Click Through Rate Formula: A Practical Guide to Counting It Correctly

Quick answer: The click through rate formula is clicks divided by impressions, multiplied by 100. A page with 2,000 impressions and 50 clicks has a 2.5% CTR. The maths never goes wrong. What goes wrong is the inputs: how each platform counts an impression and a click, averaged averages, mixed brand searches and tiny samples.

Ask anyone who works in search for the click through rate formula and you will get the same answer in a second: clicks divided by impressions, times 100. That answer is correct, and it is also the least interesting part of the topic. In our experience, nearly every misleading CTR figure a Singapore business owner has been shown was calculated with perfect arithmetic. The error sat upstream, in what was counted as an impression, what was counted as a click, and how separate numbers were combined.

This piece is about the calculation itself, not about what a “good” CTR is for a given ranking position or how to write titles that earn more clicks. It walks through the formula with simple numbers, shows how Google Search Console, Google Ads, Google Business Profile and email platforms each define the two inputs, explains the averaging mistake that inflates reports, and shows how to set the whole thing up in a spreadsheet. If you want to see where measurement sits within wider search work, our SEO services overview explains the scope. The conclusion is simple: trust the formula, question the inputs.

The Click Through Rate Formula, With Worked Arithmetic

Click through rate (CTR) is the share of people who saw a link and then clicked it. Written out in full:

CTR = (clicks / impressions) x 100

An impression is one showing of your link, ad or email to someone. A click is one recorded tap or click on it. Dividing the second by the first gives a fraction; multiplying by 100 turns it into a percentage.

Some worked examples with plain numbers:

• If a page has 2,000 impressions and 50 clicks, the CTR is 50 / 2,000 = 0.025, which is 2.5%.
• If a listing has 800 impressions and 12 clicks, the CTR is 12 / 800 = 0.015, which is 1.5%.
• If a link has 40 impressions and 4 clicks, the CTR is 4 / 40 = 0.1, which is 10%.

You can also run the formula backwards. If you know a page earns about 1.5% CTR and it gets 10,000 impressions a month, you can expect roughly 150 clicks. If you want 300 clicks at the same CTR, you need about 20,000 impressions. That reverse calculation is useful for setting realistic expectations: it shows whether a traffic target depends on more visibility (impressions) or on a better listing (CTR).

Two small rules keep the arithmetic honest. First, clicks cannot exceed impressions on the same platform for the same item and period. If a report shows a CTR above 100%, the numerator and denominator came from different places, such as clicks from one tool and impressions from another. Second, a CTR of zero with zero impressions is not a CTR at all. It is a blank, and a spreadsheet should show it as blank rather than 0%, or it will drag averages down.

The third example above, 10% from 40 impressions, looks impressive. Hold that thought. It returns in the section on sample size, because a number can be calculated correctly and still mean very little.

What Each Platform Counts as an Impression and a Click

Every platform uses the same formula, but each one defines its inputs differently. That is why a CTR from one tool cannot be compared directly with a CTR from another. The table below summarises the main sources a Singapore SME is likely to see.

Google Ads CTR is defined by Google as the number of clicks your ad receives divided by the number of times your ad is shown. The structure matches organic CTR, but ads sit in fixed, labelled placements and invalid clicks are filtered out. Comparing an ad CTR with an organic CTR for the same keyword compares two different things.

Email CTR usually divides unique clicks by emails delivered. Click-to-open rate (CTOR) divides unique clicks by unique opens instead. Since Apple introduced Mail Privacy Protection in 2021, many opens are recorded automatically when Apple Mail preloads images, so CTOR has become less reliable. Delivered emails are a steadier denominator.

Google Business Profile reports views and interactions by its own rules, so treat them as a separate measurement. For businesses that depend on the map listing, our local SEO service page explains how those signals fit together.

How Google Search Console Counts Impressions, Clicks and Position

Google Search Console is Google’s free tool that reports how your site appears in Google Search. Its Performance report shows clicks, impressions, CTR (defined by Google as clicks divided by impressions) and average position. The input definitions are where the surprises are.

Impressions. Google says an impression means a user has seen, or potentially seen, a link to your site. In general, an impression is counted when your link appears in the current page of results, whether or not the user scrolled down to it, as long as they did not need to click to see more results. Inside independently scrolling widgets, such as a carousel, or expandable sections, such as “People also ask”, the item usually has to be scrolled into view or expanded to count. Scrolling away and back during the same search does not add a second impression.

Clicks. A click is counted when someone clicks a link that takes them to a page outside Google Search. Clicking a result, coming back and clicking the same link again counts once. Clicks that start a new Google search, such as tapping a suggested related query, are not counted for your site.

Position. Average position is the topmost position your site held in each search, averaged across searches where it had an impression.

