Local websites · 90 days
Ads brought 55 of every 100 visits. Search visitors stayed to read the most.
Ads55%47%
- Direct24%58%
- Search18%80%
- Social1.8%59%
- Other sites1.6%70%
AI assistants0.6%75%
Visits by source, bots excluded. Email sent 14 visits and is left off.
The short answer
A small business website that does not advertise gets its visits from search and direct in almost equal parts: 46% and 45% on the median tracked website. The median website that runs ads gets 62% from ads. Social, other websites and AI assistants add a few percent. Search visits bring the most leads, 4.8 per 100.
What to know
- Without ads, the median tracked website gets 46% of its visits from search and 45% direct.
- With ads, paid clicks are 62% of the median website's visits, direct 21% and search 13%.
- Pooled across every visit to the tracked local websites, ads brought 55%, because the websites that advertise are the busy ones.
- Search turns 4.8 of every 100 visits into a lead on the median website, against 3.9 for direct and 3.0 for ads.
- Social media and links on other websites each brought under 2% of all visits, and email brought 14 visits in 90 days.
- AI assistants sent 1% of the median website's visits, and on the local websites 75% of those visitors stayed to read.
A small business website gets nearly all of its visits from three places: search engines, people arriving direct, and paid ads. The balance between them turns on one decision, whether you pay for clicks. Venbit Intelligence Research tracked small business websites for 90 days with every bot removed. The median website that does not advertise got 46% of its visits from search and 45% direct, and the median website that advertises got 62% from ads.
Check your mix
Compare your traffic sources with websites like yours
Enter last month's visits from each source and say whether you run ads. The tool draws your mix beside the median website like yours and works out what each source's visits would bring.
Visits last month, by source
In Google Analytics: Reports, Acquisition, Traffic acquisition. Leave a source empty if it sent none.
Enter your visits by source to draw your mix beside the median website. Until then, here is the median website that does not advertise.
Against the median website that does not advertise
Medians of tracked websites' own shares, so they do not add to 100. Social and other sites have no median yet: pooled is their share of all visits.
- Ads
- Median
- 0%
- Search
- Median
- 46%
- Direct
- Median
- 45%
- Social
- Pooled
- 1.8%
- Other sites
- Pooled
- 1.6%
- AI assistants
- Median
- 1.2%
Venbit Intelligence Research, tracked websites, bots excluded. A website runs ads when a fifth or more of its visits were paid clicks. Social and other sites show their pooled share of all visits on tracked local websites. Lead rates are the median website's among websites with 100 or more visits from that source. Nothing you type leaves your browser.
Every website in the research is tracked in the Venbit Approved System. Most are local businesses with a phone that rings, alongside a few software products. The figures cover three months from late summer into autumn, and no website is named here or anywhere else.
Websites without ads get visits from search and direct
Search means a click from a results page on Google, Bing or another engine, along with any link tagged as search. Everything that arrives with no label at all lands in direct: a typed address, a bookmark, a link in a text message. For a local business that also includes the taps on the website button of a Google Business Profile inside the Maps app, which arrive with no referrer.
I expected search to lead the websites without ads by a wide margin. It leads by one point. People hear a local company's name from a neighbour or see it on a van or a yard sign. Anyone who then types the web address, taps a link a friend texted or opens a bookmark lands in direct. The guide to direct traffic shows how much of that pile you can move to its real source by tagging the links you control.
Without ads, a small business website lives on search and direct in almost equal parts.
Websites that advertise get most visits from ads
The websites where paid clicks made up a fifth or more of visits look nothing like the rest. On the median one, ads brought 62% of visits, direct 21% and search 13%. Those websites also got far more visits, a median of 935 a month against 278 for the websites without ads, a gap the guide to visitors per month covers in full.
Tracked websites · autumn 2026
Where the median website's visits come from
- Ads
- Search
- Direct
- AI assistants
The median website's share of its own visits. Medians, so they do not add to 100.
| Runs ads | No ads | All tracked websites | |
|---|---|---|---|
| Ads | 62% | 0% | 28% |
| Search | 13% | 46% | 24% |
| Direct | 21% | 45% | 32% |
| AI assistants | 0.6% | 1.2% | 1.0% |
| Median visits a month | 935 | 278 | 609 |
| Share of visits on a phone | 69% | 47% | 64% |
Multiply the shares by the visit counts and the search visits come out close: 13% of 935 is about 122 a month, and 46% of 278 is about 128. That is rough arithmetic, a median share times a median count, and it proves nothing about any one website. It fits a simple reading of the table, though: the ads came on top of search and did not take its place.
