[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"glossary:\u002Fglossary\u002Fctr-curve":3,"glossary-related:\u002Fglossary\u002Fctr-curve":593},{"id":4,"title":5,"author":6,"body":7,"bottomLine":531,"cluster":532,"cta":533,"date":534,"definition":535,"description":536,"draft":537,"extension":538,"factChecker":539,"faq":540,"image":539,"meta":553,"meta_description":554,"meta_title":555,"navigation":556,"path":557,"related":558,"reviewer":539,"seo":560,"sitemap":561,"sources":562,"stem":583,"summary":584,"takeaways":585,"term":398,"tier":591,"updated":539,"__hash__":592},"glossary\u002Fglossary\u002Fctr-curve.md","What Is a CTR Curve? (Organic CTR by Position)","sebastian",{"type":8,"value":9,"toc":519},"minimark",[10,15,53,59,70,74,82,103,113,121,125,128,248,251,327,335,339,357,370,374,380,384,460,464,476,485,496,505],[11,12,14],"h2",{"id":13},"how-a-ctr-curve-is-built-from-search-console-data","How a CTR curve is built from Search Console data",[16,17,18,19,24,25,29,30,34,35,39,40,42,43,47,48,52],"p",{},"A CTR curve, or organic CTR by position, combines two metrics from the ",[20,21,23],"a",{"href":22},"\u002Fblog\u002Fanalyze-seo-performance-with-google-search-console","Performance report",". ",[20,26,28],{"href":27},"\u002Fglossary\u002Fsearch-console-ctr","CTR"," is clicks divided by impressions. ",[31,32],"cite-ref",{"n":33},"1"," ",[20,36,38],{"href":37},"\u002Fglossary\u002Fsearch-console-average-position","Average position"," is the topmost position a link to your site held in the results, averaged across all impressions. ",[31,41],{"n":33}," Plot one against the other for every query and the points fall along a downward slope: the higher the position, the larger the share of ",[20,44,46],{"href":45},"\u002Fglossary\u002Fsearch-console-impressions","impressions"," that become ",[20,49,51],{"href":50},"\u002Fglossary\u002Fsearch-console-clicks","clicks",".",[16,54,55,56],{},"The curve itself is the summary of that cloud. Group your queries into position buckets (1, 2, 3, 4, 5, then 6 to 10 and 11 to 20, where data gets thinner) and compute each bucket as total clicks divided by total impressions. Do not average the CTRs of the rows: a query with 20 impressions would weigh as much as one with 20,000. Data in the Queries tab is aggregated by property, so each query row has one position and one CTR for your whole site. ",[31,57],{"n":58},"2",[16,60,61,62,66,67],{},"If you work from the BigQuery bulk export instead of the report, position is stored as a zero-based sum. Google's table reference gives average position as ",[63,64,65],"code",{},"SUM(sum_top_position)\u002FSUM(impressions) + 1",", and advises to almost always aggregate your metrics. ",[31,68],{"n":69},"4",[11,71,73],{"id":72},"expected-ctr-reading-a-query-against-the-curve","Expected CTR: reading a query against the curve",[16,75,76,77,81],{},"The value of the curve at a query's position is that query's ",[78,79,80],"strong",{},"expected CTR",". Multiply it by the query's impressions and you get expected clicks; subtract actual clicks and you have the gap. That one comparison sorts a query into one of two problems:",[83,84,85,92],"ul",{},[86,87,88,91],"li",{},[78,89,90],{},"Below the curve at a good position",": the ranking is there, the snippet is not earning it. The fix is the title and meta description.",[86,93,94,97,98,102],{},[78,95,96],{},"On the curve at a weak position",": the snippet earns what that position allows. The fix is the ranking, which is the ",[20,99,101],{"href":100},"\u002Fblog\u002Fstriking-distance-keywords","striking distance"," case.",[16,104,105,106,109,110,112],{},"Google's own version of this analysis is a bubble chart of average position against CTR, split into quadrants by the average of each axis. ",[31,107],{"n":108},"5"," For queries at a top position with a low CTR, it lists three causes: competitors showing rich results from structured data, ranking for a query users do not connect with your site, and searchers who already found what they needed, such as opening hours or a phone number, on the results page. ",[31,111],{"n":108}," Only the first two are worth a rewrite, so look at the results page before you change a title.",[16,114,115,116,120],{},"The ",[20,117,119],{"href":118},"\u002Fseo-low-hanging-fruit-finder","Low-Hanging Fruit Finder"," scores keywords in positions 4 to 30 against your property's own CTR-by-position curve where there are enough top-ranking keywords, and falls back to published averages where data is thin.",[11,122,124],{"id":123},"example-one-sites-ctr-curve","Example: one site's CTR