Three details catch people out when calculating CTR from Search Console exports:

1. Property versus page grouping. When results are grouped by property, several links to your site in one result element can count as a single impression. Grouped by page, each page gets its own impression. That means page-level impressions can add up to more than the property total, and the CTRs will differ.
2. Anonymised queries. Google withholds some rare queries to protect searcher privacy. They are included in the chart totals but not listed in the query table, so the rows will not add up to the total.
3. Search type. The report defaults to web search. Image, video and news search types, and the separate Discover report, each have their own counting rules. Never add them together.

When we audited sites where the owner had built their own CTR reports, a common fault was summing query rows and dividing by the chart total: two inputs from different scopes. If the bigger problem is pages not appearing at all, that is an indexing question our technical SEO service covers.

The Averaging Mistake: Weighted vs Simple Averages

This is the single most common calculation error in CTR reporting, and it usually makes results look better than they are.

Say you have two pages. Page A has 10,000 impressions and 100 clicks, a CTR of 1%. Page B has 100 impressions and 10 clicks, a CTR of 10%. What is the combined CTR?

The tempting answer is to add the two percentages and halve them: (1% + 10%) / 2 = 5.5%. That is a simple average, and it is wrong for this purpose. It treats a page seen 100 times as equal to a page seen 10,000 times.

The correct answer is to recompute from the totals: total clicks divided by total impressions. That is 110 / 10,100 = about 1.09%. This is a weighted average, because each page counts in proportion to its impressions. The simple average overstated the real figure by roughly five times.

The same trap appears in several places:

• Monthly reports. Averaging twelve monthly CTRs to get an annual CTR gives quiet months the same weight as busy ones. Add up the year’s clicks and the year’s impressions, then divide.
• Keyword lists. Averaging the CTR column of a query export lets dozens of tiny queries with one click each outweigh the queries that actually drive traffic.
• Device or country splits. Averaging mobile and desktop CTRs ignores that one usually has far more impressions than the other.

Most agencies will tell you that average CTR across your keywords is a handy health score. That frequently backfires, because a simple average is dominated by low-volume queries, which tend to have extreme CTRs in both directions, and it can move sharply without any real change in traffic. Keep the totals, not the percentages. Store clicks and impressions in your reports, and calculate CTR fresh at every level you summarise. A percentage on its own cannot be combined correctly with another percentage; the raw counts can.

This also explains why your own spreadsheet sometimes disagrees with Search Console’s headline CTR. Google computes the headline figure from totals. If yours averages the rows, the two will rarely match.

Brand vs Non-Brand: Two CTRs Hiding in One Number

A branded search is one that includes your business name, such as someone typing your company name plus “opening hours”. A non-branded search describes what you sell without naming you, such as “aircon servicing Tampines” or “family lawyer Singapore”.

These two groups behave so differently that combining them produces a number that describes neither. People searching for your name are looking for you, so they click your result at high rates. People searching for a service are comparing options, scanning ads, map listings and competitors, so they click any single result far less often. When the two are blended, the brand searches prop up the total CTR and hide how the site performs with new customers.

The blend also shifts over time for reasons unrelated to search work. A newspaper mention, a busy festive season or a social media moment can lift branded searches for a few weeks. Total CTR rises, and someone credits it to a title tag change that had nothing to do with it. The reverse happens too: a successful push into new non-branded topics adds many low-CTR impressions, and total CTR falls even though the business is reaching more new customers than before.

The fix is to split the report before calculating anything. In Search Console, you can filter queries using a custom regular expression (a pattern-matching rule) that excludes your brand name and its common misspellings, then run the formula on each group separately. If your Search Console account shows a branded queries filter, that does the same job with less setup. Include short forms and Chinese or Malay versions of your name if customers use them.

In our experience, the non-branded CTR is the number that tells an owner whether search visibility is turning into new business. Branded CTR is mostly a reflection of how well-known you already are. For owners running a lean team, our small business SEO page explains how reporting is usually kept simple enough to read in a few minutes.

Small Samples and Statistical Noise in Plain English

Remember the link with 40 impressions and 4 clicks, a 10% CTR. If one more person had clicked, it would be 12.5%. If one fewer had clicked, 7.5%. A single person moves the figure by 2.5 percentage points. That is not a performance change; it is noise, the random variation you get whenever you count a small number of events.

The principle is easy to hold without any statistics training: the fewer impressions, the less a CTR means. A rough sense of the spread helps:

• With 100 impressions and 3 clicks, the true underlying CTR could plausibly sit anywhere from about 1% to about 8%. The 3% you calculated is a very loose estimate.
• With 1,000 impressions and 30 clicks, the plausible range narrows to roughly 2% to 4%.
• With 10,000 impressions and 300 clicks, it tightens to a fraction of a percentage point either side of 3%.

