Affiliate offer tracking: compare CPA offers by payout and EPC

A higher CPA payout does not automatically make an offer better. Compare payout, EPC, conversion state, and the attribution path on the same traffic before changing the route.

Tracked gives that comparison a concrete operating path: keep the page and CTA stable, route eligible offers behind a click slot, return the click ID through a postback, and inspect the selected offer in reports and journeys.

If your offers run through Digistore24, use the Digistore24 postback setup to carry Tracked's click ID through cid before comparing commission.

For other networks, follow the Everflow click-ID and Partner Postback recipe or the MaxBounty offer-tracking checklist. Each network has its own field and callback contract.

Payout and EPC answer different questions. Payout is the amount credited for one qualifying conversion. EPC means earnings per click, but its denominator and reporting window can vary. A network might use unique clicks, approved conversions, or a defined period. Record the definition before comparing numbers.

A clearly hypothetical payout-versus-EPC example

The arithmetic below is illustrative only. It is not a forecast, customer result, or promise about any offer.

Metric	Hypothetical Offer A	Hypothetical Offer B
Payout per approved conversion	$40	$70
Test clicks	100	100
Approved conversions	5	2
Gross payout	$200	$140
Hypothetical EPC	$2.00	$1.40

The calculation is gross payout divided by test clicks. Offer B pays more per conversion, but Offer A produces more gross payout and higher hypothetical EPC because more clicks became approved conversions.

Do not compare a network’s published EPC to your own raw payout without checking the denominator. Ask whether the figure includes rejected, pending, or reversed events, and whether it uses clicks, unique clicks, or another traffic unit. Also keep currency, fees, refunds, and chargebacks visible. “EPC” is useful only when both sides mean the same thing.

Keep the comparison like for like

Change one decision at a time. For a clean offer comparison, keep these inputs aligned:

• traffic source, audience, geography, device, and time window;
• landing page, page variation, headline, and primary CTA;
• click slot and outbound parameter mapping;
• attribution window and conversion event state;
• currency, fee treatment, approval rules, reversals, and reporting timezone;
• click count, exposure share, and the period used for the comparison.

If you change the page and offer together, the conversion-rate difference belongs to both. Mobile traffic versus desktop traffic also breaks the comparison. Write down the contract before routing the first click.

Use a stable click slot for the offer test

Keep the landing page CTA pointed at a campaign-relative path such as /click/1. Attach both eligible offers to that slot and assign deliberate weights. The page remains stable while the campaign selects an offer behind the click.

This fits offers with the same intent. If they require different copy or disclosures, use separate page or click-slot designs and label the test accordingly.

Weights are user-configured routing controls. A 50/50 setting will not divide every small block exactly in half, and Tracked does not turn it into an automatic winner decision. The weighted offer rotation guide covers slots, pools, and test design.

Follow the postback chain

The offer comparison is only as good as the attribution evidence behind it:

Ad → Tracked campaign URL → landing page → /click/1 → Offer A or B
   → advertiser stores click ID → approved event → postback with click ID
   → Tracked report and event journey

Check each boundary:

1. Create the click. Open a fresh campaign visit, follow a /click/N offer slot, and capture the offer-click ID.
2. Forward the identifier. Confirm the offer request receives the real value, not a literal macro or an altered query string.
3. Store the identifier. The advertiser or network must retain it with the lead, order, or other conversion record.
4. Return the event. Call Tracked /postback with the saved click_id. An untyped callback is the primary conversion and can use the passed payout or configured offer payout. For a secondary funnel signal, send conversion_type or its type alias; explicitly typed callbacks stay secondary and receive zero payout, even if payout is sent.
5. Verify the result. Confirm the callback resolves to the selected offer and appears once in the expected report and journey.

An HTTP 200 proves transport at most. It does not prove that the click matched, the event was accepted, a duplicate was handled correctly, or the right offer received credit. The postback URL guide gives a fuller test sequence.

Read EPC with event quality

An EPC difference can come from conversion rate, payout, approval status, or the denominator. Inspect the event state behind the number. A high raw payout with pending or reversed events may not compare to a lower payout with approved events.

Use Tracked reports and event journeys to identify the selected page, slot, offer, callback, event, and supplied revenue. Keep approval and reversal rules in the comparison notes. If event names differ, map them to stable categories before comparing them; the event mapping and postback backfill guide explains that boundary.

If you extend the comparison to profit, keep payout separate from allocated spend. Tracked allocates campaign spend to an offer using mapped campaign and day offer-click share; that report is not the same as raw payout or EPC.

The practical decision is not “which offer advertises the largest payout?” It is “which offer produces the stronger, verified result under the same traffic and event definition?” If Offer B pays more but has a lower like-for-like EPC, investigate approval timing before changing the route.

Where Tracked fits

Tracked keeps the comparison attached to the campaign route and page. You can host, proxy, or redirect the destination, keep CTAs on stable slots, attach weighted offers, pass click IDs, receive postbacks, and inspect reports and journeys. Hosted pages can return to the visual editor when the next test needs a page change.

The advertiser or network still controls its payout contract and callback syntax. Use its actual macro names and event states. A configured offer split does not promise a payout or performance outcome.

Frequently asked questions

Is EPC the same as the offer payout?

No. Payout is tied to a qualifying conversion. EPC divides credited earnings by a defined click denominator, so it also reflects conversion rate, approval state, and the reporting definition.

How many clicks do I need before choosing an offer?

There is no universal click count that makes every comparison valid. Set a test window and exposure rule, keep the inputs aligned, and wait until the event and approval data you need are available. Avoid declaring a winner from an unexamined small sample.

Can I compare two offers on one landing page?

Yes. Keep the page and CTA stable, attach both offers to the same appropriate click slot, and record the selected offer in the report. Use separate slots when the offers require different intent, copy, or eligibility rules.

Start free and compare your next CPA offers in Tracked.