curve",[16,126,127],{},"The table below is constructed to illustrate the pattern, using non-branded queries only. Your curve will differ in level; the shape, steep at the top and flat on page two, is what to expect.",[129,130,131,150],"table",{},[132,133,134],"thead",{},[135,136,137,141,144,147],"tr",{},[138,139,140],"th",{},"Position bucket",[138,142,143],{},"Impressions",[138,145,146],{},"Clicks",[138,148,149],{},"CTR (expected CTR)",[151,152,153,167,180,194,207,220,234],"tbody",{},[135,154,155,158,161,164],{},[156,157,33],"td",{},[156,159,160],{},"4,000",[156,162,163],{},"1,120",[156,165,166],{},"28.0%",[135,168,169,171,174,177],{},[156,170,58],{},[156,172,173],{},"3,500",[156,175,176],{},"560",[156,178,179],{},"16.0%",[135,181,182,185,188,191],{},[156,183,184],{},"3",[156,186,187],{},"3,000",[156,189,190],{},"330",[156,192,193],{},"11.0%",[135,195,196,198,201,204],{},[156,197,69],{},[156,199,200],{},"2,800",[156,202,203],{},"224",[156,205,206],{},"8.0%",[135,208,209,211,214,217],{},[156,210,108],{},[156,212,213],{},"2,600",[156,215,216],{},"156",[156,218,219],{},"6.0%",[135,221,222,225,228,231],{},[156,223,224],{},"6–10",[156,226,227],{},"9,000",[156,229,230],{},"270",[156,232,233],{},"3.0%",[135,235,236,239,242,245],{},[156,237,238],{},"11–20",[156,240,241],{},"12,000",[156,243,244],{},"96",[156,246,247],{},"0.8%",[16,249,250],{},"Now read two queries against it:",[129,252,253,278],{},[132,254,255],{},[135,256,257,260,263,265,267,269,272,275],{},[138,258,259],{},"Query",[138,261,262],{},"Position",[138,264,143],{},[138,266,146],{},[138,268,28],{},[138,270,271],{},"Expected CTR",[138,273,274],{},"Expected clicks",[138,276,277],{},"Gap",[151,279,280,304],{},[135,281,282,285,288,291,294,296,298,301],{},[156,283,284],{},"Query A",[156,286,287],{},"3.2",[156,289,290],{},"1,500",[156,292,293],{},"45",[156,295,233],{},[156,297,193],{},[156,299,300],{},"165",[156,302,303],{},"120",[135,305,306,309,312,315,318,320,322,324],{},[156,307,308],{},"Query B",[156,310,311],{},"14.0",[156,313,314],{},"2,000",[156,316,317],{},"16",[156,319,247],{},[156,321,247],{},[156,323,317],{},[156,325,326],{},"0",[16,328,329,330,334],{},"Query A is the finding: position 3 should return about 11% on this site and it returns 3%. That is a snippet problem worth about 120 clicks a month without any ranking work. Query B earns exactly what position 14 allows, so a new title will not help; moving it to position 5 would, at 6% of 2,000 impressions, return about 120 clicks instead of 16. Both are ",[20,331,333],{"href":332},"\u002Fblog\u002Flow-hanging-fruit-seo","low-hanging fruit",", but they need opposite fixes.",[11,336,338],{"id":337},"how-to-build-your-own-ctr-curve","How to build your own CTR curve",[340,341,342,345,348,351,354],"ol",{},[86,343,344],{},"Open the Performance report, select the Web search type and a date range of three months or more.",[86,346,347],{},"Filter out branded queries, then export the Queries table.",[86,349,350],{},"Round each query's position into the buckets above.",[86,352,353],{},"For each bucket, sum clicks and impressions and divide.",[86,355,356],{},"Add an expected-clicks column per query (impressions × bucket CTR) and sort by the gap.",[16,358,359,360,364,365,369],{},"The full manual procedure, including regex filters for branded terms, is in ",[20,361,363],{"href":362},"\u002Fblog\u002Fhow-to-find-low-hanging-fruit-keywords","how to find low-hanging fruit keywords in Search Console",". If two of your pages split the impressions of one query, fix that first; ",[20,366,368],{"href":367},"\u002Fblog\u002Fwhat-is-keyword-cannibalization","keyword cannibalization"," puts both pages below the curve for reasons no title can solve.",[11,371,373],{"id":372},"the-thing-people-get-wrong","The thing people get wrong",[375,376,377],"blockquote",{},[16,378,379],{},"The common mistake is to take a published industry CTR table and treat the gap as lost clicks. Those tables average many sites and many result layouts. Your site's curve can sit well above or below them at every position, so a query that looks under-performing against an industry table may be exactly on your own curve. Build the curve from your data first; use a published table only where your data is too thin.",[11,381,383],{"id":382},"ctr-curve-vs-ctr-vs-average-position","CTR curve vs CTR vs average position",[129,385,386,399],{},[132,387,388],{},[135,389,390,392,394,396],{},[138,391],{},[138,393,28],{},[138,395,38],{},[138,397,398],{},"CTR curve",[151,400,401,415,428,447],{},[135,402,403,406,409,412],{},[156,404,405],{},"What it is",[156,407,408],{},"Clicks ÷ impressions for one row",[156,410,411],{},"Topmost position, averaged over impressions",[156,413,414],{},"CTR per position bucket across many queries",[135,416,417,420,423,425],{},[156,418,419],{},"Unit",[156,421,422],{},"One query, page or total",[156,424,422],{},[156,426,427],{},"The whole site (or a segment)",[135,429,430,433,436,439],{},[156,431,432],{},"Answers",[156,434,435],{},"How often this row is clicked",[156,437,438],{},"Where this row ranks",[156,440,441,442,446],{},"How often this row ",[443,444,445],"em",{},"should"," be clicked",[135,448,449,452,455,457],{},[156,450,451],{},"Useful alone",[156,453,454],{},"No",[156,456,454],{},[156,458,459],{},"Only as a benchmark for rows",[11,461,463],{"id":462},"where-the-curve-can-mislead","Where the curve can mislead",[16,465,466,469,470,473,474],{},[78,467,468],{},"Branded queries."," Google notes that branded queries typically lead to higher-ranking pages and higher click-through rates. ",[31,471],{"n":472},"6"," Mixed into the curve, they inflate the top positions and make every non-branded query look weak. The branded queries filter in Search Console splits the two automatically, though Google says some queries may be misidentified. ",[31,475],{"n":472},[16,477,478,481,482,484],{},[78,479,480],{},"SERP features share a position."," All links in an AI Overview are assigned the single position the AI Overview occupies, and every item in a carousel shares the carousel's position. ",[31,483],{"n":33}," A link inside an AI Overview at position 1 does not earn what a plain position-1 result earns, so queries with these features sit below the curve without anything being wrong with the page.",[16,486,487,490,491,495],{},[78,488,489],{},"Device mix."," The same position earns a different CTR on mobile and desktop, and query mix differs between them. Build separate curves when one device dominates; the ",[20,492,494],{"href":493},"\u002Fblog\u002Fmobile-desktop-ctr-difference-by-position","mobile and desktop CTR comparison"," shows how.",[16,497,498,501,502,504],{},[78,499,500],{},"Missing rows."," The Performance report omits rare queries to protect user privacy, and tables show at most 1,000 rows. ",[31,503],{"n":184}," A curve built from a report export is therefore missing part of your long tail, so treat thin buckets with care.",[16,506,507,510,511,513,514,518],{},[78,508,509],{},"Property vs page aggregation."," The Queries tab aggregates by property, the Pages tab by page, and chart totals can differ from table totals for that reason. ",[31,512],{"n":58}," Build the curve and the queries you compare against it from the same view. If you need to act on the result, the ",[20,515,517],{"href":516},"\u002Fblog\u002Fmerge-or-redirect-competing-seo-pages","merge-or-redirect decision"," covers pages that share a query.",{"title":520,"searchDepth":521,"depth":521,"links":522},"",3,[523,525,526,527,528,529,530],{"id":13,"depth":524,"text":14},2,{"id":72,"depth":524,"text":73},{"id":123,"depth":524,"text":124},{"id":337,"depth":524,"text":338},{"id":372,"depth":524,"text":373},{"id":382,"depth":524,"text":383},{"id":462,"depth":524,"text":463},"A CTR curve turns CTR from a number without context into an expectation: at this position, on this site, a query should earn this share of clicks. Build it from your own non-branded queries, as clicks over impressions per position bucket. Then the decision for each query is simple: below the curve at a good position means fix the title and snippet, on the curve at a weak position means work on the ranking.","quick-wins","lhf","2026-11-16","A CTR curve is the expected organic click-through rate at each ranking position: the share of impressions that turn into clicks at position 1, 2, 3 and so on. Built from your own Search Console data by grouping queries by average position, it gives each query an expected CTR to compare its actual CTR against.","How a CTR curve is built from Search Console data, how to read a query's expected CTR against it, and where the curve misleads.",false,"md",null,[541,544,547,550],{"q":542,"a":543},"What is expected CTR?","Expected CTR is the click-through rate a query should earn at its current position, read off your site's CTR curve. Multiply it by the query's impressions to get expected clicks; the difference to actual clicks is the gap a title or ranking change could close.",{"q":545,"a":546},"Should I use a published CTR curve or my own?","Use your own wherever you have enough data. Published curves average many sites, while your brand, niche and the SERP features on your queries change how