to one creative change.
Reconcile the campaign record
At the evaluation cutoff, reconcile sent and delivery states with the current source, prior-buyer exclusions with the purchase record, unlock count with observed revenue, and creator minutes with the workload note. Record the reconciliation timestamp and reviewer.
Do not force totals to match by editing the smaller number. Differences can reveal reversals, delayed events, duplicate records or incompatible windows. Preserve the discrepancy, investigate its source and decide whether the evidence grade must change.
If the platform later updates a total, add a new snapshot rather than erasing the cutoff value. A stable historical view makes later campaign comparisons reproducible.
Turn gaps into the next measurement task
Every unavailable field should create one bounded task: fix asset-version naming, preserve audience snapshots, standardise evaluation cutoffs, record prior purchasers or clarify the revenue definition. Assign an owner and due date before the next comparable campaign.
Prioritise gaps that block the core denominator or purchaser exclusion. Decorative dashboard fields should not outrank the minimum evidence required to avoid duplicate experiences and false performance claims.
After the fix, test the tracker with one campaign and audit it manually. A clean single record is a stronger foundation than importing months of inconsistent history into a polished chart.
Keep a short data dictionary beside the sheet. Define every state, timestamp, revenue field and workload category in plain language so a second reviewer can reproduce the row without relying on the original operator's memory.
Create an amendment log for every corrected campaign: field, original value, replacement value, source, reason, reviewer and timestamp. Recalculate derived fractions from the amended raw values, then preserve both snapshots. This prevents a corrected row from silently changing a past report and makes it possible to explain why two exports disagree.
When importing older campaigns, assign a completeness grade before comparing them. A historical row with an asset and revenue total but no audience snapshot should remain descriptive. Do not fill its missing denominator with the current subscriber count or promote it into a comparable group simply to lengthen the trend line.
Limitations and source notes
Limitations: available platform states, attribution, reversals and audience overlap can constrain comparison. Results from one creator, offer or window do not establish universal PPV performance. Use the tracker to describe observed campaigns, not to promise future revenue.
The consistent-field tracker, evidence grades and calculation panel are original SirenCY editorial tools informed by official platform and government performance-measurement sources.