The Click-Through-Rate Decay Curve and SERP-Position Economics

TL;DR: The click-through-rate-by-position curve that operators inherited from the early 2010s (a clean exponential decay where position 1 captures roughly a third of clicks and position 10 captures barely two percent) is not the curve a real SERP exhibits in 2024. Featured snippets, People Also Ask blocks, video carousels, image packs, and the rollout of AI-generated answer boxes have flattened the curve at positions 1 to 3 and compressed the long tail. The published studies from Sistrix, AWR, Backlinko, and First Page Sage agree on the direction but disagree on the magnitude, and the curves diverge sharply by query intent. The operator question is no longer "what does the CTR curve look like" but "what is the marginal value of one position of improvement for this specific query class on this specific SERP layout."

A note on the named companies and sources. The Sistrix 2020 and 2023 studies, the Advanced Web Ranking quarterly reports, Backlinko's 2023 ranking study (Brian Dean and team), First Page Sage's industry CTR breakdowns, and the Google Search Quality Rater Guidelines appear throughout as the available public reference points. Quantitative ranges framed as advisory-engagement observation come from anonymized partner operators across e-commerce, B2B SaaS, and publishing verticals in the 200K to 80M monthly impressions range, not from the cited vendors or studies.

---

The Curve Everybody Quotes, and Why It Stopped Working

Almost every published presentation on SEO economics opens with a chart of click-through rate against SERP position. The chart usually shows position 1 at somewhere between 28 and 35 percent, position 2 around 15 percent, position 3 around 10 percent, and a long tail that drops below 2 percent by position 10. The curve has the shape of a clean power-law decay, and the strategic implication that gets drawn from it is direct: every position of improvement is roughly worth a multiple of the next position's traffic, and the most valuable single change is the move from position 2 to position 1.

The curve was approximately right in 2011. It was already loose by 2017. In 2024 it is wrong enough that operating decisions made against it tend to be wrong. The three patterns that broke it were, in rough chronological order, the colonization of position zero by featured snippets, the rise of mixed-content SERPs (image packs, video carousels, People Also Ask, local packs), and the rollout of AI-generated summaries above the organic blue links. Each pattern shifted clicks away from the canonical positions and into surfaces the original CTR studies did not measure.

The Sistrix 2020 study, drawing on roughly 80 million keywords across Google's index, already showed a CTR for position 1 of about 28.5 percent on desktop and 22.4 percent on mobile, well below the 31 to 35 percent that the older AOL leak (2006) and the early Optify studies (2011) had reported. The Sistrix 2023 update, looking at the same query universe with the newer SERP layouts, showed position 1 down further in many query classes, with the largest drops on transactional and "near me" queries where the local pack and shopping carousels absorbed the disproportionate share of the click. The AWR quarterly studies tracked a similar direction.

What the Major Public Studies Actually Say

Four public studies are cited often enough that they have become the de facto reference points for CTR-by-position discussion: the Sistrix 2020 and 2023 studies, the Advanced Web Ranking quarterly reports, the Backlinko 2023 ranking study, and the First Page Sage breakdowns by industry. The numbers diverge on magnitude but converge on shape, and the divergences are themselves informative.

The spread between the studies at position 1 is wide (22.7 to 39.8 percent), and that spread is itself a finding. First Page Sage's 39.8 percent figure comes from a narrower industry sample (B2B and high-intent commercial verticals) where the SERP layout still skews "ten blue links" and the position-1 brand is often a recognized authority; Sistrix 2023's 22.7 percent median is computed across a much broader query universe that includes the heavy SERP-feature load on consumer queries. The Backlinko number (27.6 percent) sits between the two, sampling a mid-volume query universe with mixed intent.

The position 2 to 5 numbers cluster more tightly across studies (within 3 to 5 absolute percentage points), and from position 6 downward the studies essentially agree. This suggests that the position 1 measurement is the one most sensitive to the SERP-feature environment, and that the long tail behaves more predictably than the head.

The shape comparison is the most useful reading of this chart. All five curves drop steeply from position 1 to position 3, decline more gradually from 3 to 6, and flatten into a slow tail from 6 to 10. The Sistrix 2023 curve sits below the others through most of the range, consistent with the broader SERP-feature presence in its sample. The First Page Sage curve is steeper at the top because its sample is dominated by query classes where the legacy "ten blue links" layout still applies and brand authority concentrates clicks on position 1.

