Quick answer
CPC (cost per click) is ad spend divided by clicks. Platforms count billed clicks, but your site sees fewer sessions (slow pages, blocked scripts, repeat clicks), so spend ÷ sessions is usually higher than the platform CPC, and that gap is the leak worth checking.
Formula: CPC = spend ÷ clicks (on-site CPC = spend ÷ sessions)
What is CPC?
In online advertising, CPC (cost per click) is what an advertiser pays, on average, for one click on an ad. Google Ads defines average CPC as "the average amount that you've been charged for a click on your ad", calculated by dividing the total cost of your clicks by the total number of clicks. Search ads are usually bought this way, and many social and display platforms report a CPC even when they charge per impression.
The catch is in the word "click". The platform counts clicks on its own servers, at the moment someone taps the ad. Your website counts sessions, which start only when the landing page loads and your analytics script runs. The two are never equal, and the number that matters for your budget is the one you can see on your own site. That is why it pays to calculate two CPCs, not one:
- Platform CPC = spend ÷ clicks the ad platform billed.
- On-site CPC = spend ÷ sessions your analytics recorded from that campaign.
On-site CPC is almost always the higher of the two. The gap between them is the share of paid clicks that never became a measurable visit, and it is worth investigating before you touch a bid.
The CPC formula
CPC = spend ÷ clicks
You will meet several related numbers, and they are easy to mix up:
| Term | Formula or meaning | Where you see it |
|---|---|---|
| Average CPC | Total cost of clicks ÷ clicks | Ad platform report |
| Max CPC | The most you are willing to pay for one click (your bid cap) | Bid settings |
| Actual CPC | What one click really cost after the auction | Ad platform, per click |
| On-site CPC | Spend ÷ sessions from the campaign | Your own analytics |
| CPC from CPM | CPM ÷ (1,000 × CTR) | Impression-bought channels |
Max CPC is a ceiling, not a price. Average CPC is what you pay; on-site CPC is what each visit you can actually measure cost you. The last row lets you compare a channel bought on CPM with one bought per click: a $4.00 CPM with a 0.67% click-through rate is about $0.60 per click. The CPC calculator and the CTR calculator do these conversions for you.
Platform CPC vs on-site CPC
| Platform CPC | On-site CPC | |
|---|---|---|
| Denominator | Billed clicks | Sessions that reached your site and loaded the analytics script |
| Counted by | The ad platform | Your analytics tool |
| Filtered out | Invalid clicks: Google Ads says "you won't be charged for invalid clicks or impressions" | Visitors who left before the page loaded, blocked scripts, bot sessions (if your tool separates them) |
| Several clicks in one visit | Each click counts | Usually one session |
| Use it for | Comparing bids and keywords inside the platform | Comparing channels and judging what a real visit costs |
Neither number is "wrong". They measure two different steps of the same journey, and the difference between them tells you how much paid traffic leaks out between the ad and the landing page.
Worked example: 1,000 clicks, 870 sessions
Sample data: the paid search campaign from the CPM example. One month, $800 spend, 1,000 billed clicks according to the ad platform, 870 sessions tagged with that campaign in the store's own analytics.
| Calculation | Result | |
|---|---|---|
| Platform CPC | $800 ÷ 1,000 clicks | $0.80 |
| On-site CPC | $800 ÷ 870 sessions | $0.92 |
| Clicks without a session | 1,000 − 870 | 130 (13%) |
| Premium paid per real visit | $0.92 ÷ $0.80 − 1 | 15% |
The paid social channel in the same dataset leaks more: $2,400 for 4,000 clicks is $0.60 platform CPC, but 3,320 sessions make on-site CPC $0.72, with 680 clicks (17%) missing and a 20% premium per visit.
