If you are shipping content and building links, the hard part is not doing the work — it is knowing whether the work is paying off. And most SEO measurement is theater: a screenshot of a keyword hitting number one, a Domain Rating that ticked up two points, a traffic line that is technically going up.
None of that pays rent. This guide separates the metrics that tell you something from the ones that just make you feel good, ties each to a tool that actually reports it, and shows you how to read Search Console and GA4 honestly in a 2026 landscape where impressions can climb while clicks fall and "rank number one" no longer means what it used to. The goal is not a prettier dashboard — it is knowing which pages to double down on and which to cut. For affiliates, that judgment ultimately runs on the same return logic as ROI vs ROAS.
Measurement exists to make decisions — kill, keep or scale a page or cluster — not to decorate a report. A vanity metric is one that rises without your business getting better, and three offenders dominate. Raw total traffic hides that a spike may be low-intent, branded, or a single viral post that never converts. Keyword rankings in isolation are hollow: a number-one result swallowed by an AI Overview, or ranking for a term nobody searches, tells you nothing without impressions, clicks and conversions around it. And Domain Rating or Domain Authority are third-party scores that Google does not use as ranking factors — they correlate with rankings only because both track backlinks, so treat them as a compass, not a scoreboard. The honest counter-metrics are the ones tied to outcomes: qualified, non-branded organic traffic, conversions, and revenue or earnings per visitor.
Group metrics by the decision they inform. Visibility — average position, tracked rankings, impressions and share of voice — tells you whether you are even eligible for traffic. Traffic — organic clicks in Search Console and sessions in GA4 — tells you whether people actually arrive, with the honest 2026 note that clicks are falling relative to impressions as AI answers absorb queries. Engagement — GA4 engagement rate, engaged sessions and average engagement time — tells you whether the traffic is the right traffic and the page satisfies intent. Conversions — key events, conversion rate, assisted conversions, and for affiliates revenue and earnings per click — tells you whether any of this makes money. And technical health — index coverage, Core Web Vitals and crawl stats — is a hygiene floor, telling you whether Google can even access and serve your pages.
| Metric | Tool | What it tells you | Lead / lag |
|---|---|---|---|
| Average position | Search Console | Roughly where you rank (it is an average) | Leading |
| Impressions | Search Console | How often you appeared | Leading |
| Organic clicks | Search Console | Actual visits from Google | Lagging |
| Engagement rate | GA4 | Whether traffic is the right traffic | Lagging |
| Conversions / key events | GA4 | Actions completed | Lagging |
| Revenue / EPC per visitor | Network + GA4 | The un-gameable bottom line | Lagging |
| Index coverage / CWV | Search Console | Can Google access and serve pages | Leading (hygiene) |
Three tools, three jobs. Google Search Console is the source of truth for how you appear in Google — impressions, clicks, click-through rate, average position, queries, indexing and Core Web Vitals — because it is Google's own first-party data; its limits are roughly sixteen months of history, sampled rows on high-volume sites, a blunt average position, and no indication of which queries triggered an AI Overview. GA4 is for on-site behavior and conversions — what people do after they land and whether they convert — but it measures sessions and events, not SERP appearance. Rank trackers like Ahrefs and Semrush give daily positions on specific keywords, competitor share of voice, and AI Overview presence, filling the gaps Search Console leaves; the caveat is that their traffic numbers are modeled estimates, not Google's ground truth. The one-line rule: Search Console for how you appear, GA4 for what visitors do, a rank tracker for how you stack up and where AI is eating your clicks. Getting the underlying tracking right is its own discipline, covered in affiliate tracking explained.
Open the Performance report, turn on all four metrics, and then filter — always. Sitewide numbers lie, so drill down by query, page, country and device, where the real decisions live. The single highest-ROI report for a content operator is the striking-distance filter: show positions eleven to twenty with decent impressions, and you find near-wins where an on-page refresh, a few internal links or a link build often pushes you onto page one. Learn to read the AI-Overview signal too — impressions up but clicks flat or down on a query set is the classic pattern of being shown while an AI answer satisfies the searcher; Search Console will not label it, so corroborate with a rank tracker. And keep hammering one caveat: average position is a weighted average across all impressions, so a jump from eight to six can be a mix shift rather than real improvement. Read position alongside impressions and clicks, never alone. Finding those opportunities connects back to the keyword research process.
Last-click attribution gives all the credit to the final touch before conversion. It is the affiliate industry default and how most networks actually pay, but it systematically under-credits SEO and top-of-funnel content — the informational post that started the journey gets zero last-click credit even though it kicked everything off. Assisted conversions surface those earlier touches so you can see the "best X for Y" article feeding the "X review" page that gets the last click. The practical takeaway is to judge informational pages on their assisted contribution and downstream movement, and commercial or comparison pages on last-click conversions and earnings per click. Tie your GA4 outbound-click events to actual commissions, because a page can look weak on conversion count yet earn well if its few clicks are high-value. The mechanics of multi-touch credit are covered in attribution models.
Track five to eight metrics, not thirty: non-branded organic clicks, organic impressions (to catch the impressions-up, clicks-down divergence), non-branded average position on a defined target keyword set, organic engagement rate or engaged sessions, organic conversions, conversion rate, and revenue or earnings per visitor, with indexed-page count and Core Web Vitals as an optional hygiene check. Three rules keep it honest. Run it monthly, because SEO is too slow for weekly reads and its feedback loop runs months. Segment by page or cluster rather than sitewide — a view of your "VPN comparison" cluster versus your "VPN how-to" cluster beats one blended number every time. And benchmark against your own prior periods, not against someone else's Domain Rating. Leading indicators like indexing, rankings and impressions move first; lagging ones like clicks, engagement, conversions and revenue move later — watch the leading signals for early direction but grade success on the lagging ones over a quarter.
Zero-click is now the norm — leading 2026 clickstream analysis finds that fewer than a third of US Google searches send a click to the open web — so raw impressions in isolation are increasingly a vanity number. AI Overviews depress click-through when present, with independent studies putting the drop on affected queries somewhere in a wide but clearly negative range; the exact magnitude varies by methodology, so treat it as directional consensus rather than a fixed constant. The mechanic to understand is that you can earn two impressions — the AI Overview and the classic result — and still lose the click, which is why impressions and clicks must be read together. Track AI Overview presence and citations with your rank tracker, since Search Console will not tell you which queries triggered one, and reframe the goal: ranking number one means less than it did, while qualified traffic and conversions matter more than raw volume. Being cited inside an AI answer is becoming its own objective. Building the depth that earns those citations is the point of topical authority.
Give it a full quarter. Rankings and impressions can move in weeks, but clicks and conversions typically take three to six months, longer on competitive terms. Judging in week two reads only the earliest leading signal and calls the game too soon.
Usually not a bug. It is the classic AI-Overview and zero-click signature: you are being shown more, sometimes twice, but users are clicking less. Verify with a rank tracker that flags AI Overview presence, and shift your scoring toward conversions.
There is not one, but if forced, non-branded organic conversions or revenue per visitor, segmented by cluster — it is the metric closest to money and hardest to fake. Pair it with non-branded clicks for context.
No — GA4 redefined it as the inverse of engagement rate, a different formula from the old one-pageview definition, so you cannot compare GA4 figures to the old analytics. Google now centers engagement rate, and "conversions" were renamed key events.
The SEO foundations affiliates need: how search works, what ranks, and how to build content that earns durable, high-intent organic traffic in the...
Core · 13 min readKeyword research is not about the biggest numbers — it is about demand you can capture and monetize.
Core · 13 min readOn-page SEO makes you eligible to rank, not guaranteed.