Query Intent and the Multimodal SERP

The single most important refinement to the CTR curve is intent. Google's own documentation distinguishes between informational, navigational, transactional, and (more recently) commercial-investigation intent, and the SERP layout differs systematically across the four. The CTR curve looks one way on a "how to" informational query, another way on a "best running shoes for flat feet" commercial-investigation query, and a third way on a "buy nike pegasus 40" transactional query.

The informational curve is the most affected by featured snippets and AI summaries. When a query is answered in the snippet, the snippet source captures a disproportionate share of clicks and the rest of the SERP cannibalizes itself. In partner data across publisher operators, when a featured snippet appears, the source URL's CTR runs roughly 35 to 47 percent of total impressions, while positions 2 through 5 collectively drop to a combined CTR in the high teens. The total click volume on the query is also smaller than it would be without the snippet, because some users find the answer and never click.

The transactional curve is the most affected by shopping carousels, local packs, and image carousels. A query like "best wireless earbuds 2024" returns a shopping carousel at the top of the SERP on most devices, and the organic position 1 sits below the carousel. The position-1 CTR in this layout runs roughly 7 to 12 percent, not the 22 to 32 percent the blended curve suggests, because the shopping units absorb the high-intent click before the user reaches the organic list.

The navigational curve is the most stable across SERP layouts, because navigational queries (a brand name, a known URL) reliably resolve to a single dominant result and the user clicks it almost without exception. Position-1 CTR on a navigational query runs in the 60 to 80 percent range, with the remaining traffic split across knowledge-panel links, sitelinks, and the occasional related result. The blended-intent curves understate this case because the navigational queries are a small share of total query volume.

The intent split changes the operating math substantially. A SaaS marketing team chasing "best CRM for small business" is in a commercial-investigation layout where positions 2 and 3 are nearly as valuable as position 1 (the difference is 17.3 versus 12.84 versus 10.4 percent CTR); the marginal cost of going from 3 to 1 is rarely worth the engineering effort. A publisher chasing "what is hyperbolic discounting" is in an informational layout where capturing the featured snippet (the position-zero outcome) is worth ten times as much as ranking 2 with no snippet. The operating decision is structurally different, and the blended curve hides it.

How AI Answer Boxes and Featured Snippets Reshape the Top

The featured snippet predates AI-generated answers by nearly a decade; it has been a SERP feature since 2014. The pattern it created (one organic result lifted into a callout above the blue links, with text extracted directly from the page) gave operators a clear strategic prize: structure your content so it could be extracted as the snippet, win the snippet, capture the disproportionate share of clicks. The 35 to 47 percent CTR on snippet sources is well-documented across the public studies.

The newer AI-generated answer boxes (Google's Search Generative Experience in 2023-2024, the AI Overviews rollout, and the corresponding features in Bing's Copilot and the various smaller search engines) extend the snippet pattern with a more substantial visual footprint and the ability to summarize across multiple sources rather than extract from one. The early measurement on AI Overviews shows two distinct effects: the source URLs cited in the overview see a CTR lift roughly comparable to (sometimes slightly below) the classic snippet CTR, and the non-cited organic positions below the overview see a CTR drop of roughly 18 to 34 percent compared to the same query when no overview is present.

The net effect on the curve is asymmetric. The top 1 to 2 cited sources capture more clicks than they would have at the equivalent organic position; positions 3 through 10 capture meaningfully fewer. The total clicks on the query may be lower (some users get the answer and do not click), but the distribution of remaining clicks is more concentrated at the top. The flattening of the curve in the 1 to 3 range gives way to a sharper drop into the long tail.

The "1 (cited)" row breaks the convention of using a single number per row because the position 1 organic result behaves very differently depending on whether the AI Overview cites it or not. A cited position 1 sees a CTR lift (the overview functions as additional visual real estate above the result); an uncited position 1 sees a substantial drop (the overview captures the click that would otherwise have gone to the top organic result). The strategic implication for publishers is that being cited in the overview is now nearly as important as ranking organically, and the citation patterns of the overview are not entirely correlated with the organic ranking.

The Unit Economics of Moving Up the Curve

The CTR curve has direct unit-economic implications. If a query has 10,000 monthly impressions, the difference between position 6 (perhaps 4.27 percent CTR) and position 3 (roughly 9.84 percent) is 557 incremental monthly clicks. The difference between position 3 and position 1 (in the non-snippet informational layout, roughly 26.84 percent) is another 1,700 clicks. The marginal returns are highly nonlinear, and the cost to achieve each move is rarely proportional to the click gain.