Where did the 130 search clicks go? A sample breakdown of the kind of audit worth running:
| Cause | Sample count | How to check |
|---|---|---|
| Left before the landing page and script loaded | 55 | Landing-page load time on mobile; compare gap by device |
| Script blocked by a browser extension | 40 | Gap is larger on desktop browsers popular with blockers |
| Second click on the same ad within one visit | 25 | Clicks per session in the platform's own report |
| Redirect dropped the tracking parameters, or the URL was broken | 10 | Open the final URL of every ad; look for a 404 or a redirect that strips ? |
| Total | 130 | 55 + 40 + 25 + 10 |
Only the last two causes are outright mistakes you can fix in an afternoon, but the first is the most expensive: every visitor who gave up on a slow page was paid for. If the gap suddenly widens from 13% to 25% after a site release, look at the landing page before you look at the auction.
This is the core point: platform CPC tells you what the platform charged, on-site CPC tells you what a measurable visit cost, and the difference between them is a leak to investigate, not a rounding error.
What drives CPC
CPC is set in an auction, so the price you pay depends on who else wants the same click.
- Competition for the keyword or audience. More advertisers bidding on the same query push prices up. Brand terms are usually cheap; generic, high-intent terms are expensive.
- Ad quality. Google Ads shows a Quality Score from 1 to 10 per keyword, built from expected click-through rate, ad relevance and landing page experience. Google is explicit that Quality Score "is not an input in the ad auction"; it is a diagnostic that points at what to improve. Google's Ad Rank page still lists ad and landing-page quality among the auction factors and notes that higher quality ads can often lead to lower CPCs.
- Device, location and time. The same keyword can cost more on mobile, in big cities or during business hours.
- Bid strategy. Automated strategies that chase conversions will pay more per click when they expect a click to convert. A rising CPC alongside a falling CPA can be a good trade.
Common CPC mistakes
- Using platform CPC to compare channels. Each platform counts clicks its own way. Divide every channel's spend by sessions in one analytics tool before ranking them.
- Lowering bids to fix a high on-site CPC. If the gap between platform and on-site CPC is the problem, a lower bid buys the same leak more cheaply. Fix the landing page and tagging first.
- Ignoring bots. Automated clicks that slip past the platform's filters can create sessions that never buy. If your analytics counts them as people, on-site CPC looks better than it is and conversion rate looks worse.
- Losing the campaign tag. Without UTM parameters or a click ID on the landing URL, paid sessions fall into organic or direct, so on-site CPC looks terrible and organic search looks better than it is. See paid search.
- Optimising CPC instead of profit. A cheap click that never converts is expensive. Read CPC next to cost per acquisition and ROAS.
Finding the click gap in TapCub
The example above, platform $0.80 against site $0.92, is the comparison TapCub is built to show. TapCub does not buy ads, and its cost per click is always spend divided by sessions, so the platform CPC stays in your ad account; TapCub supplies the on-site side.
- Import the spend. In All-platform analytics › Business › Channel ROI, use the Costs tab to add spend per day, channel, source and campaign, one row at a time or as a CSV.
- Read on-site CPC. On the ROI tab, cost per click is spend divided by sessions, by channel, source or campaign. For the sample search campaign it shows $0.92. Put it next to the platform's average CPC for the same dates; the difference is your leak.
- Check the paid sessions are labelled. In Web analytics › Sources, visits carrying
gclid,msclkid,wbraid,gbraid,yclidorbd_vidare classed as paid search even without UTM parameters, and ad click IDs have their own table that stores only the type and a hash. See how TapCub tracks ad click IDs. - Keep bots out of the denominator. Bots and AI crawlers are counted separately in Web analytics › Bots, so automated traffic does not inflate the session count that cost per click is divided by.
Two limits to state plainly: some ad blockers block analytics scripts, including TapCub, and TapCub does not try to get around them (details); and TapCub does not adjust bids. It shows where clicks go missing so you can decide what to fix.
Sources
- Google Ads Help, Cost-per-click (CPC): Definition.
- Google Ads Help, Average cost-per-click (Avg. CPC): Definition.
- Google Ads Help, Invalid clicks: Definition.
- Google Ads Help, About Quality Score for Search campaigns.
- Google Ads Help, About Ad Rank.
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.

Customer Experience Consultant
Writes about live chat, visitor conversations, and self-service, sharing practical ways to make customer communication clearer and more effective.