You Found Flagged Calls — Now What? A Broker's Action Playbook

Most pay-per-call operations eventually invest in some form of quality control. They listen to calls, flag the bad ones, and feel good about catching problems. But here is the uncomfortable truth: detection without action is just expensive observation.

The gap between flagging a call and doing something about it is where most operations leak money. You know the coached calls exist. You know which publishers send them. And yet the payouts keep going out, the bad traffic keeps flowing, and nothing structurally changes.

This is not a technology problem. It is a process problem. Here is the playbook.

Step 1: Verify the Flag

Not every flagged call is a confirmed issue. Before you take action, verify.

Pull the recording. Read the transcript. Understand the context. A flag for "coached call" needs human confirmation — was the caller genuinely being fed fabricated information, or was a transfer agent helping someone articulate a real situation? Those are very different scenarios.

A flag for a compliance issue might be a caller who clearly does not match the campaign vertical, or it might be an edge case that needs a judgment call. Either way, you need to listen before you act.

Build a simple verification workflow: flag comes in, assigned reviewer listens within 24 hours, reviewer marks the flag as confirmed, false positive, or needs escalation. Track your false positive rate — if it is above 10-15%, your detection criteria need tuning.

Step 2: Categorize Severity

Not all flags carry the same weight. A coached call where a caller is being fed false medical information to qualify for a health insurance transfer is a fundamentally different problem than a caller who slightly exaggerated their interest level.

Establish severity tiers:

Critical — Immediate action required. Coached calls with fabricated information, DNC violations, TCPA violations, clear fraud rings. These warrant same-day response: pause the publisher, hold payouts, begin documentation.

High — Action within 48 hours. Repeated compliance issues from the same publisher, callers who consistently misrepresent qualifications, patterns that suggest systematic problems rather than one-off incidents.

Medium — Monitor and document. Isolated compliance flags, borderline cases, first-time issues from otherwise clean publishers. Document the instance, note the date and details, and watch for patterns.

Low — Log and review monthly. Minor quality concerns, disposition mismatches, calls that are technically compliant but low value. These feed into publisher scorecards during regular reviews.

Step 3: Have the Publisher Conversation

This is where most brokers hesitate. Confronting a publisher about traffic quality feels adversarial, especially when they are sending volume. But data changes the dynamic entirely.

Lead with specifics, not accusations. "We flagged 14 calls from your traffic this week — here are the call IDs, timestamps, and the specific issues identified" is a fundamentally different conversation than "your traffic quality is bad."

Share the evidence. Provide transcript excerpts. Show the pattern. Most legitimate publishers want to fix quality issues because their reputation depends on it. The ones who get defensive or dismissive when presented with clear evidence are telling you something important.

Document every conversation. Date, attendees, what was discussed, what was agreed upon. This matters later.

Step 4: Execute the Clawback

When you have confirmed fraudulent calls, you need a clawback process. This is not optional — it is how you protect your margins and signal to your publisher network that fraud has financial consequences.

Your publisher agreements should already include clawback provisions. If they do not, fix that immediately. Standard language covers payouts for calls confirmed as fraudulent, coached, or non-compliant within a defined review window (typically 30-90 days).

The clawback process: compile the list of confirmed fraudulent call IDs with evidence, send a formal notice referencing the specific contract clause, deduct from the next payout cycle or invoice separately. Keep records of everything.

Publishers who push back on legitimate clawbacks with clear evidence are publishers you do not want in your network.

Step 5: Track Patterns, Not Just Incidents

A single flagged call is an incident. Five flagged calls from the same publisher in a week is a pattern. Fifteen flagged calls from the same sub-ID is a problem that requires structural response.

Track flag rates by publisher over time. Track them by campaign, by sub-ID, by time of day. Fraud is rarely random — it clusters around specific sources, specific time windows, and specific campaign types.

The brokers who catch fraud early are the ones who look at trends, not just individual flags. A publisher whose flag rate creeps from 2% to 5% to 8% over three weeks is showing you a trajectory. Act on the trajectory, not just the threshold breach.

Step 6: Enforce a Clear Escalation Framework

Publishers need to know the consequences before they face them. Publish your quality standards and make the escalation path explicit:

First offense (isolated incident): Written notice with evidence. Documented conversation. Clear expectations set.

Second offense (repeated issue): Formal warning. Reduced volume cap. Increased monitoring frequency. 30-day probation period.

Third offense (pattern confirmed): Publisher suspension. All payouts held pending full audit. Requirement to demonstrate corrective action before reinstatement.

Systematic fraud confirmed: Permanent removal from network. Full clawback of fraudulent payouts. Notification to industry contacts where appropriate.

The key is consistency. If you enforce consequences for one publisher but not another, your framework means nothing.

Step 7: Document Everything for Legal Protection

In the pay-per-call industry, disputes happen. Publishers challenge clawbacks. Buyers question traffic quality. Regulatory bodies request records. In every scenario, documentation is your shield.

For every confirmed flag, maintain: the original call recording and transcript, the specific flag and evidence supporting it, reviewer notes and verification, all publisher communications, any clawback actions taken, and the outcome.

This is not busywork. A single TCPA violation can carry penalties of $500 to $1,500 per call. When a buyer comes to you with a compliance concern, your ability to show a documented QC process with specific actions taken is the difference between retaining that relationship and losing it.

The Underlying Principle

Having the data is necessary but not sufficient. The brokers who actually reduce fraud and improve traffic quality are the ones who build a repeatable process around their data — verify, categorize, communicate, enforce, document, repeat.

Every step in this playbook gets easier and faster with practice. The first clawback conversation is awkward. The twentieth is routine. The first publisher removal feels risky. After the third, your remaining publishers understand that quality standards are real.

This playbook works at any scale, but the manual overhead grows with call volume. Tools like ConvoQC handle the detection layer — flagging coached calls, compliance issues, and DNC/TCPA concerns on every call within minutes — so your team can focus on what matters: verification, publisher communication, and enforcement. The process is what makes the difference. The tooling just determines how fast you can run it.