SEO Metrics: What Each Number Can and Cannot Tell You, and Why Average Position Misleads
· Royking Niba
SEO metrics are the numbers that tell you whether search is sending the right people to your site and whether they do anything once they arrive. The core four come straight from Google Search Console: clicks, impressions, click-through rate and average position. Around them sit outcome metrics from your analytics (sessions, leads, sales) and health metrics (indexing, page experience, links). This guide quotes Google’s own definitions, shows a decision table of what each metric can and cannot tell you, and works through a model of how average position can get worse while traffic grows, which is the single most common misreading in SEO reports.
The four Search Console metrics, in Google’s words
Every SEO metric that matters starts with how Google counts. The Search Console help page “What are impressions, position, and clicks?” carries no last-updated date, so I checked it on 8 October 2026. These are its definitions, quoted exactly:
- Impression: “An impression means that a user has seen (or potentially seen) a link to your site in Search, Discover, or News.”
- Click: “any click that sends the user to a page outside of Google Search, Discover, or News is counted as a click”.
- CTR: “The calculation of (clicks ÷ impressions).”
- Average position: “the topmost position occupied by a link to your property or page in search results, averaged across all queries”.
Two details in that page change how you should read the numbers. First, position only counts the best result you had: “Query 1 returns your property at positions 2, 4, and 6: its position is counted as 2 (the topmost position).” Second, average position is weighted by impressions, not by queries: “In Search Console, the metric is shown as average position, which averages the position value for all impressions”. A query that shows you 30,000 times moves the average thirty times harder than a query that shows you 1,000 times. Hold on to that sentence, because the worked model below depends on it.
The chart totals can sometimes differ from the table totals.
Google Search Console Help, Performance report
The Performance report help page explains why: queries, countries, devices and dates are “aggregated by property”, while pages and search appearance are aggregated by page. If you add up the rows of the Pages tab and get a different total from the chart, nothing is broken. Report the chart total and say which view each table came from. The same page warns that “The newest data can be preliminary, meaning it’s still being collected and might change in the next few hours”, so never close a weekly report on the last day’s figures.
A decision table: what each SEO metric can and cannot tell you
This table is my own working framework, not a Google document. The left columns say where the number comes from. The right columns are the important ones: the question it answers well, and the question people wrongly ask it.
| Metric | Source | Answers well | Cannot tell you | Report it |
|---|---|---|---|---|
| Clicks | Search Console | How many visits Google Search sent | Whether those visitors were the right people | Weekly and monthly, always |
| Impressions | Search Console | How often you were shown, which reveals new rankings early | Whether anyone saw the result (it counts “potentially seen”) | Monthly, next to clicks |
| CTR | Search Console | Whether a title and snippet win the click at a fixed position | Anything at site level, because it mixes positions | Per page or per query only |
| Average position | Search Console | Movement for one query or one small, stable query group | Overall progress, because new low rankings drag it down | Per query, never as a site headline |
| Organic sessions | Your analytics | On-site behaviour of search visitors | Search demand, and it will not match Search Console clicks | Monthly |
| Organic conversions | Your analytics or CRM | Whether search makes money | Which query produced the sale, in most setups | Monthly, as the headline |
| Indexed pages | Search Console page indexing report | Whether Google keeps the pages you want kept | Whether those pages rank | Monthly, and after any migration |
| Page experience | Search Console Core Web Vitals report | Real-user speed and stability at the 75th percentile | Rankings on its own, since Google says there is no single page experience signal | Quarterly, or when a template changes |
| Referring domains | A third-party link index | Rough trend of who links to you | What Google counts, since no tool sees Google’s link graph | Quarterly |
| Tracked rankings | A rank tracker | Daily movement on a fixed keyword list | Traffic, because one location and device is not your audience | Weekly, as a diagnostic |
The pattern in the right-hand column is deliberate. Clicks and conversions belong on the first line of any report. CTR and average position are diagnostic tools that only make sense when you hold the query set still, and they mislead when you average them across a whole site. If you want the deeper reason page experience sits so low, my post on page experience walks through the 75th-percentile arithmetic.
A worked model: average position gets worse while traffic grows
This is a model with invented, labelled numbers, built to show a mechanism. It is not a client result. The CTR figures are assumptions chosen to be plausible for a page-one and a page-three result; your real rates will differ, and the conclusion holds for any rates where page one beats page three.