Those ranges come from standard margin-of-error reasoning, and the exact bounds depend on the method used, but the pattern is what matters. Ten times the impressions gives you a far more trustworthy CTR.

Practical rules that follow from this:

1. Do not judge a change on a few days of data. A title change on a page with a modest number of daily impressions needs weeks, not days, before the before-and-after comparison means anything.
2. Compare like periods. Weekday and weekend searching differ, and Singapore has strong seasonal patterns around Chinese New Year, the mid-year school holidays and year-end. Compare four weeks with four weeks, or a month with the same month last year.
3. Group small pages. Instead of judging dozens of low-traffic pages one by one, group them by type, such as all service pages or all location pages, and calculate CTR from the group’s combined clicks and impressions.
4. Be suspicious of extremes. The highest and lowest CTRs in any export almost always belong to the smallest rows.

When we review CTR claims in reports, the first thing we check is the impression count behind each percentage. A big percentage with a small denominator should be read as “not enough data yet”, not as a win.

Case Study: What a Restaurant’s Numbers Can and Cannot Tell You

Our restaurant case study documents a five-month engagement with a modern Asian sit-down restaurant (65 covers) in Tiong Bahru, Singapore. It does not report a click through rate, and we will not calculate one from its figures, because doing so would break the rules this piece is about.

The starting position. At baseline the restaurant had 410 monthly organic visitors, 4 keywords ranking on page 1, 1,200 Google Business Profile (GBP) views a month, 14 GBP reviews and 3 photos, a local pack position outside the top 10 for its main terms, a mobile Lighthouse score of 48/100 and approximately 2 monthly reservations via organic search.

What the programme did. The page sets out five phases: a GBP rebuild (months 1-2), a review velocity strategy (months 1-5), website technical SEO including menu and LocalBusiness schema (months 1-2), a local content strategy (months 2-5) and local citation building across 25 Singapore F&B directories (months 1-3).

The ending position at month 5. Monthly organic visitors reached 1,206 (+194%). Keywords ranking on page 1 went from 4 to 18. GBP views rose from 1,200 to 4,800 a month (+300%). Reviews grew from 14 to 78, the restaurant moved into the top 3 of the local pack, and monthly reservations from organic search went from approximately 2 to 23, which the page reports as +21 a month.

Why it matters for the formula. Notice what the page puts side by side: GBP views, counted by Google Business Profile, and website visitors, counted by a different system on the website. It would be easy to divide one by the other and call it a click through rate. That would mix a numerator and a denominator from two platforms with different counting rules, exactly the error the table above warns against. Notice also that the page reports reservations as an absolute change, +21 a month, rather than a percentage from a very small baseline, which is the honest way to present small numbers. The page also notes that menu schema lets diners see dish names and prices before clicking, so part of the decision now happens before any click is counted. The results came from the whole programme across all five phases, not from any one change. Owners in F&B can see how that work is scoped on our restaurant SEO page.

Setting Up the Click Through Rate Formula in Google Sheets

A simple spreadsheet removes most of the errors above, as long as it stores raw counts and calculates percentages last.

Step 1: Set up columns. Column A for the page or query, column B for impressions, column C for clicks. Paste these straight from a Search Console export, keeping the export’s grouping (by page or by query) consistent across the whole sheet.

Step 2: Add the CTR column. In D2, enter =IF(B2=0,"",C2/B2). This divides clicks by impressions and leaves the cell blank when there are no impressions. Copy it down the column. Then select column D and use Format, Number, Percent, so 0.025 displays as 2.50%. There is no need to multiply by 100 in the formula; the percent format does it for display.

Step 3: Add a correct total row. Below the data, put =SUM(B2:B200) for total impressions and =SUM(C2:C200) for total clicks, and for total CTR use =SUM(C2:C200)/SUM(B2:B200). Never use =AVERAGE(D2:D200) for the total; that is the simple-average trap.

Step 4: Add a minimum-impressions flag. In E2, enter =IF(B2<100,"low data","") so small rows are labelled before anyone draws conclusions. Choose the threshold to suit your traffic.

Step 5: Split brand and non-brand. Add a column marking each query as brand or non-brand, then use =SUMIF on clicks and impressions for each group and divide.

Related formulas worth keeping on the same sheet:

• Conversion rate = conversions / clicks (or sessions) x 100. If 150 clicks produce 6 enquiries, the conversion rate is 4%. As with CTR, check that both numbers come from the same source and period.
• Cost per click (CPC) = ad spend / clicks. If S$300 of spend produces 150 clicks, the CPC is S$2. This only applies to paid placements; organic clicks have no media cost, though they have content and site costs.
• Cost per enquiry = spend / enquiries, which is usually more useful to an owner than either of the above.

If you want an outside view of how your own reports are built, the Singapore SEO Agency team reviews measurement as part of looking at a site.

