A CPA network has frozen a payout, named bots, misleading ads, or poor traffic quality, and asked for an explanation. The first task is to preserve tracker evidence. Do it before a tracker is cleaned, a domain is changed, a team member loses access, or raw logs rotate out. A screenshot of a dashboard is rarely enough. The dispute will turn on whether the records show a consistent route from click to conversion and whether that route can be verified.
Our legal practice works with traffic arbitrage teams, media buyers, and agencies in disputes over CPA and CPL payouts. CPA means payment for a target action. CPL means payment for a submitted lead. We review Keitaro and Binom tracker logs, separate real user activity from technical noise, and prepare the factual basis for a claim to a network, advertiser, bank, or court where appropriate. Russian-speaking teams often work with foreign networks, so correspondence and supporting materials may need to be prepared in English.
A hold and an ad account ban are different disputes. A hold concerns money already earned or claimed under the network terms. A Meta Business Manager or Google Ads restriction concerns platform access and policy enforcement. Do not combine them in one complaint. Each case needs its own facts, records, and requested remedy.
Preserve Tracker Evidence Before You Argue
The network may have already formed an internal view of the traffic. That does not mean its conclusion is final. But arguing with general statements such as “the traffic was clean” usually makes the position weaker. The response should begin with records: what was launched, when it was launched, where clicks went, what filters operated, and what happened after the click.
Start by making a protected copy of the tracker data. Export the relevant campaign reports for the disputed period in their original available format. Keep the date range visible. Preserve click logs, conversion logs, postback records, landing-page data, offer URLs, traffic-source parameters, sub IDs, and any available IP, user-agent, referrer, timestamp, device, language, and geographic fields.
Do not edit source files to make them easier to read. Create a separate working copy for analysis. The original export should remain unchanged, with the date and source noted in a short preservation record. If several people have tracker access, identify who exported the files and when. This is basic evidence discipline. It helps explain later that the data was not rebuilt after the dispute began.
Screenshots still have a role, but a limited one. They can record the state of a dashboard, campaign settings, cap status, payout terms, and notices inside the affiliate account. Take screenshots that show the full browser window, account identifier, and date where possible. They should support the raw data, not replace it.
What Records Usually Matter in a Hold Dispute
A useful evidence set tells one coherent story. The exact set depends on the allegation and the tracker configuration. For a claim of bot traffic, click and conversion timing, repeated technical identifiers, abnormal event density, and the pattern of postbacks can matter. For an allegation of misleading advertising, the relevant materials are usually creatives, pre-landing pages, landing pages, campaign settings, approval history, and the applicable terms.
Preserve the version of the network’s terms that applied when traffic was sent. Save the offer card, payout amount, cap, allowed geographies, prohibited methods, and messages from the account manager. If the network changed terms after the campaign started, the timeline becomes central. A current terms page may not prove what applied during the disputed dates.
Keep invoices, payout requests, prior payout confirmations, and correspondence about the hold. If the network provided a reason code or an anti-fraud report, retain the complete message. Do not quote only one sentence from it. The wording matters, including whether the network alleges invalid leads, duplicate registrations, incentivized traffic, policy violations, or a broad right to reject traffic under its terms.
Where there is a payment issue, the contract side and technical side must meet. The tracker may show that a conversion occurred, but it may not answer whether the advertiser accepted the lead or whether a contractual validation period expired. Conversely, a broad fraud allegation does not remove the need for the network to explain which traffic it rejected and on what basis. That is why we compare logs with the offer terms, payout history, and correspondence rather than relying on a single report.
How We Examine Keitaro and Binom Logs
Tracker logs are not self-explanatory. A high conversion rate is not automatically proof of fraud, just as a low conversion rate is not proof of valid traffic. The question is whether the data contains patterns that support or contradict the reason for the hold.
We first establish the disputed period, offer, campaigns, and claimed amount. Then we identify the data sources that can be preserved and compare timestamps across clicks, redirects, postbacks, and conversions. We check whether sub IDs connect traffic-source records to tracker events and whether the event sequence is technically plausible.
A review may find ordinary explanations for irregular-looking data. A delayed postback can create a timing gap. A redirect chain can affect referrer visibility. A traffic source may aggregate device data differently than the tracker. At the same time, identical technical markers across a large number of supposed users, impossible time intervals, or repeated conversion patterns may create a real problem for the claim.
The point is not to label every suspicious signal as fraud or every lead as valid. The point is to state what the available records actually show, what cannot be established, and what additional data the network or advertiser should disclose. That distinction is often more useful in a formal claim than emotional correspondence.
Do Not Damage the Evidence While Responding
The first response to a hold is often written in a hurry. Avoid sending altered tables, partial extracts without explanation, or unsupported promises to provide anything the network requests. Once a statement is sent, it can be compared against later documents.
Do not delete campaigns, change tracker settings for the disputed period, overwrite landing pages, or ask employees to clean their chat history. Operational changes may be necessary to stop further spend, but preserve the historical configuration first. Record when the campaign was paused and why.
Do not create retrospective documents. A recreated report can be useful as an explanation, but it must be clearly identified as an analytical summary. It is not the same as a contemporaneous tracker export. Mixing the two gives the other side an easy argument that the evidence was shaped after the fact.
If the network requests logs, clarify the period, requested fields, format, and purpose before sending a large data set. Some fields may be irrelevant to the dispute. Other fields may be essential, such as sub IDs and precise timestamps. A focused response is better than a data dump that creates new inconsistencies.
From Evidence to a Formal Claim
After preservation, the next step is an assessment. We determine whether the documents support a demand for payment, a request for a detailed fraud rationale, or a narrower attempt to resolve an accounting issue. Not every hold should become a claim. If the logs and offer terms show a material weakness, it is better to identify it early than spend time on a position that cannot be supported.
Where the evidence supports the client’s position, the claim should be specific. It should identify the parties, offer, period, amount, applicable terms, tracker records, communications, and the requested action. It should not make accusations that the records do not prove. If correspondence is with an international network or platform, we prepare the factual presentation in English when needed.
The same approach applies when a blocked advertising account affects the evidence. For a Meta restriction or Google Ads restriction, preserve the ban notice, account ID, campaign history, billing records, creatives, policy messages, and prior appeals. That material is used for an appeal about access. It should not be substituted for evidence in a payout hold dispute.
Questions Clients Ask
Can tracker screenshots prove valid traffic?
Usually, no. Screenshots help confirm what the interface displayed, but exports and logs are more useful for checking events, timestamps, redirects, and conversion records.
What if the network will not disclose its anti-fraud data?
The network may refuse to provide its internal scoring model or advertiser-side details. We can still assess its stated grounds against your tracker data, the offer terms, and the payment history, then formulate a reasoned request or claim.
Can we preserve evidence after access to the tracker is lost?
Sometimes. Existing exports, team archives, server backups, account notices, traffic-source reports, and correspondence may still establish part of the record. The available position depends on what was saved before access ended.
If a CPA or CPL payout is on hold, do not begin with a long explanation to the network. Preserve the original records, isolate the disputed period, and assess the evidence before making demands. Send us the hold notice, the stated reason, the offer terms, and available Keitaro or Binom exports. Our lawyers will review whether the facts support a payment claim and prepare the next communication without mixing it with a separate Meta or Google Ads account dispute.