A measured trial on your own numbers.
We are at the founding-site stage: choosing a small number of marketplaces for measured trials. Each trial reads covered orders against uncovered ones on the site's own numbers, under a protocol agreed in writing before it runs. Small volume is fine.
Marketplaces and services platforms in England and Wales that earn revenue on each transaction, from commission, a take rate or a booking fee. The cover has to be paid for out of that revenue. If your revenue per transaction does not carry it, we say so at the benchmark, not three meetings in.
The cover takes three kinds of dispute: marketplace goods; site services and money owed; deposits and holdings.
Five things, and none of them changes your own process.
Place the cover line
At the moment a transaction forms. One sentence from an approved set; your legal team reviews it once.
Route a dispute only at final refusal
One flag at the exit of your own dispute process. Nothing routes earlier, and we take no part while your process runs.
Pass the records for that dispute
With the participant's authorisation: the transaction, the contract terms and the refusal history. The participant gathers nothing.
Send your monthly transaction count
That count is the meter. It is reconciled against your records each month. How the fee works.
Carry the split for the trial
During the trial window, transactions are assigned at random to a covered group and an uncovered group. The participant does not see the split. After the trial, every transaction is covered.
The wording at the point of transaction.
Your site's name goes in place of "the site". Stated at purchase, so no participant finds it out later. The valuation report is the participant's to read and to keep. We never send you per-claim data. You receive counts and outcomes in aggregate, never an individual claim.
Some claims are routed and valued but not taken on, and the wording says so. The exclusions in full.
One measurement across all live founding sites.
The measurement falls at the earlier of a pooled transaction count set in the trial protocol, or twelve months from the first site's go-live. A small site's rate is warranted mostly on the network's numbers.
From our own records, we measure the disputes routed, admitted, valued and taken on, their outcomes and the time to each. From your records, we measure conversion and average transaction value in the covered group against the uncovered group. We also measure the refund and goodwill spend on disputes your process could not close.
You see your raw result, its standard error, the network figure, the weight on your own result and your warranted figure, with the arithmetic. Then the rate decision: the published rate, or a lower rate set from your warranted figure. A site may instead go live across the whole book from day one.
You can stop on one month's notice at any time. If the trial does not show the cover paying for itself, the rate never steps up. We can end the paid arrangement if submitted disputes exceed the share set in the written agreement, or if measured value comes in below the floor. Either way, disputes already routed are completed, and every transaction covered while you were with us stays covered.
Email dex@calmlyresolve.com with the subject "Measured start". Tell us the site and roughly what it earns on a typical transaction.
Start a measured trialWhat happens next, in order:
- We send the benchmark: a one-page estimate of what the cover would cost and carry on your numbers, from public data. Every inferred number is flagged as an estimate.
- Thirty minutes on your reaction. Five numbers matter most: average transaction value and its spread, take rate, cost per dispute, the share of disputes your process closes, and conversion by value band.
- If it holds up: the trial protocol in writing, then the written agreement, conditional on committed money standing behind the cover. Tom Manuel, the founder, joins at the point of terms.
Product screen, pre-launch. Demonstration data; figures are assumptions.