The short answer
The average visit to a small business website holds about 76 seconds of active time, according to a recent Venbit Intelligence Research study of tracked local service websites. Search visits average 113 seconds and ad visits 63. Visits that became a lead had a median of 36 active seconds, so a short visit is not a failed one.
What to know
- On tracked local service websites, the average visit held about 76 seconds of active time.
- Search visits held the most attention at 113 active seconds, and 80% of them were engaged.
- Ad visits averaged 63 seconds, and only 47% of them stayed 10 seconds, opened a second page or became a lead.
- Half of the visits that became a lead held 36 active seconds or less, so a short visit can be a successful one.
- Time on site is three different numbers depending on the tool: wall-clock time, GA4 engagement time and active time.
- GA4's Average engagement time divides by active users and not by visits, so it reads higher whenever people come back.
The average visit to a small business website holds about 76 seconds of active time, a little over a minute. When Venbit Intelligence Research timed visits to tracked local service websites, it counted only the seconds a page was on screen and in focus. Search visits held 113 seconds on average and ad visits 63. The tool below runs both kinds of clock on your own visit to this page, then places your site's figure beside the research for the source most of your visits come from.
Time your own visit
Time your visit, then compare your site
Two clocks run on your visit to this page. Switch to another tab for a few seconds, come back, and see which one stopped. Then enter your own average and pick where most of your visits come from.
Time since you opened this page
0:00
Never stops
Active time
0:00
Starts with the page's script
- Timed from one page load to the next
- 0:00 on this page until you open another one
- Times the active clock stopped
- 0 so far. Switch tabs and come back to add one.
Seconds, or minutes and seconds: 75, 1:15 or 1m 15s
Average active seconds a visit
- Search1:53
- Ads1:03
- Direct1:16
- AI assistants1:11
- Social1:26
- Other sites1:27
All sources together: 1:16 a visit.
- Research average
- 1:16
- All sources
- You
- n/a
- Enter your average
- Engaged visits
- 56%
- 10s, a 2nd page or a lead
- Share of visits
- 100%
- Tracked websites
Venbit Intelligence Research, tracked local business websites, bots excluded. Active time counts only while the page is visible and its window has focus, and it starts when the page's script runs, as the tracker's does. A visit is engaged at 10 active seconds, a second page or a lead. The clocks run on this page, and nothing you type leaves your browser.
I wrote this guide because time on site is three different numbers depending on the tool that reports it, and most of the benchmarks that rank for it come from large or mixed sets of websites. The figures here come from small business websites tracked over three months. Every bot was removed before counting, and no website is named here or anywhere else.
Average time on site is 76 active seconds
That is a pooled average, so every visit counts once and the busier websites weigh more than the quiet ones. Ads brought 55% of the visits, and ad visits were the shortest of any source, which pulls the pooled figure toward theirs.
Average active seconds a visit, by source
- Search113s
- Links on other websites87s
- Social86s
- Direct76s
- AI assistants71s
- Ads63s
Search visits held 50 seconds more than ad visits. People who search have typed the problem in their own words and picked a page from a list, and more than a third of the ad visits came from Meta's apps, where the person was scrolling a feed a moment before the tap. Across all tracked websites, software, content and B2B sites included, search visits averaged 99 active seconds and 2.71 pages, so search visitors stay longer on a local service website than on the others.
Search visits stay longest and ad visits shortest
An average in seconds hides how many visits stayed at all. Venbit Intelligence counts a visit as engaged when it holds 10 active seconds, opens a second page, or sends a form or another lead, and Google Analytics 4 uses a similar rule built on the same 10 seconds. A visit that is not engaged is a bounce, and the split by source follows the clock with one exception.
| Source | Share of visits | Average active time | Engaged | Not engaged (bounce) |
|---|---|---|---|---|
| Search | 17.6% | 113 s (1:53) | 80% | 20% |
| Links on other websites | 1.6% | 87 s (1:27) | 70% | 30% |
| Social | 1.8% | 86 s (1:26) | 59% | 41% |
| Direct | 23.6% | 76 s (1:16) | 58% | 42% |
| AI assistants | 0.6% | 71 s (1:11) | 75% | 25% |
| Ads | 54.8% | 63 s (1:03) | 47% | 53% |
| All sources | 100% | 76 s (1:16) | 56% | 44% |
AI assistant visits are the exception. They passed the engaged line at nearly the rate of search visits, 75% against 80%, yet held 42 fewer seconds. My reading is that the assistant has described the business before the click, so the person arrives to check a fact or two and then acts, which is a good visit for a business to get. The row is small, under 1% of all visits, so read it with care.
