The 40% threshold is the line where AI-driven query cannibalization becomes the dominant fact in SEO budget decisions. Below 40% you can argue AI search is a new-channel opportunity. At 40% it's the dominant channel-share shift, and the strategy stack flips.
Q2 2026 baseline puts AI-cannibalized organic clicks at roughly 17% — already enough to disrupt mid-market SEO programs but not yet the strategic inflection. The forecast question is when 40% lands. Three scenarios bracket the answer from Q2 2027 (aggressive) to Q2 2028 (conservative); the probability-weighted blend points to Q3 2027.
What follows is the model, the five levers that move the curve, the per-quarter implications, and the playbook for in-house SEO and content teams operating against the inflection. The forecast is a planning input — not a single-point prediction.
- 01The 40% threshold matters because it flips strategic posture from 'new channel' to 'dominant share shift'.Below 40%, organic-search investment can defend share. At 40%, AI search captures more click volume than traditional organic for major informational query mixes — strategy must shift from defending blue-link share to capturing AI-citation share.
- 02Q2 2026 baseline: 17% of organic clicks already cannibalized.Estimated from SimilarWeb LLM traffic + Authoritas AI Overview tracking + Profound visibility data. Concentrated in informational and product-comparison queries; transactional queries less affected. The 17% baseline is well above the 8-12% that 2025 forecasts projected for Q2 2026.
- 03Probability-weighted forecast: 40% by Q3 2027 (blended of three scenarios).Aggressive (0.30 weight): Q2 2027. Base case (0.50 weight): Q4 2027. Conservative (0.20 weight): Q2 2028. Probability-weighted blend lands at Q3 2027. The 12-month bracket between aggressive and conservative reflects real uncertainty in adoption pace.
- 04Five levers move the curve: AIO query coverage, AIO citation rate, Perplexity/ChatGPT MAU, query-intent cannibalization mix, and advertiser pull-back.AIO query coverage is the largest single lever. Perplexity + ChatGPT MAU growth is a close second. Query-intent mix matters because informational queries cannibalize earlier than transactional. Advertiser pull-back is the conservative-scenario driver — if Google's AIO ad-revenue hit forces a pull-back, the curve flattens.
- 05Plan against Q3 2027 — but accept the bracket. Build flexibility into channel-investment ramps.The 12-month bracket between scenarios means you need a curve-tracking discipline, not a single-point plan. Run quarterly reads on the five levers; ramp AI-citation investment as signals strengthen; maintain organic-share defense investment until the cannibalization curve clearly crosses 30%.
01 — The ThesisWhy 40% is the load-bearing number.
The 40% threshold is not magic — it's where two strategic regimes flip. Below 40%, organic-search investment defends meaningful share against AI cannibalization. At 40%, the cannibalized share becomes larger than the defendable share for major query mixes, and the strategy must shift from defense to capture. Above 50% the shift is decisive; below 30% it's still a tactical question.
The 40% line is the procurement-relevant inflection because it triggers budget-shape changes. SEO budgets at mid-market enterprises are typically 60-80% organic content production and 20-40% on-site technical work. At 40% AI cannibalization, the rational reallocation moves content production from blue-link target keywords to AI-citation target prose — a different content architecture, different editorial voice, different measurement framework.
The strategic question is when the 40% line lands. That's what the forecast model answers.
"Below 40%, defend organic share. At 40%, capture AI-citation share. The strategic regimes are genuinely different."— Internal SEO strategy memo, April 2026
02 — Where We Are TodayThe 17% baseline.
Q2 2026 baseline puts AI-cannibalized organic clicks at roughly 17% across an informational + product-comparison query mix typical of mid-market B2B SaaS. The estimate blends three signals: SimilarWeb LLM traffic data (showing direct AI-search traffic growth), Authoritas AI Overview tracking (showing AIO coverage and click-through displacement), and Profound visibility data (showing citation share inside AI answers).
Of tracked queries · Q2 2026
Google AI Overviews surface on roughly 47% of tracked SEO query sets — up from 31% in Q1 2026. Coverage concentrated in informational and how-to query intent; transactional queries still mostly blue-link.
+16 pts QoQClick-through on AIO queries
When AIO surfaces, blue-link CTR drops roughly 42% from pre-AIO baseline. The displaced clicks split between AIO citations (going to cited domains) and zero-click (no domain visit at all).
Sistrix Q2 2026Q2 2026 MAU estimate
Perplexity Q2 2026 MAU estimated at 210M, up from 145M in Q1. Growth pace +45% QoQ; cumulative growth from launch tracking the ChatGPT consumer-search adoption pattern but at faster slope.
+45% QoQQ2 2026 MAU estimate
ChatGPT Search feature MAU estimated at 180M (subset of total ChatGPT 800M MAU), up from 115M in Q1. Growth tracks consumer ChatGPT adoption with a 6-9 month lag.
+57% QoQ03 — Three ScenariosThe forecast range.
We model three scenarios for when AI cannibalization of organic clicks crosses 40%. Probability weights derived from Q1+Q2 2026 base-rate fit and observed lever trajectories.
Q2 2027 · 40% (0.30 probability weight)
AIO query coverage doubles from 47% to 94% by year-end 2026; Perplexity MAU hits 800M by Q1 2027; ChatGPT search MAU hits 600M; advertiser pull-back signal does not materialize. Cannibalization compounds quarterly. The 40% line lands by mid-2027.
0.30 probabilityQ4 2027 · 40% (0.50 probability weight)
AIO query coverage grows from 47% to 70% by year-end 2026 (slower expansion); Perplexity MAU 600M by Q1 2027; ChatGPT search 450M; modest advertiser pull-back signal. Cannibalization grows steadily. The 40% line lands by late 2027. Most likely path.
