Direct traffic in analytics: where it really comes from

What the label actually means

The name suggests a loyal customer typing your web address from memory. Sometimes that’s exactly what happened. But the label isn’t based on what the visitor did. It’s based on what your analytics tool didn’t receive.

Every visit is sorted using two clues. The first is any campaign tag in the address, such as a utm_source value. The second is the referrer, the address of the page the visitor clicked from. If neither clue is present, the tool has nothing to go on, and the visit lands in the direct bucket by default. In Google Analytics 4, for example, these visits show up in the Direct channel with the source “(direct)” and the medium “(none)”.

So “direct” is really a “we don’t know” pile. Some of it is genuinely direct. A lot of it is visits whose real source was lost on the way.

The visits you’d expect to see there

These are the honest ones:

• Someone types your address into the browser bar, or lets it autocomplete
• A returning customer uses a bookmark or a pinned tab
• A person heard your name on the radio, at a market stall or from a friend, and typed it in
• You, your staff or your developer opening the site to check orders or test a change (unless you exclude logged-in users)

For an established business with regular customers, this group can be sizeable. A yoga studio’s members who check the timetable every week will often go straight there.

The visits that end up there by accident

This is where most of the over-counting comes from. A click carries no referrer when it starts somewhere that isn’t a web page, or when something removes the referrer on purpose. The usual accidental sources, many of which would have been referral traffic if the referrer had survived:

• Email. Links clicked in desktop mail apps and many mobile mail apps usually arrive with no referrer. Your newsletter is one of the biggest culprits.
• Messaging apps. A link shared in WhatsApp, Signal, iMessage or a desktop chat app rarely says where it came from. People sometimes call this “dark social”.
• Documents and files. Links inside PDFs, slide decks, spreadsheets and Word files.
• QR codes. A phone camera opening a link has no page to report.
• Secure to insecure. When a page on HTTPS links to a page on plain HTTP, browsers drop the referrer. If your site, or a page linking to you, still runs without HTTPS, you’ll lose referrals here.
• Links marked “noreferrer”. Some sites deliberately strip the referrer from outgoing links.
• Privacy tools. Some browsers, extensions and settings remove or trim referrers.
• Redirects and shorteners. A badly set up redirect or link shortener can lose the referrer and any tags along the way.
• Bots. Automated traffic often sends no referrer, so unfiltered bot visits inflate the direct count.

Why the direct count runs high

Here’s a made-up week for the yoga studio: 1,200 visits, 480 of them labelled direct. That looks like a strong, loyal following.

Now look at the detail. On Monday at 7am there’s a spike of 210 visits, all landing on the new spring timetable page. Monday at 7am is when the studio’s email newsletter goes out. Those 210 people didn’t type the timetable address from memory. They clicked a newsletter link in their mail app, which sent no referrer.

There’s a simple test you can run on your own reports. Look at which pages direct visits land on. People type a home page. They don’t type yourstudio.co.uk/classes/thursday-yin-yoga-beginners/. When long, deep URLs get lots of direct visits, those visits almost certainly came from a link somewhere, and you’ve lost the record of where.

This matters because direct visits get credit nobody earned. If the newsletter brought 14 bookings but they all show up as direct, you might decide the newsletter isn’t worth the effort, while direct looks brilliant. You’d be making a decision on a mislabelled number.

How UTM parameters reduce direct traffic

You can’t control what other people’s apps do with referrers. You can control the links you share yourself, and that covers most of the accidental direct pile.

UTM parameters are small tags added to the end of a link. They travel in the address itself, so they survive even when the referrer is stripped. A newsletter link might look like this:

yourstudio.co.uk/timetable/?utm_source=newsletter&utm_medium=email&utm_campaign=spring-timetable

When someone clicks it, your analytics reads the tags and files the visit under email, newsletter, spring timetable, instead of direct. Tag these first:

1. Every link in your email newsletters and automated emails
2. Links you post in WhatsApp groups or broadcast lists
3. QR codes on flyers, posters, packaging and receipts
4. Links in PDFs, guides and downloadable worksheets
5. Your bio links on social profiles, since in-app browsers can lose the referrer

Two rules keep the tags useful. Never add UTM tags to links between your own pages, because that overwrites the visitor’s real source mid-visit. And pick one naming style and stick to it, so “Newsletter” and “newsletter” don’t show up as two different sources.

Analyzine tracks UTM campaigns and ties orders, enrolments and bookings back to the source, medium and campaign that brought them, so a tagged newsletter link gets its own line in the revenue report instead of disappearing into direct.

Other ways to shrink the unknown pile

UTM tags do most of the work, but a few other fixes help:

• Run the whole site on HTTPS, and redirect old HTTP addresses to it, so secure sites linking to you keep their referrer.
• Check your redirects. If you’ve changed permalinks or moved domains, make sure redirects keep the query string, or your UTM tags will vanish.
• Exclude yourself. Your own visits to check orders shouldn’t count. Most WordPress analytics tools let you skip logged-in administrators.
• Filter bots. A caching or security plugin that blocks obvious bots keeps junk out of every report, not only the direct one.

Organic vs direct, and does it affect SEO?

Neither, really. Organic means visits from unpaid search results, and those normally arrive with the search engine as their referrer, so they’re counted as organic traffic. Direct is the opposite: no referrer at all.

As for SEO, Google has said it doesn’t use your analytics data for rankings, so the direct number in your reports doesn’t move your position. What does matter is the real behaviour hiding inside it. A rising count of people typing your name is a sign your brand is getting known, and that tends to show up in branded searches too.

Accept that some direct will always remain. Your aim isn’t zero. It’s a direct share small enough that the rest of your sources tell a believable story. Start with next week’s newsletter: tag every link, send it, and compare Monday’s direct count with the week before.

Frequently asked questions