Does AI Drive More Direct Traffic? What Data Shows

Your Direct traffic is growing fast. Before you celebrate, check how much of it is real people and how much is just bots.

By Snehal Singh | Published: October 22, 2026

AI Overview Summary

Direct traffic is the group of visits your analytics tool cannot trace to a source. Many marketers hope some of it is hidden AI traffic from tools like ChatGPT. A study of 12 websites found no proof of that yet, because total Direct visits jumped by 140.9% this year while visits showing real engagement stayed flat, which suggests most of the growth came from bots. The study also found no reliable relationship between AI mentions and later Direct visits, so the better way to measure AI influence is to ask customers how they discovered you and to track your brand mentions in AI answers.

Table of Contents

• Why Everyone Thinks AI Traffic Is Hiding in Direct
• What One Company Found
• The Bot Problem Nobody Mentioned
• The Correlation Trap
• How to Check Your Own Direct Traffic
• A Better Way to Measure AI Impact
• Mistakes That Look Like Strategy
• How Ad2Connect Approaches AI Traffic Measurement
• Key Takeaways
• Frequently Asked Questions

Why Everyone Thinks AI Traffic Is Hiding in Direct

Imagine your traffic report shows a significant jump in Direct visits, and it would be a relief to discover that artificial intelligence was responsible for it. Many marketers believe exactly this, because the idea feels logical and some early data appears to support it.

In one study of AI Mode users, 18% of exits involved people leaving AI Mode and going directly to another website without clicking any visible link.

Similarweb research points in a similar direction, since people who received an AI brand recommendation often visited that brand within seven days. However, fewer of those visits appeared as Direct (19.9%) than for other visitors (38.8%), because most of them arrived through search.

So the idea is certainly possible, but the real question is whether we can actually see it in our analytics today.

What One Company Found

There is one real example. A growth expert named Eric Wu shared how the company Ramp traced its traffic.

His team looked at Direct visits to long blog page addresses, which nobody would realistically type by hand, so those visits were unlikely to be genuine Direct visits.

They removed anything that resembled an email campaign and compared the remaining visits with search visitors. Some of that Direct traffic behaved just like search traffic, because people stayed on the page, scrolled, and clicked in similar ways, and customers who were asked later confirmed that AI had sent them.

The result: Ramp tied 30–40% of its Direct traffic to answer engines.

That raised a fair question. Can other websites find the same signal?

The Bot Problem Nobody Mentioned

An SEO researcher tested this idea on the Google Analytics 4 accounts of 12 websites, comparing August with September 2026 and leaving out homepages, login pages, and pages with unknown sources.

At first, the news looked great, because eight of the 12 sites showed Direct growth and some grew by more than 100%. Then came the disappointing part: much of that traffic turned out to be automated bots.

On one site, Google Analytics counted 409,000 Direct visits, yet only 198 of them were engaged, which means real people hardly interacted with the pages at all.

Across the January to September sample, total Direct traffic rose 140.9%, but engaged Direct traffic changed by just ?1.0%, which means the real, human part of Direct traffic stayed almost flat, and when the same numbers are adjusted for the number of days in each month the growth is only +2.3%, so the huge jump in your Direct traffic report is mostly automated visits and not new customers arriving from AI answers.

The study also searched for clear AI footprints inside Direct traffic, and on average they made up only 0.003% of all Direct visits.

The finding is straightforward: hidden AI traffic could not be detected in the analytics data without customers telling the researchers directly.

The Correlation Trap

The researcher did not stop there, and he ran a second test on four brands to see whether a brand that appears more often in AI answers receives more branded searches or Direct visits a few weeks later.

Across 632 tests, the early results looked exciting. One site showed a strong link of 0.72 between AI share of voice and Direct traffic, and the link peaked about three weeks later, which fits a story where people see a brand in AI and visit afterwards. A second site showed a link of 0.47 between ChatGPT mentions and Direct visits.

But a strong link does not prove a cause, because two lines can rise together simply since both are trending upwards.

The Fair Test

To check this, he deliberately matched the AI data with the wrong weeks, because if the real timing mattered, the real data should beat these mismatched versions. It did beat most of them, but several wrong matches looked just as strong or even stronger.

The link also changed over time. For one site it appeared only in the second half of the year, while for the other it appeared only in the first half. The samples were small as well, with only 17 and 35 weeks of usable data.

So the answer is not a clear no, since an effect may exist, but it could not be proven.

How to Check Your Own Direct Traffic

You do not need expensive tools to begin, and a simple three-step plan is enough to get a clearer picture.

Step 1: Remove Bots

Split your Direct traffic by day, page, country, browser, device, and screen size, and look for sudden spikes from visitors with an identical setup and almost no engagement, which usually indicates a bot.

Step 2: Remove Pages That Naturally Get Direct Visits

Some pages always get Direct traffic. Leave these out of your test:

• Homepages and country root pages.
• Pricing and subscription pages.
• Login, sign-up, and account pages.
• Help pages, about pages, and privacy policies.
• Payment pages and order confirmations.

Step 3: Look at Long Blog Pages

Focus on long page addresses that nobody would type, because Direct visits there are more likely to be hidden referrals, and compare their behaviour with your search visitors.

A Better Way to Measure AI Impact

For years, marketing was measured with clicks, mostly because analytics tools made clicks easy to count. AI changes that path, since a person may see your brand in an answer, remember it, and visit later without any click being recorded.

So we need a wider view, and the most practical way is to combine three signals that cover each other’s blind spots:

• Where your brand appears. Track mentions across a fixed set of prompts, AI tools, and markets.
• How people found you. Ask new leads and customers where they first heard of you. Write down their exact words, including the AI tool they used.
• What happens next. Link those answers to qualified leads, sales, and revenue.

Put all three on one monthly report and read it with a few simple questions in mind. If your AI visibility rises but no customer mentions AI, your tracked prompts may not match how buyers actually search, and if more customers mention AI, check whether those customers go on to buy and stay.

Mistakes That Look Like Strategy

Many teams make the same errors when they chase hidden AI traffic:

• Celebrating total Direct growth without checking engaged sessions.
• Assuming all rising Direct traffic comes from AI.
• Trusting one strong correlation from one small data set.
• Testing many combinations and showing only the best result.
• Relying only on clicks to judge AI impact.
• Skipping the simple “How did you hear about us?” question.

How Ad2Connect Approaches AI Traffic Measurement

At Ad2Connect, a boutique digital marketing agency in Malad, we do not take a traffic number at face value, and we first clean the data by removing bot visits and pages that always receive Direct traffic before studying what real people actually do.

Our team works as a best SEO company in Mumbai for brands that want clear answers instead of vanity charts, so we track brand mentions in AI answers, add “how did you hear about us” questions to enquiry forms, and connect the responses to qualified leads and revenue.

The same thinking guides our paid campaigns, because as a performance marketing agency in Mumbai we check whether growing traffic actually turns into enquiries and sales, so that your advertising budget goes where it genuinely works.

Key Takeaways

• Direct growth is not AI growth. Most of the jump in Direct traffic was bots.
• Check engaged sessions. Real Direct traffic stayed almost flat.
• Be careful with correlations. A strong link can disappear under a fair test.
• Ask your customers. Self-reported answers are the clearest proof of AI influence today.
• Use three signals. Track brand mentions, customer answers, and revenue together.

Ad2Connect is a digital marketing agency helping brands grow through data-driven SEO, performance marketing, and content strategy. If your Direct traffic looks strange and you are unsure what is behind it, talk to our team about a traffic quality audit.