An affiliate sees approval rate slip and assumes the advertiser is cutting it. The advertiser sees the same drop and assumes the traffic got worse. Both sides look at the same numbers and reach opposite conclusions. That's why approval rate (the share of leads that end up confirmed and paid) is the most common point of conflict in CPA: it directly determines both the affiliate's earnings and the advertiser's budget. We covered why approval matters more to affiliates than the rate itself in What Makes a CPA Offer Attractive to Affiliates?. Here, we break down why approval drops, how it looks from each side, and what to do to bring it back.
One Metric, Two Perspectives
|
Question |
Affiliate |
Advertiser |
|
What approval rate means |
The share of traffic that actually earns money |
The share of paid leads from the target audience |
|
What counts as a good lead |
One that gets confirmed and paid |
One that meets the criteria and reaches a real customer |
|
How a drop looks |
"They're cutting my approval" |
"The traffic got worse" |
|
What's usually missing |
A specific reason for each rejection |
Data on which source the bad traffic came from |
Almost every approval drop falls into one of six causes below. In each of them, both the affiliate and the advertiser own part of the problem.
1. The Traffic Doesn't Match the Offer
The most common cause: leads keep coming in, but not from the audience the offer was built for.
-
Affiliate side: traffic comes from a GEO or audience the offer doesn't cover. Age, income, region, and other criteria get ignored because "it converts anyway."
-
Advertiser side: the offer's criteria are vague, or they changed and affiliates were never told.
-
How to fix it: write down the target audience and criteria in the offer terms, and announce any changes to affiliates in advance, not after the fact.
On Leadgid: offer terms are agreed with your account manager at the contract stage, so criteria don't have to be reverse-engineered from rejected leads.
2. Errors and Incomplete Data in the Lead
A lead can be real and still get rejected because of data problems.
-
Affiliate side: the form lets invalid phone numbers and emails through, autofill populates fields with errors, or the same person gets submitted multiple times.
-
Advertiser side: the form or CRM mishandles valid data, and good leads get filtered out at the door.
-
How to fix it: check validation on the form, set up deduplication, and test the application-to-CRM flow on a small volume before scaling.
3. Fraud and Incentivized Traffic
Bots and fake traffic drag approval down, and both sides lose: the affiliate loses earnings, the advertiser loses trust in the source.
-
Affiliate side: part of the traffic can turn out to be low quality even when the affiliate doesn't know it, for example when buying from unreliable sources.
-
Advertiser side: without automatic screening, fraud only surfaces after payment.
-
How to fix it: screen traffic before a lead is confirmed, not after, and investigate suspicious sources by SubID instead of cutting off the affiliate entirely (more in Tracking and Analytics: How to Know Where Your Customer Came From).
On Leadgid: built-in fraud detection screens traffic as it comes in, so advertisers pay only for genuine applications.
4. The Landing Page Promises One Thing, the Offer Delivers Another
A user submits a lead expecting one thing, gets something different during processing, and drops off.
-
Affiliate side: the pre-lander and creatives promise terms the offer doesn't have, such as an amount, a term, or "approval with no checks."
-
Advertiser side: the landing page isn't localized for the GEO, or its terms don't match what the user saw on the affiliate's page.
-
How to fix it: align creatives and pre-landers with the offer terms, and check that what the ad promises matches what the person actually gets.
5. How the Advertiser Handles Leads After Submission
Sometimes approval drops not because of the traffic, but because of what happens to the lead afterward.
-
Affiliate side: the affiliate can't see what happens to a lead after it's sent, so they can't tell bad traffic from bad processing.
-
Advertiser side: slow callbacks, an overloaded call center, or changed internal scoring criteria reduce the share of confirmed leads regardless of traffic quality.
-
How to fix it: monitor processing speed, document any change in criteria, and give a specific rejection reason for each lead instead of a generic one.
On Leadgid: approval stats are available in real time, broken down by rejection reason, so both sides can see exactly where leads are being lost.
6. Tracking Glitches
Sometimes a lead is confirmed, but the information never reaches the affiliate, and approval in their report is understated.
-
Affiliate side: discrepancies between the affiliate's own stats and the network report get mistaken for an approval cut.
-
Advertiser side: the postback isn't sent, or it's delayed, and statuses get lost.
-
How to fix it: check the postback setup, reconcile statuses for the same period, and only then draw conclusions about traffic quality.
An approval drop is rarely one side's fault. More often it's a mix of causes: vague criteria, weak validation, fraud in one source, or a processing problem. So the first step is to find the cause, not the culprit: look at rejection reasons, split traffic by source, and reconcile data from both sides. If you're still choosing a network and want to know in advance how it handles reporting and disputed leads, start with the checklist: How to Choose a CPA Network: A Checklist for Advertisers.
FAQ
- There's no universal number: it depends on the vertical, GEO, payout model, and the offer's criteria. It's more reliable to track your own trend and compare traffic sources against each other than to chase a "market average."
- Most often the cause is mixed. Breaking rejected leads down by reason and by traffic source shows where the problem sits: in the traffic, in the data, or in the processing.
- As soon as it's visible. The earlier both sides see the drop and its cause, the less budget and traffic get lost before the problem is fixed.
- It depends on the terms of the partnership. That's why approval criteria and the process for disputed leads should be agreed before launch, not after the first drop.