The cost asymmetry runs the other direction: moving from page 2 to page 1 typically costs less in content and link investment than moving from position 6 to position 3, which in turn typically costs less than moving from position 3 to position 1. The position 6 to 3 move is often a content-quality and on-page-optimization investment ($3K to $18K typical in partner engagements). The position 3 to 1 move is often a brand-authority and link-graph investment ($45K to $190K typical, sometimes higher and rarely lower). The math of where the marginal dollar of SEO spend belongs depends on both the click gain and the cost asymmetry, and the breakpoint sits in different places for different operators.

The chart makes the intent-dependency of the marginal economics visible. On informational queries, the position-3-to-1 move and the position-1-to-snippet move are both large, suggesting that pushing for the snippet is the highest-marginal-value SEO move in this intent class. On transactional queries, the marginal value of moving from 6 to 3 is roughly comparable to the marginal value of moving from 3 to 1, and the snippet has almost no marginal value because transactional queries rarely trigger a snippet. The commercial-investigation queries sit in between.

The economics work the other direction too. There are query classes where the position-1 push is unambiguously correct: high-volume informational queries with a stable snippet opportunity, navigational queries on competitor brand terms, transactional queries with a small SERP-feature footprint, and queries where the organic-1 result drives a particularly high conversion rate due to brand familiarity. The discipline is doing the math per query class, not applying the legacy CTR curve as a default.

Position Volatility and the "What Position Am I At" Problem

A complication that the standard CTR-curve analysis tends to elide is that position is not a single number per query. The same query, run by different users in different sessions, can return materially different SERPs. The variation comes from device (desktop vs mobile vs tablet), geo (country, region, ZIP code), language, signed-in personalization, the user's recent browsing history, the time of day, the rapid SERP-feature iteration that Google runs in production, and the rotation in search results that the documentation calls "freshness" or "QDF" (query deserves freshness).

In partner data on the same set of 1,200 commercial keywords tracked over 14 consecutive days, the median position per query drifted by roughly 0.4 to 1.8 positions on a typical day-over-day basis, with weekend volatility roughly 35 percent higher than weekday. The standard deviation of position around the median per query ran 0.9 to 2.7 positions. A query nominally ranking "position 3" might in practice be returning positions 2, 3, 4, or occasionally 6 across the population of actual SERPs returned.

This volatility means the CTR curve has to be applied to a distribution of positions, not to a point estimate. A query ranked 3 with low volatility (95 percent of SERPs returning position 2 to 4) will have an effective CTR close to the position-3 curve value. A query ranked 3 with high volatility (50 percent at position 3, 25 percent at position 2, 25 percent at position 5 or 6) will have a higher effective CTR than the position-3 value, because the position-2 SERPs lift the average more than the position-5 SERPs depress it (the curve is convex in this range). The asymmetric shape of the CTR curve interacts with the position distribution to produce an effective CTR that is not just the curve value at the median position.

Brand Strength as a CTR Multiplier

The position-CTR curve assumes that the user reads the SERP linearly from top to bottom and clicks the first result that looks relevant. In practice, users sweep the SERP and click the result they recognize, and recognition is heavily weighted by brand familiarity. A position-3 result with a strong brand can out-click a position-1 result with a weak brand, especially on commercial-investigation queries where the user is already evaluating brands.

The pattern shows up in the Sistrix 2023 data and in the Backlinko 2023 study as a consistent CTR uplift for recognized brand domains relative to the curve baseline. Across the partner data we have audited, the brand-strength uplift on commercial queries runs in the range of 1.3 to 2.4 times the curve baseline for top-tier recognized brands (the dominant retailer in a category, the dominant SaaS player), 1.1 to 1.6 times for mid-tier brands, and approximately neutral (0.85 to 1.10 times) for lesser-known brands. On informational queries the brand effect is smaller, because users are looking for answers rather than recognized providers, and the recognition signal weighs less.

The "recognized but negative-association" row is worth highlighting because it is the only multiplier below the unknown-entrant baseline. The pattern shows up for brands that the user recognizes but has a reason to avoid (a service the user previously cancelled, a brand with negative press in the user's recent reading, a known low-quality competitor in the category). The CTR drops below the curve baseline because the user actively skips the recognized name. The pattern is small in the data (it requires a population with prior brand exposure to manifest), but it is consistent enough across the operators we have audited to be worth naming.

