What Is Direct Traffic and Why Is It So High?

Tony CarnevaleNovember 21, 202612 min read
Traffic sourcesLast 30 days

347 visits

  • Ads22164% of traffic
  • Direct7020% of traffic
  • AI assistants267% of traffic
  • Social247% of traffic
  • Search51% of traffic
  • Other sites10.3% of traffic
Traffic sources on the demo site for the last 30 days: Direct is the second row, 70 of 347 visits, and it wears the site's own mark because nothing named a source.

The short answer

Direct traffic is the 32% of visits that reach the median small business website with no referrer, no ad click id and no tags, so analytics cannot name a source. Tracked websites without ads see 45%. Texts, email apps, QR codes, untagged Business Profile links and bots all land there, and tagging the links you hand out shrinks it.

What to know

  • Direct means the visit carried no referring page, no ad click id and no utm tags, so nothing named its source.
  • Across the tracked websites the median share arriving direct was 32%, and on about a third of them it was 40% or more.
  • Websites without ads lean on it hardest, with a median of 45% direct against 21% on websites that buy clicks.
  • Visitors arriving direct still became leads at 3.9 per 100 on the median website, second only to search at 4.8.
  • Bots swell unfiltered direct: 98% of visits from China and 96% from Singapore were bots, and they arrive with no referrer.
  • Tagging the Business Profile button, newsletters, texts and printed links moves those visits out of direct and names them on each lead.

Direct traffic is every visit that reaches your site with no referring page, no ad click id and no tags on the address, so your analytics has nothing to name it by. Typed addresses and bookmarks are in there, along with texts, email apps, QR codes, untagged listings and a share of bots. When Venbit Intelligence Research measured this over 90 days, the median tracked site received 32% of its visits this way. That is why the number looks high, and the link builder below is how you shrink it: tag a link for each place you hand one out, and the next visit carries its source.

Build a link

Tag a link so its visits stop landing in Direct

Paste the page the link should open, pick where people will see it and name the campaign. The verdict is the tracker's own rule applied to the link you built.

Where people will see the link

Left empty, it is gbp.

Your tagged link

https://yourbusiness.com/?utm_source=google&utm_medium=organic&utm_campaign=gbp

Without tags

Direct

Where 32% of visits land on the median tracked website

With this link

Search

Named gbp on every lead

Source
google
Medium
organic
Campaign
gbp
Line on each lead in the Inbox
gbp
Channel in Google Analytics
Organic Search

Untagged, a tap on the listing's website button arrives with no referrer or a bare Google one, and nothing on the visit says the listing sent it. Paste it into the Website field of your Business Profile. Its visits then count as search, which is what they are.

The filing comes from the rules Venbit Intelligence applies to every visit, which follow Google's published channel definitions. The 32% is from Venbit Intelligence Research, tracked websites, 90 days in autumn 2026, bots excluded. Nothing you type leaves your browser.

The rule behind that verdict fits in one line. Venbit Intelligence reads three clues on every visit, an ad click id, utm tags and the referring site, and files the visit as direct only when all three are missing. The rules are ported from the ones Google publishes for Google Analytics 4, which reports the leftover as (direct) / (none). Every analytics tool ends up with this bucket, and its size says more about your links than about your customers.

Direct is 32% of visits on the median site

What surprised me is where direct ranks against the named sources. Take each website's own mix and find the middle value for each source, and direct comes out at 32%, ahead of ads at 28% and search at 24%. Those three medians come from different websites, so they do not add up to 100. Read together, they say the biggest slice of traffic on the median small business site is the slice with no name on it.

Median share of visits arriving direct

  • No ads45%
  • All tracked websites32%
  • Run ads21%
Venbit Intelligence Research, tracked websites, bots excluded. A website counts as running ads when a fifth or more of its visits were paid clicks.

Advertising splits the set in two. On the websites that buy clicks, a median 21% of visits arrive direct, because ad clicks carry a click id and fill the report with named visits. The websites with no ads sit at 45%, level with search at 46%, so an owner who never advertises can expect close to half the traffic to arrive unnamed. Every source's share, website by website, is in where small business website traffic comes from.

The sources hiding inside direct traffic

Direct is a leftover bucket, so the useful question is what falls into it. Each source below reaches the site with nothing attached, for a reason that has to do with how the link was opened.

  • Typed addresses and bookmarks

    The slice owners picture when they hear the word. A person who types your address or opens a saved bookmark sends no referrer.