many clicks each position earns. A published table is a fallback for positions where your own data is thin.",{"q":548,"a":549},"How many impressions does a CTR curve need?","Google publishes no minimum. A position bucket built from a few hundred impressions moves a lot from week to week, so merge thin buckets such as positions 6 to 10 and 11 to 20, and use a longer date range before you trust a point.",{"q":551,"a":552},"Why is my CTR at position 1 lower than the curve suggests?","Common reasons are rich results from competitors, a query you rank for that searchers do not associate with your site, or a searcher who already got the answer on the results page. Check the results page for that query before rewriting the title.",{},"A CTR curve is the expected CTR at each ranking position. How to build yours from Search Console and spot queries that under-earn for their rank.","CTR Curve: Organic CTR by Position, Explained",true,"\u002Fglossary\u002Fctr-curve",[27,37,50,45,100,332,559,362],"\u002Fblog\u002Fhow-to-easily-find-quick-wins-to-boost-your-traffic",{"title":5,"description":536},{"loc":557},[563,567,570,573,576,580],{"publisher":564,"name":565,"url":566,"accessed":534},"Google Search Console Help","What are impressions, position, and clicks?","https:\u002F\u002Fsupport.google.com\u002Fwebmasters\u002Fanswer\u002F7042828",{"publisher":564,"name":568,"url":569,"accessed":534},"Performance report (Search results): Overview and basic setup","https:\u002F\u002Fsupport.google.com\u002Fwebmasters\u002Fanswer\u002F7576553",{"publisher":564,"name":571,"url":572,"accessed":534},"About Search Console data","https:\u002F\u002Fsupport.google.com\u002Fwebmasters\u002Fanswer\u002F96568",{"publisher":564,"name":574,"url":575,"accessed":534},"Table guidelines and reference (bulk data export)","https:\u002F\u002Fsupport.google.com\u002Fwebmasters\u002Fanswer\u002F12917991",{"publisher":577,"name":578,"url":579,"accessed":534},"Google Search Central Blog","Optimizing website performance with a Search Console bubble chart","https:\u002F\u002Fdevelopers.google.com\u002Fsearch\u002Fblog\u002F2022\u002F04\u002Fperformance-optimization-bubble-chart",{"publisher":577,"name":581,"url":582,"accessed":534},"Introducing the branded queries filter in Search Console","https:\u002F\u002Fdevelopers.google.com\u002Fsearch\u002Fblog\u002F2025\u002F11\u002Fsearch-console-branded-filter","glossary\u002Fctr-curve","The expected organic click-through rate at each ranking position, built from your own Search Console data.",[586,587,588,589,590],"A CTR curve maps each ranking position to the click-through rate a site normally earns there; the value at a position is that query's expected CTR.","Build each point as total clicks divided by total impressions for the position bucket, never as an average of row-level CTRs.","Your own curve beats a published industry table, because brand, niche and SERP layout change the level of the curve.","A query far below the curve at a top position is a title and snippet problem; a query on the curve at position 14 is a ranking problem.","Branded queries, AI Overviews and carousels distort the curve, so build it from non-branded queries and compare like with like.",1,"jzx4EamnPsaI8tomY05P_7zGtDx4xD2DuiHZCpSjhZ0",[594,597,600,603,606,609,612,614],{"path":27,"label":595,"summary":596},"CTR (Search Console)","Click-through rate — clicks divided by impressions — the share of appearances in search that turned into a visit.",{"path":37,"label":598,"summary":599},"Average position (Search Console)","The topmost position your site held each time it appeared for a query, averaged across all of those impressions.",{"path":50,"label":601,"summary":602},"Clicks (Search Console)","One click is counted when a user leaves Google for your page from a search result; repeat clicks on the same link in one search count once.",{"path":45,"label":604,"summary":605},"Impressions (Search Console)","Counted each time a link to your site appears in a loaded page of search results, whether or not anyone clicks it.",{"path":100,"label":607,"summary":608},"Striking distance keywords","Queries where a page already ranks just off page one, typically positions 11–30, one nudge from real clicks.",{"path":332,"label":610,"summary":611},"Low-hanging fruit SEO","Prioritising the rankings closest to paying off: low-difficulty keywords where a small push moves traffic.",{"path":559,"label":613,"summary":539},"SEO Quick Wins: The 15-Minute Search Console Playbook",{"path":362,"label":615,"summary":539},"How to Find Low-Hanging Fruit Keywords in Search Console"]