Quick answer
Traffic share is the percentage of total traffic that one channel, page or website receives: a channel's sessions ÷ all sessions on your site, or your visits ÷ all visits in a market. A share only means something next to the numbers behind it.
Formula: channel traffic share = channel sessions ÷ total sessions
What traffic share means
Traffic share is the percentage of total traffic that one channel, page or website receives. The phrase carries two different meanings, and the difference between them is the denominator:
| Channel share (on-site) | Market share of traffic | |
|---|---|---|
| What it measures | One channel's slice of your own traffic | Your site's slice of a whole market or category |
| Formula | Channel sessions ÷ all sessions on your site | Your visits ÷ all visits in the market |
| Denominator | Your own sessions in the same period | Estimated visits to every site in the category |
| Data source | Your own analytics (e.g. TapCub, GA4) | Third-party estimates (e.g. Similarweb) |
| Question answered | Where does my traffic come from? | How visible am I against competitors? |
Channel-share example: if organic search brings 4,000 of your 10,000 sessions, its channel share is 40%. Market-share example: if your site gets 90,000 visits a month in a category estimated at 600,000 visits, your market share of traffic is 15%.
How the channels themselves are defined is a separate question — see source channel. This article is about computing and reading the share.
The thesis throughout: a share only means something in comparison. Read it next to the numbers behind it, and when a share moves, split the move into the channel's own growth and everyone else's decline.
The traffic share formula
Four formulas, one per denominator:
- Channel traffic share = channel sessions ÷ total sessions × 100. A session is a group of interactions within a time window; in GA4 a session ends after 30 minutes of inactivity (About Analytics sessions). Count sessions, not pageviews, unless you say otherwise.
- Market share of traffic = your visits ÷ total visits in the market × 100. Similarweb defines traffic share this way for websites, channels and keywords. The market total is an estimate, not a measured census.
- Page traffic share = page visits ÷ total site visits × 100. Semrush's Top Pages report uses exactly this: "Traffic Share reflects the percentage of the website's traffic driven to a particular page" (announcement, 2022-04-25).
- Search click share = clicks ÷ impressions × 100. This is the click-through rate in Search Console's performance report: of all the times your page appeared in search results, how often it was clicked.
Denominator rules, applying to all four:
- Same period. Numerator and denominator must cover the same window; do not compare a rolling 30-day share with a calendar-month share.
- Same counting basis. Sessions, visits and users are different units; both sides must use the same one.
- Humans only. Bot sessions in the denominator distort every channel's share at once (next section but one).
Sample data for formula 4: a page earns 480 clicks from 12,000 impressions in October (480 ÷ 12,000 = 4.0%) and 500 clicks from 10,000 impressions in November (500 ÷ 10,000 = 5.0%). Clicks rose, but the share rose mostly because the denominator shrank — the same trap as the channel example below.
Worked example: two months of channel shares
Sample data: a site's human sessions by channel, October versus November.
| Channel | Oct sessions | Oct share | Nov sessions | Nov share |
|---|---|---|---|---|
| Organic | 4,000 | 40% | 4,200 | 43.75% |
| Direct | 3,000 | 30% | 2,700 | 28.125% |
| Referral | 2,000 | 20% | 1,800 | 18.75% |
| Social | 1,000 | 10% | 900 | 9.375% |
| Total | 10,000 | 100% | 9,600 | 100% |
Each month's shares sum to 100%: 40 + 30 + 20 + 10, and 43.75 + 28.125 + 18.75 + 9.375.
Organic's share rose 3.75 percentage points (40% → 43.75%). Did organic grow? Only partly. Decompose the change in two steps:
- Own growth. Hold October's total fixed: 4,200 ÷ 10,000 = 42%. Organic added 200 sessions, worth +2.0 points (200 ÷ 10,000).
- Denominator shrink. Now use November's actual total: 4,200 ÷ 9,600 = 43.75%. The other three channels lost 600 sessions between them, and the smaller denominator adds another +1.75 points.
+2.0 + 1.75 = +3.75 points. Meanwhile total traffic fell 4% (10,000 → 9,600). A dashboard showing only "organic share up" would hide a shrinking site. Decision rule: never read a share alone — put the channel's actual sessions and the total beside it, and split any change into own growth and denominator effect.