- Month 1: the site ranks for one group of queries. They produce 10,000 impressions at an average position of 4. Assumed CTR 8 percent, so 800 clicks.
- Month 2: nothing changes for that group, still 10,000 impressions at position 4 and 800 clicks. A new article starts ranking at position 25 for a large group of long-tail queries: 30,000 impressions at an assumed 0.3 percent CTR, so 90 extra clicks.
| Metric | Month 1 | Month 2 | What a careless report says |
|---|---|---|---|
| Clicks | 800 | 890 | Up 11.25 percent |
| Impressions | 10,000 | 40,000 | Up 300 percent |
| Site CTR | 8.0% | 2.225% | “CTR collapsed” |
| Average position | 4.0 | 19.75 | “Rankings fell 15 places” |
The average position comes from Google’s impression weighting: (10,000 × 4 + 30,000 × 25) divided by 40,000 is 19.75. Not one ranking got worse. Every original query is exactly where it was, traffic is up, and the site has a new article climbing. Yet two of the four headline numbers look like a disaster. The fix is simple: segment before you average. Compare the same query group month on month, and report new rankings as their own line. A client or manager who sees “average position 4 to 19.75” without that segmentation will reasonably ask what went wrong, and the honest answer is that the metric was used for a job it cannot do.
The model has limits worth stating. Real query groups overlap, CTR varies with SERP features such as AI Overviews, and a new article can also cannibalise an old one, which really would push some positions down. That is why the segmented view matters: it lets you tell this harmless pattern apart from a real decline.
Reading a drop: which metric moved first
When traffic falls, the order in which the metrics moved is the best clue you have. Google’s guide to debugging search traffic drops, last updated 10 December 2025, lists the main causes as an algorithmic update or site-wide security or spam issue, seasonality, a technical issue or changing interests, and a reporting glitch. It recommends looking at a 16-month view because “This will help you analyze the traffic drop in context,” and suggests you “check them on Google Trends to understand if the drop was only for your website or throughout the web.”
| Pattern in Search Console | Most likely reading | First check |
|---|---|---|
| Impressions and clicks fall together, positions stable | Demand fell: seasonality or a fading trend | Same months last year, then Google Trends |
| Positions fall across many queries on the same dates | An algorithmic update or a site-wide quality issue | Google’s ranking update history, then the penalty recovery checklist |
| Clicks fall, impressions and positions stable | Lower CTR: a SERP feature, a worse title, or a competitor’s snippet | Search the query yourself and look at the page |
| One section’s pages lose impressions entirely | Indexing or crawling: noindex, robots.txt, redirects, server errors | Page indexing report and my crawl errors guide |
| A slow slide on older pages over many months | Content decay as fresher pages overtake them | Content decay triage |
| Every metric drops to near zero on one day | A reporting glitch, a lost property, or a manual action | Search Console messages and the Manual actions report |
A minimum SEO metrics set for a monthly report
- Organic conversions and their value, from your analytics or CRM. This is the line that justifies the work.
- Clicks from Search Console, month on month and year on year, so seasonality is visible.
- Clicks for a fixed query group that you agreed at the start, with its average position, so movement is measured on stable ground.
- New queries with impressions, reported separately, so growth does not hide inside a worse average.
- Top pages gaining and losing clicks, with one sentence of explanation each.
- Indexing and page experience status, as a pass or a named problem, not a wall of numbers.
Everything else is supporting evidence. If a metric would not change a decision this month, leave it out of the summary and keep it in the appendix. Choosing what to track starts before the report, at the keyword stage, which is why I keep a separate keyword research checklist for agreeing the fixed query group in the first place. For a whole-site health pass, the SEO audit checklist covers the technical side and the SEO content audit covers which pages deserve the work.
Related reading
- Google Penalty Recovery: how I diagnose and reverse a traffic collapse
- Keyword research checklist: twelve checks before you commit to a keyword
- SEO proposal: what to include, and the questions Google says to ask
- Content decay: finding the pages that are quietly losing traffic
- Page experience: what Google measures and what it does not
- SEO audit checklist
- Crawl errors and how to triage them
Leave a Reply