Ads sit at the other end of the table. Fewer than half of ad visits, 47%, held 10 active seconds, opened a second page or became a lead, and the other 53% left before any of those happened. Ads & SEO in Venbit Intelligence flags the ads whose clicks leave within five seconds, and the Feed's ad post says what to change.
Send it to /pricing, where the price it advertised is the first thing on the screen.
Visits that became a lead were short
- 36 s
- median active time of the visits that became a lead
- About 2 min
- median length of those visits from the first event to the last
- 49%
- of those visits opened a single page
This was the surprise in the data for me. Half of the visits that turned into a lead held 36 active seconds or less, and a quarter of them lasted 20 seconds or less from the first event to the last. A visitor on a phone can land, tap the number in the header and be gone inside that time.
Treat time on site as a sign of interest, and judge the website by its leads.
The two clocks measure the same visits in different ways. Measured on the wall clock, the median lead visit lasted about 2 minutes, a figure that includes time the page sat open behind another tab or app. The active clock leaves that time out. Nearly half of these visits, 49%, opened one page, which the guide to pages per visit looks at in detail, and the conversion rate guide has the lead rates by source.
Time on site means three different numbers
Ask three analytics tools for time on site and you can get three answers for the same visit. None of them is broken. They start and stop the clock at different moments, and they divide by different things.
Wall-clock time
Time from the first thing a visit records to the last, background tabs included. Universal Analytics, Google's older version, timed a session from its first hit to the first hit on its last page, so a one-page visit with nothing else recorded measured zero.
Google Analytics 4 engagement time
Google defines user engagement as the time your web page was in focus. Its Average engagement time divides that total by active users, so it reads higher than a per-visit figure when people come back. Average engagement time per session is the per-visit version.
Venbit Intelligence active time
Seconds count only while the tab is visible and its window has focus, and they carry from page to page through the visit. Engaged time on the Analytics tab is the total divided by visits, the measure this page reports.
The one-page visit is where the methods split furthest apart. In the research, 61% of visits opened a single page, and a clock that runs from one page load to the next has nothing to measure on any of them. How far those one-page visitors read is in how far visitors scroll.
Bounce rate moves with the clock it is built on. A rule that called every one-page visit a bounce would put 61% of these visits in that bucket, while the engaged rule puts 44% there, because a one-page visit that holds 10 active seconds or ends in a lead counts as engaged. On the same visits, the choice of rule moves bounce rate by 17 points, so compare the rules before you compare two tools.
| How it is timed | What it records |
|---|---|
| Wall clock, first event to last | 4:10 |
| Universal Analytics session duration, no events set up | 0:00 |
| Google Analytics 4 user engagement, page in focus | 1:10 |
| Venbit Intelligence active time | 1:10 |
Measured side by side, the gap between tools is large. Plausible, an analytics company, ran its own tool beside GA4 on one recipe website for 28 days in 2025, and for the eight pages it published, GA4's engagement figure came out between 3% and 80% lower. That is one website, tested by the company whose number came out higher, so read it as proof that the method moves the figure by a wide margin.
GA4: Average engagement time
- Engaged seconds divided by active users
- A person who visits three times counts once
- Reads higher than a per-visit figure
GA4: Average engagement time per session
- Engaged seconds divided by sessions
- Each visit counts once
- The match for the 76 seconds on this page
When you take a number from GA4 to hold against this page, use the per-session metric. If people come back to your site two or three times before they get in touch, the per-user figure runs well above the per-session one for the same seconds, and the comparison flatters your site.
Read a low time on site by its source
A low average is a symptom, and the source of the visits says where to look. Split your figure by source first, then hold each line against the table above, because a site that lives on ads will post a low average with nothing wrong.
- 1
Check the page search visitors land on
At 113 active seconds, search visits stay longest of any source, so a search figure well under that points to a mismatch: the wrong page ranks for the search, or the page hides the service and the town below a large picture. Read each landing page as the searcher would, and use where your website ranks on Google to see which page Google shows.