How bots distort channel shares
Bots land somewhere too. A bot request with no referrer, if a tool counts it as a session, lands in direct traffic, so one channel's raw count absorbs them while every other channel's share is pushed down.
Sample data: November again, but 900 bot sessions arrive in direct. The raw total becomes 9,600 + 900 = 10,500.
| Channel | Human sessions | Human share | Raw sessions (with bots) | Raw share |
|---|---|---|---|---|
| Organic | 4,200 | 43.75% | 4,200 | 40% |
| Direct | 2,700 | 28.125% | 3,600 | 34.29% |
| Referral | 1,800 | 18.75% | 1,800 | 17.14% |
| Social | 900 | 9.375% | 900 | 8.57% |
| Total | 9,600 | 100% | 10,500 | 100% |
Direct's share jumps from about 28.13% to 34.29% (+6.16 points), while organic falls from 43.75% to 40%. With bots in the denominator, organic looks flat month-over-month: 40% in October, 40% in November. Humans-only, it rose 3.75 points. The bots did not just inflate direct — they erased a real organic gain.
If-then rule: if a channel's share jumps while its actual sessions stay flat, check the bot report before celebrating; if every other channel's share drops at once, suspect the denominator rather than four simultaneous declines.
Common mistakes when reading traffic share
| Mistake | What it hides | Fix |
|---|---|---|
| Reading share without actual numbers | November's total fell 4% while organic's share rose | Always show channel sessions and the total beside the share |
| Mixing counting bases | Per-arrival rows can exceed the session total; session-first-source rows sum to it | Pick one basis and label it; see how channels are decided |
| Comparing different periods or definitions | A 30-day share against a calendar-month share, or sessions against users | Same window, same metric, same filters on both sides |
| Treating estimated market share as measured | Third-party panels model category visits; small sites are noisy | Use market share for direction, not for exact point claims |
All four are the same mistake: reading a percentage as if it were a count. A share is a ratio; it moves when either side moves.
How to read traffic share in TapCub
Run the two-month example above in TapCub:
- Read shares next to counts. Web analytics › Sources lists sessions by channel, source and UTM, so organic's 4,200 of 9,600 sits beside its 43.75%. The default basis is First source of session: each session is credited once, so channel rows sum to the total. Every arrival also counts mid-session arrivals from external sites, so rows can add up to more than the total. Pick one basis before comparing October with November; how channels are decided covers the rules.
- Keep the 900 bot sessions out. In the default Detect and report separately mode, recognised bots are excluded from human visitors, sessions and pageviews, so the bot scenario above does not reach Sources: direct stays at 2,700 and organic's 3.75-point gain stays visible. The bots themselves appear on Web analytics › Bots.
- Split the move into its parts. In All-platform analytics › Operations › AI Assistant, open Why did it change on sessions and split by channel. It shows each channel's contribution to the total change versus the previous period, where contribution is that channel's change divided by the total change. In the example, the total fell by 400 sessions: organic +200, direct −300, referral −200, social −100. That is the own-growth-versus-everyone-else split the share alone cannot show.
- Read search click share separately. Monitoring › Search performance connects Google Search Console or Bing Webmaster and shows clicks, impressions and click-through rate by page, query and country. Data usually lags 1–3 days, so compare aligned windows.
What TapCub does not do: it measures your own traffic and provides no competitor or market traffic data. Market-level traffic share needs third-party estimates such as Similarweb's.
Sources
- Similarweb Knowledge Center, Traffic Share (full text verified via search index, 2026-11-15).
- Semrush News, Introducing Mobile Data and New Metrics Inside the Top Pages Report (2022-04-25).
- Google Analytics Help, About Analytics sessions.
- Google Search Console Help, Performance report (Search results).
- Similarweb blog, Traffic Share: Measuring Your Share of Success (Monique Ellis, 2023-06-14).
Was this explanation clear?
We use your answers to fix and expand this page.
Couldn’t send. Please try again in a minute.
Thanks for the feedback. We’ll use it to improve this page.You’ve already sent feedback on this page. Thanks.

Product Experience Specialist
Explores usability, user journeys, and feature adoption through practical examples of onboarding and everyday product use.