The phone row moves with the ads. On the local websites, 90% of the visits that carried Meta's click id came from a phone, and 86% of those carrying Google's. A website that starts advertising gets a phone audience the same week, so test every page an ad opens on a phone before the first click is paid for.
Pooled, ads bring 55% of all visits
Add up every visit to the local business websites in the set, leaving out the software and content websites, and ads take a larger share still. Of 149,082 visits, 81,747 came from ads.
Share of all visits, local websites pooled
- Ads55%
- Direct24%
35,111 visits
- Search18%
26,270 visits
- Social1.8%
2,750 visits
- Other websites1.6%
2,311 visits
- AI assistants0.6%
879 visits
The pooled figure and the per-website median answer two different questions. A pooled total counts every visit once, so a busy website weighs more than a quiet one, and the busy websites here are the ones that advertise. Each website gets one vote in a median, which makes it the right number to hold your own website against.
Pooled: ads are 55% of visits
- Adds every visit on the local websites together
- A website with 935 visits a month weighs more than three times one with 278
- Answers where all the visits went
Per website: ads are 28% at the median
- Takes each website's own share, then the one in the middle
- Every website counts once
- Answers what is normal for a website like yours
What counts as each traffic source
Every analytics product sorts visits into sources with rules, and the rules decide the numbers. Venbit Intelligence follows the channel rules Google publishes for GA4, with a change for ad clicks: a visit carrying a click id that only ads issue, such as Google's gclid, counts as an ad even with no tag on it.
Search
Clicks from the results on Google, Bing or another search engine, plus any link tagged utm_medium=organic, such as a tagged Google Business Profile link.
Direct
No referrer and no tag: a typed address, a bookmark, a link in a text message, an untagged Business Profile tap from the Maps app.
Ads
Any visit carrying a click id that only ads issue, such as Google's gclid, or a paid tag like utm_medium=cpc. A Facebook or Instagram ad counts here when its link carries a paid tag.
Social
Posts, profile links and shares from a social site or app. A Facebook or Instagram ad link with no paid tag lands here too.
Other websites
Links on other sites, such as a directory, a supplier or a local news story. GA4 files them under Referral and Venbit Intelligence under Other sites.
AI assistants
ChatGPT, Gemini, Copilot, Perplexity, Claude and the other assistants, read from the referrer or the tag the assistant adds to its links.
Social, other websites and email stay small
- 1.8%
- of all visits came from social media
- 1.6%
- came from links on other websites
- 14
- visits came from email in 90 days
Social media brought 1.8% of the visits. At least nine in ten visits carrying Meta's click id are not in that count, because their ad links carried a paid tag and the visits landed under ads. Links on other websites added 2,311, and email added 14 visits in 90 days, too few to say anything about.
The small sources also turned fewer of their visits into leads. On the median website with at least 100 visits from each, social brought 1.4 leads per 100 visits and other websites 1.1, against 4.8 for search. A directory listing or a partner's link earns its place when it sends the right customer, but neither one will carry a local website.
AI assistants send about 1 visit in 100
AI assistants are the newest source and the smallest of the ones worth tracking. The median website got 1.0% of its visits from them: 0.6% for the advertisers, whose paid visits swamp everything else, and 1.2% for the websites without ads. About one in five tracked websites got 2% or more.
GPTBot, ClaudeBot and most other AI crawlers read pages without running scripts, so the pages read here are the ones loaded the way a browser does.
The visitors they send stay on the page. On the local websites, 75% of AI visitors stayed to read, second only to search at 80%. ChatGPT sent 2,435 of the AI visits across all the tracked websites, far ahead of Gemini at 82. The guide to ChatGPT traffic breaks the AI visits down by assistant and by month.
Search visits bring the most leads
Where visits come from is half the picture. The other half is what each source turns into, and on that count search wins by a clear margin.
Median leads per 100 visits, by source
- Search4.8
- Direct3.9
- Ads3
- Social1.4
- Other websites1.1
A lead is a form sent, a phone number tapped, or the goal the owner chose to count. Each figure is the median website's own rate, which compares a search visit and an ad visit on the same kind of website. Search wins because the visitor typed the problem into a box, and the results page sent them to a business that solves it.
| Source | Share of all visits | Stayed to read |
|---|---|---|
| Search | 18% | 80% |
| AI assistants | 0.6% | 75% |
| Other websites | 1.6% | 70% |
| Social | 1.8% | 59% |
| Direct | 24% | 58% |
| Ads | 55% | 47% |
Ads bring the most visits and the smallest share of visitors who stay. Fewer than half of ad visitors stayed to read, against four in five from search. A paid visit still became a lead 3.0 times per 100 on the median website, not far below the median of 3.4 per 100 for all visits, so the clicks are worth something. The weak point is the page the ad opens.