Field notes: In our restaurant case study, a modern Asian sit-down restaurant (65 covers) in Tiong Bahru, GBP views rose from 1,200 to 4,800 a month and monthly organic visitors went from 410 to 1,206 over five months, while monthly reservations from organic search went from approximately 2 to 23. The page reports these as separate measurements from separate systems and gives no click through rate. That is the right way to read them: each figure is valid within its own platform, and none of them should be divided by another to manufacture a rate the page never measured.

Our Take

The click through rate formula is trivial: clicks divided by impressions, times 100. Nobody gets the division wrong. The mistakes live in the inputs. Each platform defines an impression and a click in its own way, so CTRs from different tools cannot be compared or mixed. Averaging percentages instead of recomputing from total clicks and total impressions inflates results. Blending branded and non-branded searches hides how you perform with new customers. And small samples produce dramatic percentages that mean almost nothing. Store raw counts, calculate CTR last, split brand from non-brand and check the impression count behind every percentage. If you would like to see the kinds of measurable outcomes we report on, our case studies set out the figures client by client.

Frequently Asked Questions

What is the click through rate formula?

The click through rate formula is clicks divided by impressions, multiplied by 100 to give a percentage. For example, if a page has 2,000 impressions and 50 clicks, its CTR is 50 divided by 2,000, which is 0.025, or 2.5%. The same formula is used by Google Search Console, Google Ads and email platforms, but each one defines impressions and clicks differently, so the results are not directly comparable.

How does Google Search Console count an impression?

Google says an impression means a user has seen, or potentially seen, a link to your site. In general, an impression is counted when your link appears on the current page of results, whether or not it was scrolled into view, as long as the user did not have to click to see more results. Inside carousels and expandable sections, the item usually needs to be scrolled into view or expanded. Scrolling away and back during one search does not add another impression.

Why does my calculated CTR not match Search Console’s total CTR?

The usual reason is averaging. Search Console calculates its headline CTR from total clicks divided by total impressions. If you average the CTR column of an export, small rows get the same weight as large ones, and the answer drifts. Another reason is that anonymised queries are included in the chart totals but not listed in the query table, so the rows will not add up to the total. Grouping by page versus by property also changes impression counts.

What is the difference between a weighted and a simple average CTR?

A simple average adds up individual CTR percentages and divides by how many there are, treating every row equally. A weighted average divides total clicks by total impressions, so each row counts in proportion to its impressions. For CTR, the weighted approach is correct. If one page has 10,000 impressions at 1% and another has 100 impressions at 10%, the simple average says 5.5%, while the true combined CTR is about 1.09%.

What is the difference between email CTR and click-to-open rate?

Email CTR usually divides unique clicks by emails delivered, although some platforms use emails sent. Click-to-open rate divides unique clicks by unique opens, so it measures how persuasive the content was for people who opened. Since Apple’s Mail Privacy Protection began recording many opens automatically, open counts have become inflated, which makes click-to-open rate less reliable. CTR based on delivered emails is usually the steadier measure to track over time.

Is Google Ads CTR calculated the same way as organic CTR?

The formula is the same: clicks divided by the number of times the ad was shown. The inputs differ. Ads appear in fixed, labelled placements on the results page and across the Google Network, and Google filters out clicks it judges invalid. Because the placements and counting rules are different, comparing an ad CTR with an organic CTR for the same keyword does not tell you which performs better in any meaningful way.

How many impressions do I need before a CTR is reliable?

There is no single cut-off, but the principle is clear: the fewer impressions, the less the CTR means. With 100 impressions, one or two extra clicks swing the percentage noticeably. With 1,000 or more, the figure becomes much steadier, and with 10,000 it is quite stable. For small pages, group similar pages together and calculate CTR from their combined totals, and compare periods of at least several weeks rather than days.

Should I separate branded and non-branded searches when calculating CTR?

Yes. People searching for your business name are looking for you and click at high rates, while people searching for a service are comparing options and click any one result less often. Blending them makes total CTR look healthier than your performance with new customers really is. Filter branded queries out in Search Console using a regular expression or the branded queries filter if available, and calculate CTR for each group separately.

How do I calculate CTR in Google Sheets?

Put impressions in column B and clicks in column C. In D2, enter =IF(B2=0,””,C2/B2) and copy it down, then format column D as a percentage. For the total, use =SUM(C2:C200)/SUM(B2:B200), not an average of the CTR column. Adding a flag such as =IF(B2<100,”low data”,””) helps you avoid reading too much into small rows.

How is CTR related to conversion rate and cost per click?

They are separate steps in the same journey. CTR measures how many people who saw a link clicked it. Conversion rate measures how many of those clicks became an enquiry or sale, calculated as conversions divided by clicks or sessions. Cost per click divides ad spend by clicks and only applies to paid placements. All three use the same discipline: both inputs must come from the same source and period.