0.50 probabilityQ2 2028 · 40% (0.20 probability weight)
AIO query coverage flatlines around 60% (Google ad-revenue defense); Perplexity hits adoption plateau at ~400M MAU; ChatGPT search cannibalization slows. Cannibalization grows but at decelerating pace. The 40% line lands in mid-2028. Unfavored but credible.
0.20 probabilityAI search cannibalization curve · 6-quarter forecast bands
Forecast: SimilarWeb + Authoritas + Profound + Sistrix · scenario-weighted · Apr 202604 — The 5 LeversWhat moves the curve.
Five levers determine which scenario plays out. Watch them quarterly; the curve shifts whenever a lever moves faster or slower than expected.
AIO query coverage
% of tracked queries surfacing AIOLargest single lever. Q2 2026 at 47%; aggressive scenario projects 94% by year-end. If Google extends AIO coverage aggressively (informational + transactional + comparison), cannibalization accelerates. If Google narrows coverage to defend ad revenue, curve flattens.
Largest leverAIO citation rate
Domains cited per AIO answerCitation rate determines whether displaced clicks go to cited domains or to zero-click. Q2 2026 averages 3.2 citations per AIO answer. Higher citation rate softens cannibalization (clicks redistribute); lower rate hardens it (more zero-click).
Distribution leverPerplexity + ChatGPT search MAU
Direct AI-search traffic growthAggregated MAU tracks adoption pace independent of Google AIO. Q2 2026: Perplexity 210M + ChatGPT search 180M. Combined growth pace 50%+ QoQ. If sustained, direct AI-search traffic surpasses 1B MAU by Q2 2027 and pulls forward the cannibalization curve.
Adoption leverQuery-intent cannibalization mix
Informational vs transactional splitInformational queries cannibalize first; transactional queries lag. Q2 2026 mix: 31% informational AIO-cannibalization, 11% transactional. As AI engines improve transactional answer quality (product comparison, pricing, availability), the mix shifts toward higher overall cannibalization.
Mix leverAdvertiser pull-back signal
Google ad-revenue defensive movesConservative-scenario driver. If Google AIO ad-revenue impact triggers a defensive pull-back (narrowing AIO coverage, pushing toward sponsored-AIO model, or de-prioritizing AIO surface), the curve flattens. Watch Q3-Q4 2026 Google earnings ad-revenue commentary.
Defensive lever05 — Implications by QuarterWhat this means for SEO budgets.
The forecast translates into specific budget-shape implications by quarter. The pacing matters because channel reallocation lags strategic decision by 6-12 months in most mid-market enterprises.
Cannibalization 17% → 24% · diversify quietly
Below 30% cannibalization, organic-share defense still earns. Begin diversifying content production to support AI-citation share — but don't pull defensive investment yet. The cost of premature reallocation is real.
Defend + diversifyCannibalization 24% → 31% · ramp AI-citation investment
Cannibalization clearly above 25%. Ramp AI-citation investment to 30-50% of content production budget. Maintain organic-share defense for transactional and product-comparison queries; reallocate informational-query budget toward AI-citation capture.
Ramp · AI-citationCannibalization 31% → 40% · strategy regime flip
Cannibalization crosses 40% in base case. Strategy regime flips — AI-citation share becomes the dominant channel investment for informational query mix. Organic-share defense narrows to transactional + branded query support. Editorial voice and content architecture shift.
Flip · capture-modeCannibalization 40%+ · capture-mode steady state
Steady-state in capture mode. Most informational content production tuned for AI-citation capture; defensive organic investment narrowed to query-mixes still defending share. SEO budget shape stabilizes in new equilibrium roughly 18-24 months after threshold crossing.
Steady state"Channel reallocation lags strategic decision by 6-12 months. The teams that wait for 40% to land first will spend 2028 catching up."— SEO leadership memo, April 2026
06 — What To Do This YearConcrete moves for 2026.
Build AI-citation tracking now
Stand up Profound or Authoritas tracking on your top-200 informational queries. Track citation share quarterly. The data is the leading indicator for whether your content architecture is positioned for capture-mode.
Highest leverageAudit content architecture against AI-citation tactics
From the GEO contrarian audit: opinion density, named-author attribution, verb-rich attribution inside prose, prose-first markdown rendering. Score top-traffic articles against these axes. Most teams fail at least 2 of 4 on the highest-traffic templates.
AI-citation hygieneDiversify content-production allocation
Shift 20-30% of content-production budget toward AI-citation-optimized templates by year-end. Don't pull defensive investment yet, but begin building the muscle and editorial voice. The lead-time on editorial voice change is 6-12 months.
Allocation shiftSet quarterly lever-tracking discipline
Run quarterly reads on the five levers. AIO query coverage, AIO citation rate, Perplexity + ChatGPT search MAU, query-intent cannibalization mix, advertiser pull-back signal. The discipline catches scenario shifts before they bind.
Curve discipline07 — ConclusionPlan against the curve, not the headline.
The strategic regime flip is the load-bearing event — not the specific quarter.
The 40% line will land somewhere between Q2 2027 (aggressive) and Q2 2028 (conservative). The probability-weighted blend points to Q3 2027. The 12-month bracket reflects real uncertainty in lever trajectories — not lazy forecasting.
The decision-relevant fact is the regime flip itself, not the specific quarter. Teams that build AI-citation tracking now, audit content architecture against the tactics that move the citation needle, diversify content-production allocation, and set quarterly lever-tracking discipline will be ready when the threshold lands. Teams that wait until 40% lands will spend 2028 catching up.
We will publish the next forecast update late July 2026 alongside the Q2 quarterly report, with refreshed lever data and recalibrated scenario weights. Bookmark this page; we will edit-in-place to update the probability-weighted forecast point as new data lands.