- 2
Repeat the ad's offer on the first screen
Ad visits average 63 seconds, so hold ad traffic to that figure and not to the site average. Open the page each ad sends people to, on a phone, and make sure its first screen repeats the offer in the ad. In Venbit Intelligence, Ad performance shows time on site for your ad clicks and the share that left within five seconds, which gives you the place to start.
- 3
Count direct phone taps before judging direct
Direct covers typed addresses, bookmarks and links in texts and emails, and a share of those visitors already know you and came back for the phone number. A short direct visit that ends in a phone tap did its job. What direct traffic is explains what lands in that line.
- 4
Send social and link clicks to the matching page
People who click a post or an article expect the thing they read about, and those visits average 86 seconds from social and 87 from links on other websites. Point each link at the page about that thing, and ask a website that links to your home page to link the service page.
- 5
Put each source's time beside its leads
When time rises and leads do not, the page holds attention and loses people at the point of contact. The share of visitors who reach your contact page is the next number to read.
See engaged time for your own site
Venbit Intelligence puts Engaged time and Bounce rate in the row of figures under the Analytics chart, each with the period before under it, as the screen at the top of this page shows. Beside them sits a Typical visit card with the middle visit's engaged time, and both count the same active seconds as the research above.
The middle visit's engaged time: half of this period's visits spent longer on the site, half spent less. Engaged time counts only while the tab is visible and focused, so a tab left open in the background doesn't stretch it.
The Typical visit card answers a different question. Engaged time is the average, which a handful of long visits can lift, while Typical visit is the middle visit, with half the visits above it and half below. When the average climbs and the typical visit holds still, a few people stayed much longer, and the pages they read are worth a look.
One visit at a time, the trail shows where the seconds went. It lists each page the person opened with a clock beside it, what they scrolled and clicked on each one, and whether the visit became a lead, so a short visit that ended in a phone tap reads as the success it was. The Feed and Analytics page shows the rest of the tab, and the full list of features covers everything else.
How we measured this
Every figure on this page comes from Venbit Intelligence Research: visits to local service websites tracked in the Venbit Approved System, 21 August to 18 November 2026. The software, content and B2B websites in the wider set are left out, apart from the one search figure that covers every tracked website. Bots were identified and removed before anything was counted, and the bot traffic guide shows how many there are: they inflate visit counts and drag time down in tools that count them.
- 1
The clock starts with the page's script
When the tab is in front and its window has focus, active seconds start to count.
- 2
It stops when the tab loses view or focus
Switching tabs, minimising the window or clicking into another app stops the count, and coming back restarts it. A tab can be the selected one in a window behind another app and still report itself as visible, so the tracker checks focus as well.
- 3
The seconds carry across pages
Every page of the visit adds to one total. While a page stays open in view, the tracker sends a short signal after 15 seconds, again 30 seconds later, again 60 seconds after that and then every 2 minutes, so a long read on one page is counted.
- 4
Ten seconds marks an engaged visit
Ten active seconds, a second page, or a form or another lead marks the visit as engaged, and a visit that never gets there counts as a bounce.
The data has limits, and I would rather state them than have you find them. These are websites whose owners chose to track them closely, which makes them a well-measured sample of local service businesses but not a random one. The averages are pooled, so the websites with the most visits weigh the most, and a tab left in front on a computer whose owner has walked away keeps counting until another window takes focus or the machine sleeps.
See engaged time for your own site
Venbit Intelligence counts active seconds with the bots removed, sets each figure against the period before, and puts a clock beside every page of every visit. One line to install, free for 14 days, no card needed.
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 (you are here)
- 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?
- 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.
Sources
- Venbit Intelligence Research: visits to local service websites tracked in the Venbit Approved System. Bots excluded. Method and limits on the research page.
- Google Analytics Help, Analytics dimensions and metrics (Average engagement time, User engagement, Average engagement time per session)
- Google Analytics Help, [GA4] Engagement rate and bounce rate (engaged session)
- Google Analytics Help, User engagement (time with the web page in focus)
- Google Analytics Help, Session Duration, Avg (Universal Analytics)
- Plausible, GA4 could be underreporting average engagement time by up to 80%, as compared to Plausible (one website, eight pages, 28 days, 2025)