AI assistants are missing from the lead chart on purpose. Few tracked websites got 100 or more AI visits, and their median was 1.5 leads per 100, while the 879 AI visits pooled across the local websites brought 87 leads, 9.9 per 100. Two answers that far apart mean the sample is too small to settle it, and the ChatGPT against Google guide works through both.
Google and Meta send nearly all ad clicks
On the local websites, Google's ad click id came on 52,084 visits and Meta's on 29,473. Microsoft's came on 908. Why those counts sit below the clicks each ad account reports is the subject of why ad clicks do not match website visits.
The two big platforms send different visitors. Across all the tracked websites, the median website turned 3.1 of every 100 Google ad visits into a lead and 1.2 of every 100 Meta ad visits, counted on the websites with 100 or more such visits. A Meta ad reaches people scrolling a feed, and a Google search ad reaches people who asked for the service.
Check your own traffic sources
Your mix means something only next to websites like yours and only once the labels are right. Work through these in order.
- 1
Open the report that splits visits by source
In GA4 it is Reports, then Acquisition, then Traffic acquisition, read by session default channel group. Venbit Intelligence shows it as Traffic sources on the Analytics tab. Use a full month, and compare it with the same month last year if your trade is seasonal.
- 2
Tag every link you control
Add utm tags to the website link on your Google Business Profile, your email signature, your newsletters and any QR code. Those visits then stop landing in direct and show up under the source that earned them.
- 3
Compare with websites like yours
Put the numbers in the tool above and say whether you run ads. A website without ads that gets 12% of its visits from search has a different problem from an advertiser at 12%.
- 4
Judge each source by its leads
Visits say where people came from, and leads per 100 visits say which source pays. The guide to a good conversion rate has the numbers by kind of business.
- 5
Grow the source with the best rate first
On the median website that is search: a page for each service, the towns you serve named on it, and a complete Google Business Profile. Ads come next, once the page each ad opens works on a phone.
See your traffic sources in Venbit Intelligence
The Traffic sources list on the Analytics tab gives each source a row with its own mark, its visits, the change against the period before and its share of the traffic. When one source carried most of a rise or a drop, a line above the list names it. Open a row and you get its visits and leads, the share who stayed to read or left within seconds, its visits by day, the pages it lands on, the devices it came on and how its leads came in.
Search3,575▲ 2%42% of traffic
Ads1,787▲ 13%21% of traffic
Direct1,362▲ 3%16% of traffic
AI assistants681▲ 30%8% of traffic
Social680▲ 3%8% of trafficOther sites425▲ 6%5% of traffic
Google Ads1,173 clicks
Meta604 clicks
Ads get two more views of their own. The Campaigns section lists every campaign with its clicks and the leads it brought, and the Ad performance report flags ads whose visitors leave within five seconds. For AI assistants, the AI search section shows which assistant sent each visitor. Feed and Analytics and Ads and SEO show the screens in full.
How we measured traffic sources
Every figure here was counted by Venbit Intelligence, which installs with one line on a website and sets no cookies. Bots are identified and removed before anything is counted. Each visit gets one source by the rules above, and a website's share is its visits from that source divided by all its visits from 21 August to 18 November 2026.
Per-website figures are medians across the tracked websites, split at the line where paid clicks were a fifth or more of visits. Pooled figures add together the visits of the local service websites, leaving out the software, content and B2B websites. Lead rates are each website's own, counted only where a source sent it 100 or more visits.
See where your own visits come from
Venbit Intelligence files every visit under its source with the bots removed and ties each form and phone tap to the source behind it.
In Venbit Intelligence
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
- Average Website Visitors per Month for a Small Business
- What Is a Good Conversion Rate for a Small Business Website?
- What Time of Day Do Website Leads Come In?
- Which Day of the Week Brings the Most Website Leads?
- How to Tell If a Website Lead Is Spam
- How Much Website Traffic Comes From ChatGPT?
- Do ChatGPT Visitors Convert Better Than Google Visitors?
- Which AI Bots Read Your Website?
- Why Ad Clicks Don't Match Website Visits
- Mobile vs Desktop Website Traffic for Small Business
- How Long Do Website Visitors Take to Contact a Business?
- Average Time on Site for a Small Business Website
- Average Pages per Visit for a Small Business Website
- How Far Do Visitors Scroll on a Website?
- Why Is My Website Getting Traffic From China and Singapore?
- What Is Direct Traffic and Why Is It So High?
- Where Does Small Business Website Traffic Come From? (you are here)
- How Many Visitors Reach Your Contact Page?
- New vs Returning Visitors for a Small Business Website
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.