Organic traffic is visits arriving from unpaid search results. It’s the headline number most SEO work is judged on, and on its own it’s a poor way to judge anything.
Where to measure it
| Source | Measures | Use it for |
|---|---|---|
| Search Console | Impressions, clicks, position, query | The truth about search performance |
| GA4 or similar | Sessions, behaviour, conversions after the click | What happened once they arrived |
| Rank trackers | Positions for chosen keywords | A sample, not a census |
| Third-party estimates | Modelled traffic (Ahrefs, Semrush) | Competitors, where you have no other option |
Search Console is the source of truth for search itself, because it’s Google’s own data. Analytics measures what happened after the click. The two will never match exactly, and expecting them to has wasted a great many meetings. Never present a third-party estimate and a first-party measurement as though they’re the same kind of number.
Segment it or it tells you nothing
A single organic traffic line hides everything that matters. The splits worth having, roughly in order of usefulness:
- Branded versus non-branded. The most important split by a distance. Branded traffic largely reflects marketing you did elsewhere, so counting it as an SEO win flatters the channel.
- By page type. Guides, product pages and the glossary behave differently and fail differently.
- By topic cluster. Shows whether topical authority is building anywhere in particular.
- By intent. Informational traffic and commercial traffic are worth wildly different amounts.
- By device. Effects that cancel out in aggregate can be large on one device, as SearchPilot’s split tests keep demonstrating.
The context most reports omit
Most pages get nothing
Ahrefs’ study of around 14 billion pages found 96.55% get no organic traffic from Google at all. The figure is quoted constantly and deserves its caveats, since the index it draws on is not a random sample of the web, which is why our library grades it weak.
Clicks are getting harder for the same rankings
Pew’s research found that when an AI summary appeared, users clicked a traditional result on 8% of visits against 15% when none appeared. Stable rankings with falling clicks is now a normal pattern rather than evidence you did something wrong.
Which is why reporting clicks without impressions and average position is close to useless in 2026. If impressions held and clicks fell, the results page changed shape around you and no amount of content work addresses that directly.
Common mistakes
- Reporting total sessions with no segmentation. Hides every useful signal.
- Counting branded traffic as an SEO win when a campaign elsewhere caused it.
- Ignoring seasonality and comparing months that were never comparable.
- Comparing part-periods. Month-to-date against a full month is not a comparison.
- Treating traffic as the goal. Traffic that never converts is a cost centre with good graphs.
Related
AI Overviews and SEO tracks what’s happening to clicks. SERP covers why position means less than it did.