The brand multiplier interacts with the curve flattening at the top. On a SERP where positions 1 to 3 are CTR-flat (the case in many commercial-investigation queries with shopping carousels and AI overviews above), the brand signal becomes the dominant lever between the three positions, and the operator who has invested in brand-search lift sees the benefit even without changing the underlying organic position. This is one of the underappreciated mechanisms by which a brand-marketing investment pays back into the SEO funnel.

Mobile Versus Desktop, and the Forgotten Layout Differences

The CTR curve is reported in the public studies usually as a blended desktop-and-mobile number, sometimes split with a separate column. The split is more important than the blended view suggests, because the mobile SERP has structurally different real estate, scroll behavior, and SERP-feature density than the desktop SERP.

The first-screen real estate on a mobile SERP holds roughly one to one and a half organic results before the user scrolls; the desktop SERP holds three to five. This means the position-1-versus-position-2 CTR gap is larger on mobile than on desktop (the user often does not scroll past the first result before refining the query), while the position-3-versus-position-4 gap is smaller (both are already below the fold). The Sistrix 2023 study reports a mobile position-1 CTR around 16 to 19 percent on commercial queries, versus 22 to 27 percent on desktop for the same query universe.

The mobile SERP also has heavier SERP-feature density (the local pack appears more aggressively, the shopping carousel takes a larger share of vertical real estate, the People Also Ask block expands inline more often). The result is that the mobile CTR curve flattens earlier (positions 1 and 2 are closer in CTR on mobile than on desktop) and the long tail is even longer (the user is less likely to scroll to positions 6 to 10 on mobile, because doing so requires more thumb work and the alternative is to refine the query).

The strategic implication is that the device split of the user base matters for the SEO economics. An e-commerce operator with 78 percent mobile traffic faces a CTR curve that flattens earlier and decays faster than the blended curve suggests; a B2B SaaS operator with 65 percent desktop traffic faces a curve closer to the published averages. The right curve to apply to the unit-economic math is the device-weighted curve for the operator's actual traffic mix, not the blended public curve.

A Practical Reframe of the Position-Economics Question

The accumulation of patterns (intent dependence, SERP-feature distortion, AI Overview reshape, brand multiplier, mobile-desktop split, position volatility) makes a single CTR curve a poor instrument for operating decisions. The reframe that has worked in advisory engagements is to abandon the single-curve view and adopt a query-class-weighted view, where each commercial query is bucketed by its actual SERP layout and intent, the relevant CTR curve is applied per bucket, and the unit-economic math is done per bucket.

The buckets that have proven operationally useful are roughly: informational with stable snippet opportunity, informational without snippet, commercial investigation with shopping carousel above the fold, commercial investigation with classic ten-blue-links layout, transactional with strong local pack, transactional without local pack, navigational on own brand, navigational on competitor brand, and the residual "other" bucket. Each bucket has a different CTR curve, a different cost structure for moving up the curve, and a different downstream conversion rate.

The cost ranges in this table vary deliberately. A snippet-capture content investment for an informational query rarely costs the same as a brand-search lift program for a commercial query. The point of the bucketing is to force the operator to ask "which bucket is this query in, and what is the actual cost of the right intervention for this bucket" before committing engineering or content budget.

The CTR curve is not a single function of position; it is a family of functions, one per query class, each with its own shape and its own marginal economics. The operator who treats it as a single curve and chases position improvements uniformly will spend the marginal SEO dollar in the wrong place most of the time.

A Note on Measurement: Where Search Console Helps and Where It Misleads

Google Search Console reports a CTR per query per page, which makes it tempting to derive an empirical CTR curve directly from the operator's own data. The temptation is mostly worth resisting, because the Search Console CTR is computed at a very specific aggregation level (the query as Google records it, the page that received the click, the position averaged across all impressions in the date range) and it conflates several distinct quantities.

The most common confusion is between the "position" Search Console reports (the average position across all impressions, weighted equally) and the position the user actually saw on any given session. A page that appeared in positions 2, 3, 5, 6, and 8 across five sessions would report an average position of 4.8, but the CTR was driven by the position-2 and position-3 sessions. The reported CTR-and-position pair does not let the operator distinguish "I ranked 4.8 with low volatility" from "I averaged 4.8 with high volatility," and the implied CTR curve is misleading either way.

