What Is Organic Traffic in Google Analytics?
Visitors who arrived from unpaid search results. It is the figure most businesses judge search work by. It is also systematically lower than what search actually delivered. Understanding why is worth more than any other explanation in this series.
What Counts As Organic Here
Somebody searched, saw your site in the unpaid results, clicked it and arrived. That is organic traffic. The important part of the definition is everything it leaves out, because those exclusions are where the misunderstandings start.
What is included. Unpaid results only.
Any search engine, not only Google, provided the click came from a result somebody did not pay for. That is the whole category.
What is excluded deliberately. Anything paid.
An advert click is recorded separately and should be. One is bought and the other is earned. A figure mixing them tells you nothing about either.
What is excluded by definition. Other routes.
Somebody arriving from another website, from social, from an email or from a document is not organic, however they originally heard of you.
What is excluded by accident. The important one.
A substantial number of visits that genuinely came from search are not recorded as organic, because the tool could not tell. Block two is that problem and it is the reason this page exists.
Why the category matters commercially. It is the earned half.
This is the traffic a business did not pay per click for, which makes it the figure everybody wants to see growing and the one most at risk of being read too simply.
It Is Systematically Understated
Consider what actually happens. Somebody searches for a plumber, finds you, reads a page and does not ring. Three days later they want the number, so they type your business name or open the site from their history. That second visit is recorded as direct. Search gets no credit for it whatsoever.
Why that matters so much. It is the normal pattern.
Very few people search, click and buy in one visit. Most consider, leave and come back, which means the visit that produced the work is frequently not the visit search was credited for.
Where the credit goes instead. Direct.
Which sounds like somebody typing your address deliberately and largely is not. Our traffic sources guide covers what that category actually holds.
The other routes to the same loss. Several.
Links opened inside applications, some email software, links from documents and certain transitions between sites all arrive without a usable source, so they land in the same place.
What makes this different from ordinary error. It runs one way.
These effects only ever move visits away from search and into an unattributed pile. Nothing in the mechanism moves anything the other way, which is what makes the bias systematic rather than random.
Who this costs. Whoever did the search work.
Which is the commercial point. The measuring performance material carries the full version of that argument.
Which Means The Number Is A Floor
Read the organic figure as a minimum rather than a count. Search delivered at least this much and certainly more. That is a genuinely useful way to hold a number even without knowing how much more.
What we can say confidently. The direction.
The error always runs the same way. That is knowable, it is structural and it does not depend on your particular site.
What nobody can tell you. The size.
How much is being lost depends on your customers, their devices, how long they take to decide and what software they use. We publish no figure for it. Anybody who quotes you one has invented it.
Why the refusal matters. It gets weaponised.
A confident percentage in a proposal makes the case for a service sound precise. The truthful version is less tidy and considerably more useful, since it survives contact with your actual data.
How to use a floor properly. For direction.
Compare it against itself over time. If the understatement is roughly stable, the trend remains meaningful even though the level is wrong.
When the floor moves for no reason. Check direct.
An organic fall alongside a direct rise of similar size is usually reclassification rather than a decline, which is worth ruling out before anybody investigates a problem that did not happen.
Branded And Non-Branded Are Both In Here
Somebody searching for your company name and somebody searching for the service you provide are completely different people commercially. This tool puts both in the same organic figure and cannot separate them, which hides the only distinction that matters for growth.
A branded search. Somebody who already knew.
They saw your van, took a recommendation, remembered an advert or used you before. Search delivered them. It did not win them.
A non-branded search. A stranger.
They typed the trade and perhaps a town and had no idea you existed. Every one of these is genuinely new demand your website earned.
Why mixing them flatters. Name searches perform.
People searching for you specifically arrive ready and convert well. A figure heavy with them looks excellent while telling you nothing about whether you are reaching anybody new.
The pattern this hides. Reputation without reach.
An established business with a good name can have healthy organic traffic that is almost entirely people who already knew about it. That is worth having and it is not growth.
Which tool can split them. Not this one.
Search reporting shows the actual queries, so the split can be made there. The performance report guide covers doing it.
Which Is Why You Need Both Tools
The two previous blocks describe the same limitation from different angles. Analytics knows what people did after arriving and is unreliable about how they got there. Search reporting knows exactly what they searched and nothing about what happened next.
What each answers here. Opposite halves.
Which searches showed your site and whether anybody chose you is a search reporting question. What those people then did on the site is an analytics question.
Why that split exists. Different vantage points.
One measures inside Google's results and the other measures on your own pages. Neither can see across the boundary, which is a limitation rather than a fault in either.
What using one alone produces. Half a conclusion.
A business looking only at analytics believes search delivers less than it does. A business looking only at search reporting has no idea whether any of it was any use.
Where the comparison is set out. Its own page.
Our comparison of the two tools covers the division properly, including why their figures will never reconcile. This page will not repeat it.
What to do practically. Connect them.
The two can be joined so some search data appears alongside site data, which is a convenience worth the five minutes it takes.
Judging It Properly
Three habits turn this figure from a number somebody quotes at you into something that supports a decision. None of them takes long and none requires anything to be configured.
By landing page, not in total. The first.
Which pages organic visitors arrive on tells you what search actually values about your site. A total tells you nothing about which work is paying.
Against a comparable period. The second.
Equal length periods, ideally the same window last year. Calendar months and seasonal boundaries both manufacture movements that never happened.
Alongside whether anything happened. The third.
Organic visits who enquired is a different figure from organic visits. It is the one worth reporting. That requires results to be configured, which is its own guide.
What the landing page view reveals. Surprises.
Frequently an article or an old page is bringing in more search visitors than the service pages the business promotes. That is strategic information rather than a curiosity.
What not to do with it. Report it alone.
A single organic figure at the top of a monthly summary is the format that causes most of the trouble in this subject. It carries none of the caveats on this page and it invites exactly the conclusion block seven warns against.
For local businesses specifically. The town pages.
Which location pages people actually arrive on is more informative than any geographic report. our local tracking guide covers why.
When A Rise Is Not Good News
Organic traffic going up is not automatically an achievement. Visitors from searches nobody buys from are worse than no visitors at all, because they inflate a figure everybody is watching while producing nothing.
How it happens innocently. A popular article.
Something written to attract attention succeeds, brings in readers with no intention of buying anything and lifts the traffic figure noticeably. The reporting improves and the business does not.
How it happens deliberately. Volume for its own sake.
Where an agency is measured on traffic, producing content aimed at easy searches is the rational response. It works, in the sense that the number goes up.
Why it is actively harmful. It buys time.
A rising figure postpones the question of whether any of this produces work. That question gets asked eventually and the answer is worse for the delay.
How to tell the difference. Look further down.
Did the extra visitors do anything. If organic rose while enquiries stayed flat, you have more of the wrong people rather than more of the right ones.
What we report instead. Results first.
Enquiries, then where they came from. Everything on this tool sits on the Google Analytics guide. Our measuring performance material owns the wider argument about traffic as an objective.
Search brought
you more than
your analytics
gives it credit for.
Somebody finds you through a search, comes back three days later by typing your name and that second visit is recorded as direct. It happens constantly, it only ever moves credit in one direction and it means the organic figure is a floor rather than a count.
How we read it properly:
If your organic traffic rose while your enquiries did not, we will say so rather than put the rise in a report and move on.
Every guide.
One tool.
Getting it installed and configured properly, reading the reports without being misled, measuring outcomes rather than activity, the advanced reporting and what to do when the numbers look wrong.