  • Links in texts and messaging apps

    A link tapped in a text or a chat app opens with nothing attached. Word of mouth passed on by phone lands here.

  • Most email apps

    Mail apps on a phone or a computer open links without a referring page. Webmail in a browser can pass one, and those visits file as email.

  • PDFs and documents

    Links inside a PDF quote, a Word file or a slide deck open from a document, so there is no referring page to pass on.

  • QR codes

    Phone cameras open the address with no referrer. An untagged code on a van, a yard sign or a flyer counts every scan as direct.

  • Apps that strip the referrer

    Plenty of apps open links in their own browser and send nothing about where the tap happened. When a link from Facebook or Instagram arrives with an fbclid, the tracker reads it as Social.

  • A secure page linking to an insecure one

    Browsers send no referrer from an https page to an http page. A site still served over http loses the referrer from most of the web.

  • An untagged Business Profile button

    The website button on a Google Business Profile arrives with a bare Google referrer or none at all. Tagged, those visits count as search.

  • Bots

    Scrapers load the page with no referrer. An analytics tool that lets them through counts them as direct visitors.

The person-to-person part of this has its own name among marketers, dark traffic or dark social: links passed along in texts, chat apps and emails, which reach your site with nothing attached. Nobody can measure it after the fact, because the visit holds no trace of the conversation that sent it.

Direct is the name for a visit whose source went missing on the way in.

You can test a link you already hand out with the link reader in how to track where website leads come from: paste it, say where the visitor came from, and it shows which clue decides the filing. The builder at the top of this page goes the other way and writes the link for you.

Find direct traffic in Google Analytics

In GA4, open Reports, then Acquisition, then Traffic acquisition, and read the Direct row of the session default channel group. Switch the first column to session source / medium and the same visits show as (direct) / (none). A jump in that row right after a change to your website points at a broken redirect or a lost tag before it points at your customers.

Google Analytics 4

  • Labels the bucket Direct, with (direct) / (none) as its source and medium
  • Files a text tagged utm_medium=sms under SMS and a print tag under Unassigned

Venbit Intelligence

  • Names the same visits Direct and gives them your site's own mark
  • Puts a tagged text or flyer under Other sites, with its campaign name on each lead
  • Sets aside direct visits from China, Singapore and Hong Kong until they act like a person

Direct visitors still turn into leads

I expected direct to sit near the bottom when I ranked the sources by lead rate. It sits second. For each website with at least 100 direct visits, I divided its leads from direct by those visits, and the median was 3.9 leads per 100. Only search did better, at 4.8 per 100.

Median leads per 100 visits, by source

  • Search4.8
  • Direct3.9
  • Ads3
  • AI assistants1.5
  • Social1.4
  • Other sites1.1
Venbit Intelligence Research, each website's own rate for each source it got 100 or more visits from, autumn 2026, bots excluded. A lead is a form sent, a phone number tapped or the goal that website set. The AI assistant, social and other-site rows rest on few websites, so read them as a comparison and not as an average.

That rate fits a good part of direct being people who already know the business, such as past customers coming back or someone who was sent the address in a text. Pool the visits of the local service websites in the set and direct still turned 3.6 of every 100 into a lead, against 3.8 for ads and 4.1 for search on the same websites.

SourceShare of visitsLeads per 100Left after one page, not engagedAverage active seconds
Ads54.8%3.853%63
Direct23.6%3.642%76
Search17.6%4.120%113
Social1.8%2.541%86
Other sites1.6%3.030%87
How visits from each source behaved on tracked local service websites, pooled. Venbit Intelligence Research, bots excluded.

The bounce column is where direct shows its other half. Direct visits left after one page without engaging 42% of the time, twice the rate for search at 20%. A bucket that holds both loyal customers and visits that open one page and vanish will look like that, and it is the reason a single direct figure on a dashboard tells you so little.

Ad clicks can arrive as direct

Paid clicks belong in this story because a quarter of them come in without a referrer. On the local service websites in the research, 26% of the visits that carried a Google Ads click id had no referring page. For Meta the share was 6%.

26%
of Google ad visits arrived with no referrer
6%
of Meta ad visits came in the same way
Venbit Intelligence Research, local service websites, visits identified by their ad click id.

Venbit Intelligence still names the platform on those visits, because the click id sits in the address and the tracker reads it before anything else. A gclid files the visit under Ads, and an fbclid files it under Social until the ad's address also carries utm_medium=paid. Analytics that ignores the click id files the same visits as direct, and so does a redirect on your own site that drops everything after the question mark.

