Pay per call disputes are how a buyer or a publisher says a specific call did not meet a specific rule. They are not a mood about the week, and they are not a request to be repaid because the caller did not buy. If you cannot name the call and the rule, you do not have a dispute. You have a wish that the invoice were smaller.
Do the listening before you file. A recording, where the campaign records, will show product, state, and whether anyone was actually on the line. A note that says 'short' without the audio is an argument. Call quality beyond duration is what the note is missing. Duration can prove a floor. It cannot prove the caller wanted what you sell.
What the file has to contain
A usable dispute names the event ID, the date and time, the amount, the rule that failed, and the evidence. Wrong state, with the ZIP the caller gave. Outside hours, with the timestamp in the campaign timezone. Duplicate, with the earlier event ID and the window in the order. Wrong product, with the words you heard. Already a customer, with the record you matched, not a memory.
'About 30% of these were junk' is not a dispute. It is a percentage with no calls behind it. Nobody can credit a percentage. Pull the IDs or do not file. If the pattern is a source, say the source and send a sample, then ask for the source to be paused. A pause is an operations move. A credit is a call-level move. They solve different problems. Mixing them is how a good path gets punished for three bad recordings.
The rule has to be one you both had before the call. A new dislike, discovered on Friday, does not reach backward. Pay per call filters are supposed to be in the order for this reason. If the filter was not written, the call that 'felt wrong' can still be billable. Fix the order for the next day. Pay the calls that met the old one.
The clock in the public terms
Your campaign order can set its own window. If it does not, the Terms and Conditions are the default, and they are specific. Routine quality disputes have to be submitted within seven business days after the event and the verification information you reasonably need are available. Invoice or settlement disputes have to be submitted within 15 calendar days of receipt. The notice has to carry event IDs, dates, amounts, the failed criterion, and the evidence, through a secure channel.
That is a contract default, not a suggestion that every RidgeRise order uses those exact days. Read your order first. If the order is silent, those are the days. General dissatisfaction is not enough. The terms say that in section 21, and they also say the receiving side should acknowledge within five business days and try to decide within 15 business days after the evidence is actually there. Those clocks are for the dispute process. They are not a promise about how fast a caller buys.
Late routine disputes are treated as accepted for settlement, except for calculation errors, concealed fraud, or compliance problems you could not reasonably have seen earlier. 'We got around to listening in week three' is not that exception if the audio was available in week one. Listen while the window is open. Do not use a late batch to hold unrelated calls. The terms also say a clawback is limited to the affected amount and the loss the agreement allows. An unresolved hunch is not a reason to hold the whole balance.
Credits, replacements, and what is not a dispute
When the call fails the written rule, the remedy in the order applies. If the order does not name a replacement, the incorrect charge comes off. A replacement call is a different event. Do not assume you are owed a free one because the first caller was rude. Rude and qualified can both be true.
These are not disputes. The agent talked over the caller. The price you quoted was higher than the ad the caller remembers, and the ad was yours. The homeowner chose another contractor after a fair visit. The senior said they would speak to a daughter. The injury victim did not sign. Those are the job you bought the call to attempt. Billable versus unqualified draws that line so you do not have to redraw it on every recording.
Fraud is a different pile. Fabricated calls, a source pretending a transfer was inbound, a caller coached to claim the wrong state. Pause first when you can show it. Then file with the IDs. A single odd call is a dispute. A source that does it on purpose is a cutoff. We can pause a path. We still need the calls, not a vibe.
Publishers reading this from the other side should send the same kind of file when a valid call was rejected: the ID, the rule you met, the audio or the log. A payout argument with no call behind it sits in the same pile as a buyer percentage. Questions to get in writing before you send traffic are in what to ask a pay-per-call network.
RidgeRise will review a dispute that names a rule we can check. We will not invent a credit rate, and we will not treat a missed sale as a bad call. The public terms are the fallback. The order is the campaign. Read both before you file.
If you are about to buy and you want the invalid list and the window in the brief, start a test campaign. Bring examples of calls you would reject and calls you would keep. That packet is more useful than a promise that disputes 'will not happen.' They will. The written rule is what makes them short.
