Conversion rate is the share of visitors who do the thing you wanted: out of 100 clicks, how many signed up, installed or bought. Two percent means two out of every hundred. It is the single number that tells you whether your traffic and your offer fit each other.
Conversion rate (CR) is the number of conversions divided by the number of opportunities, expressed as a percentage: CR = conversions ÷ clicks × 100 in affiliate reporting, or conversions divided by sessions, impressions or users in other tools. The definition only means something once both halves are fixed. “Conversion” must name the event (a lead, an approved sale, a first app open) and the base must name what you divide by (unique clicks, all clicks, sessions). Two dashboards showing different conversion rates for the same campaign usually disagree on one of those two choices.
Ad platforms and analytics tools define the metric for their own purposes. Google Ads counts conversions per ad interaction according to the conversion actions you configured. Google Analytics 4 renamed what it previously called conversions to key events on March 21, 2024, reserving “conversions” for key events that are shared with Google Ads, precisely because advertisers kept comparing two numbers built on different bases. In an affiliate tracker the conversion is whatever the postback reports, which may be the initial action or only the approved one.
A funnel has several conversion rates, and treating them as one hides where the problem is. Ad impression to click is the click-through rate; click to prelander click is the prelander’s conversion rate; prelander click to offer action is the offer’s. The end-to-end conversion rate is the product of the steps. When the overall number drops, the fix is in the step whose rate fell, which is why every link in the funnel should be tracked separately.
Conversion rate is a ratio of two counts, so it comes with a sample-size problem. A campaign with 400 clicks and 8 conversions shows 2%, but the true rate could plausibly be anywhere between about 1% and 4%; one more or one fewer conversion moves it a quarter of a percent. Decisions to pause or scale should wait for enough conversions to narrow that range, and comparisons between placements with a handful of conversions each are mostly noise.
A native campaign shows an end-to-end conversion rate that fell from 1.2% to 0.6% in a week. Tracking each step separately shows where.
| Step | Last week | This week |
|---|---|---|
| Ad impression → click (CTR) | 0.9% | 0.9% |
| Click → prelander click-through | 40% | 40% |
| Prelander click → offer conversion | 3.0% | 1.5% |
| End-to-end (click → conversion) | 1.2% | 0.6% |
The ad and the prelander perform exactly as before; the offer page converts half as well. That points away from your creatives and toward the advertiser: a changed landing page, a new form field, a price change or a broken flow on one device. Without the per-step rates the natural reaction is to rewrite the ads, which would have changed nothing.
Every field on an offer page — payout, cap, GEO, flow, KPIs, hold — decides how you get paid or get burned.
Advanced · 12 min readThe daily and weekly optimisation loop: read the data, decide, then act — rules-based thresholds for scale, hold or kill, funnel-stage priorities,...
Advanced · 11 min readHow to diagnose affiliate tracking problems: which tracker, network and platform discrepancies are normal and which are alarming, plus broken postb...
Beginner · 12 min readIf you run campaigns on gut feel, you are flying blind with your own money.
It depends entirely on the action and the traffic. SOI forms on warm traffic can convert above 10%; sales on cold native traffic often convert below 1%. The useful benchmark is the network average for that specific offer, which your account manager can share, and your own rate over time.
Usually because of the base or the timing. The tracker may divide by all clicks while the network counts unique clicks, or the network may show only approved conversions while the tracker shows every postback received. Align the definitions before you compare.
Enough that one more or one fewer conversion does not change your decision. In practice most operators wait for a few dozen conversions per placement or creative before pausing or scaling it, and more when payouts are high and conversions rare.
On March 21, 2024 Google Analytics 4 started calling its tracked goal events “key events” and kept “conversions” for the key events shared with Google Ads. The data did not change; the naming now makes clear that GA4 and Google Ads conversion counts are built differently.
References are listed as plain text on purpose; look them up by title and publisher. Updated: 2026-10-05.
Run them inside a network with real tracking, roughly 48-hour payouts and a dedicated manager.
Join the network