If your ad platform reports more clicks than your analytics shows ad visits, why ad clicks do not match website visits works through the gap one cause at a time. Ads & SEO in Venbit Intelligence reads the click id the moment a visit lands and follows each campaign to the form or phone tap it brings.

Bots make direct traffic look bigger

A scraper loads your page with no referrer, so any bot an analytics tool fails to catch lands in direct. Across all the tracked websites, 39% of the visits the tracker recorded were bots, before any filtering. The clearest pattern shows up by country.

98%
of visits from China were bots, on nearly every tracked website
96%
of visits from Singapore were bots
14,768
visits set aside as Offshore scraper
Venbit Intelligence Research, tracked websites, 2026.

Those visits arrive direct, open one page and leave. Venbit Intelligence treats every direct visit from China, Singapore or Hong Kong as a suspected scraper and counts it as a person only once it does something a person does: a click, a tap on the phone number or email, a form or a second page. If Google Analytics shows a block of direct visits from those countries, read why your website gets traffic from China and Singapore before you read anything into your direct share.

WebsiteMaintenance.com
Last 30 days
344visits
Portland, OROn Pricing
Bothell, WAOn Website Repair
Everett, WAOn Care Plans
Seattle, WAOn Home
Bot trafficBots are recorded but hidden from every number by default.
Datacenter browserScript · Ashburn, VA
No user agentScript
GooglebotGoogle · Search engine
GPTBotOpenAI · AI crawler
AhrefsBotAhrefs · SEO tool
Visits reaching the demo site: each person joins the count, and each bot, from GPTBot to a datacenter browser, is named and kept out of every number.

What bots do to a visit count, and how to compare your host's number with your analytics, is covered in how much of my website traffic is bots.

Direct never reaches zero, and it should not: bookmarks and typed addresses belong there. What you can move is every link you hand out yourself, roughly in this order of effort.

  1. 1

    Tag the Google Business Profile website button

    Set the Website field to your address with ?utm_source=google&utm_medium=organic&utm_campaign=gbp on the end. Those visits then file as search, where they belong, and stop mixing with typed addresses.

  2. 2

    Label newsletters and email signatures

    Give every link in a send utm_medium=email and a campaign named for that send. Do the same for the website line in each signature the business uses.

  3. 3

    Put a tracking link on anything printed

    A QR code, a flyer, a van wrap or a yard sign gets its own short address. Tracking links for QR codes, flyers and posts shows how to make one and read its count.

  4. 4

    Tag the links you text to customers

    Quote follow-ups and appointment reminders are links you write yourself. Add utm_source=sms and a campaign name, and the leads they bring stop reading direct.

  5. 5

    Keep ad click ids on the address

    Check that the page an ad opens keeps everything after the question mark, and add utm_medium=paid to Meta ads so their clicks count as ads.

  6. 6

    Leave partner and directory links alone

    Another business's page sends its own address as the referrer, which names it. Tagging that link would overwrite a source that is already right.

Google Business Profile· Website link
https://websitemaintenance.com/
Referrer: None sent
Its visits are filed under
Direct
A Google Business Profile link that sends no referrer files its visits under Direct, and with the tags on the end the same visit files under Search.

Pick one spelling for each source and keep it. A report treats facebook, Facebook and FB as three sources, and a campaign called test2 tells you nothing in March. The link builder lowers the case and swaps spaces for hyphens for that reason.

See direct traffic in Venbit Intelligence

On the Analytics tab, Traffic sources lists Direct beside the named sources and gives it your own site's mark, since nothing else named those visits. The demo site at the top of this page had 70 direct visits out of 347 in its last 30 days, 20%, close to the 21% median for websites that advertise.

Each lead carries the same verdict. In the Inbox, a form lead's line reads Sent the form, then the town, then the source: the campaign name when the link had one, the referring site when there was one, and the word direct when there was neither. Open a direct lead and its source strip says Came straight to the site.

Inbox1ArchivedSpam 2Deleted
/Sent the form · Minster, OH · Website Maintenance Campaign 1
/checkupSent the form · Dallas, TX · Website Maintenance Campaign 1
/Sent the form · Crockett, TX · Website Maintenance Campaign 1
/Sent the form · Miami, FL · direct
/Sent the form · Belleville, IL · direct
The demo site's Inbox: each lead's line ends with its source, a campaign name for the Meta leads and the word direct for two of the others.