The more useful Search Console workflow is to use the query-level CTR as a benchmark against the bucket-specific public-study curve, and to investigate the deltas. A query with CTR substantially above the curve baseline suggests a strong title or meta description, or a brand multiplier, or a SERP-feature presence the operator is benefiting from. A query with CTR substantially below the baseline suggests the opposite, or a different intent than the operator assumed, or a competitor capturing the click that the position would otherwise suggest. The investigation is more valuable than the empirical curve.

The other useful Search Console signal is the impression-to-click trend over time for the same query. A query where impressions are stable but CTR is declining is a query where the SERP layout has shifted underneath the page (a new SERP feature appeared, or the AI Overview started covering the query), and the operator's curve assumption has stopped applying. The investigation is then to check what changed in the SERP and decide whether the right response is to optimize for the new layout or to reallocate to a different query.

Key Takeaways

1. The CTR-by-position curve quoted in most SEO presentations dates from a different SERP era. Sistrix 2020 and 2023, AWR, Backlinko 2023, and First Page Sage report a position-1 CTR in the 22 to 40 percent range depending on sample, materially lower than the 31 to 35 percent legacy number and highly variable by query class.
2. Query intent is a stronger predictor of effective CTR than position. The informational-with-snippet, commercial-investigation-with-carousel, and transactional-with-local-pack layouts each produce CTR curves that diverge from the blended public curve by factors of 2x to 5x on individual positions.
3. AI Overviews lift the cited source's CTR and depress the non-cited positions below the overview. The early measurement on partner data suggests an 18 to 34 percent CTR drop on positions 2 through 10 when the overview is present and the position is not cited; the cited source sees a lift roughly comparable to the classic featured-snippet CTR.
4. The unit economics of moving up the curve are highly nonlinear and depend on both the click gain (which the curve gives) and the cost of the move (which the curve does not). The position 3 to 1 move is roughly four to ten times more expensive than the position 6 to 3 move, and the breakeven depends on the per-click conversion value.
5. Position volatility means a single "rank tracker" number per query is a poor summary. The effective CTR depends on the full position distribution, and the asymmetric shape of the CTR curve interacts with the variance to produce a CTR that the median position alone does not predict.
6. Brand strength multiplies the CTR at a given position, with top-tier recognized brands running 1.3 to 2.4 times the curve baseline on commercial queries and unknown brands running approximately neutral or below. The brand multiplier is one of the underappreciated SEO returns on a brand-marketing investment.
7. The mobile and desktop CTR curves differ enough that a device-weighted curve, applied to the operator's actual traffic mix, is a better instrument for unit-economic math than the blended public curve.
8. The reframe that has worked in advisory engagements is to abandon the single-curve view and bucket queries by their actual SERP layout and intent, apply the relevant curve per bucket, and do the cost-per-move math per bucket. The discipline is per-bucket, not per-position.

Citations and Further Reading

• Sistrix, "Why (Almost) Everything You Know About Google CTR Is No Longer Valid" (2020) and the 2023 update, the most-cited public reference on CTR-by-position shifts across the SERP-feature era.
• Advanced Web Ranking, "Google CTR Studies" (quarterly), tracking the CTR curve across roughly 12 million keywords with country and device splits.
• Backlinko, Brian Dean and team, "We Analyzed 4 Million Google Search Results" (2023), the most-cited ranking study with CTR-by-position data.
• First Page Sage, "Google Click-Through Rates in 2023" by industry, useful for the B2B and commercial-vertical CTR baselines.
• Google Search Central documentation on SERP features, featured snippets, and the Search Generative Experience, the canonical source for the layout mechanics.
• Optify and AOL leak (2006-2011) data, the historical baseline against which the contemporary curves are usually compared.
• Aleyda Solis and the published SEO operator literature, on the practical implications of curve flattening for ranking strategy.
• Lily Ray and the Amsive Digital ranking-update tracking, on the role of SERP layout shifts in observed CTR changes over time.
• The Search Quality Rater Guidelines (Google, updated periodically), Sections 12-14 on the rater treatment of query intent classification.
• John Mueller and Search Off the Record podcast threads on SERP-feature rollout and the question of CTR as a ranking signal (it is largely not).
• The published work of Onely (Bartosz Goralewicz and team) on rendering, indexability, and how the AI Overview citation behavior interacts with technical SEO health.