A direct lead is still a lead. The trouble comes with the next decision, because when half your leads read direct you cannot tell which flyer, listing or newsletter earned its keep.

For anything printed, Venbit Intelligence has Tracking links in the left bar. You give a link a name and the page it should open, and it hands back a short address such as vnbt.app/van to print or put behind a QR code. Each link counts its clicks, the visits that reached the page and the leads they became, and it names the clicks that came from link scanners so they do not pass for people.

The Tracking links screen on the demo site: the QR code on the van brought 32 clicks, 29 visits and 2 leads in 30 days, under its own name.

The short address adds the link's tags to every visit it sends, so the leads behind it stop reading direct too. Give each placement its own link: the van, the yard sign and the postcard each need one, or the count cannot tell them apart.

How we measured direct traffic

Every figure here comes from Venbit Intelligence Research: websites tracked in the Venbit Approved System, from 21 August to 18 November 2026, with bots identified and removed before anything was counted. A visit counts as direct when it carries no ad click id, no utm tags and no referrer from another site, the same test Google publishes for GA4's Direct channel. Leads are forms sent, phone numbers tapped or a goal the owner set; a phone tap is intent to call, since the tracker cannot see whether the call connected.

The shares of traffic are medians of each website's own share, so one large site cannot drag them. Lead rates by source are each website's own rate where it had 100 or more visits from that source. The pooled table and the ad click figures cover the local service websites, with software, content and business-to-business sites left out.

These are websites whose owners chose to track them closely, which already makes them more active online than the average shop, and about half of them advertise. Treat the figures as a well-measured sample of businesses like these and not as a census of every small business.

See what your direct traffic is hiding

Venbit Intelligence files every visit by the same rules as the link builder, keeps bots out of direct, and puts the source on every lead. One line to install, free for 14 days, no card needed.

More website analytics guides

Every question in this series, from Website Analytics for Small Business Owners.

Leads inbox4
Ad performance4
SEO tracking3
Visitor behaviour4
Feed and notifications3
Goals and measurement4
Privacy, AI and teams3
Venbit Intelligence Research19

Tony Carnevale

Founder, Venbit

Tony started Venbit in Mill Creek, WA in 2011 and still does the work: web design, SEO, and the website analytics product the company builds and runs for its own clients. The numbers and screens in his articles come from that product, not from a content mill.

Questions & Answers

  • In GA4, Direct is the channel for sessions whose source is (direct) and whose medium is (none) or (not set): no referring site, no campaign tags and no ad click id. It does not prove someone typed your address. On the median website Venbit Intelligence Research tracks, 32% of visits arrive this way.

  • Texts, email apps, PDFs, QR codes, apps that strip the referrer and untagged Business Profile buttons all land in direct, and unfiltered bots do too. Websites with no ads see the most: a median of 45% of visits on the tracked websites without ads, against 21% for websites that advertise. Tagging the links you hand out brings it down.

  • Neither on its own. Direct visitors became leads at a median of 3.9 per 100 on the tracked websites, second only to search at 4.8, so many of them already know the business. On the other side, 42% of direct visits left after one page without engaging, against 20% for search, because bots and stray taps sit in the same bucket.

  • Tag every link you control: the Business Profile website button with utm_source=google and utm_medium=organic, newsletters and signatures with utm_medium=email, texts with utm_source=sms, and anything printed with a tracking link. Keep ad click ids on your landing pages. Direct will not reach zero, since bookmarks and typed addresses belong there.

  • Dark traffic, or dark social, names visits from links shared in private: texts, chat apps and email. Those links open with no referrer, so analytics files them as direct, inside the 32% of visits that arrive direct on the median tracked website. Nothing afterwards can pull them back out; the fix is a tag on the link before you send it.

  • Yes. Scrapers load pages with no referrer, so an analytics tool that misses them counts them as direct. Across the tracked websites, 98% of visits from China and 96% from Singapore were bots, and 39% of all recorded visits were bots before filtering. A block of one-page direct visits from those countries is almost always automated.

  • On the tracked local websites, 26% of Google ad visits arrived with no referrer. They carried a Google click id, so Venbit Intelligence still filed them as ads, while analytics that ignores the click id, or a redirect that strips the query string, files them as direct. Meta ad visits arrived without a referrer 6% of the time.

More guides with a tool

See all 23 guides