Fewer Clicks, Better Buyers, and the New Content Forecast Math

A B2B content team we audited in January built its FY26 plan on 8% organic session growth. By the end of Q1, sessions to its highest-value pages were down 38% year over year, lead volume was down 11%, and pipeline was flat. The CFO asked which number was wrong. The honest answer: the model. Every input still measured a world where traffic and revenue moved together, and that world ended sometime in 2025.

The denominator collapsed on schedule

This is measured, repeated, and industry-wide. HubSpot has been public about the mechanics of its own traffic decline: informational queries that once fed the blog are now answered inside AI surfaces, search volume to key pages fell 58%, and the visitors who still click arrive with materially higher intent. Ahrefs re-ran its AI Overviews study on 300,000 keywords with December 2025 data and found the presence of an AI Overview correlates with a 58% lower click-through rate for the top-ranking page. Seer Interactive, tracking 5.47 million queries across 53 brands, watched organic CTR on AIO-present queries fall to a 1.3% floor in December 2025 before recovering to 2.4% in February 2026, which it reads as a possible "new normal" rather than a rebound.

Whatever coefficient you pick from that table, the direction is settled. Plan on it.

Why the old formula double-counts optimism

The legacy content forecast is a four-factor multiplication:

Pipeline = Sessions × visitor-to-lead CVR × lead-to-deal rate × ACV

It fails now because the first factor is falling for structural reasons while the second is rising for the same structural reason. AI answers absorb low-intent clicks; the survivors are pre-qualified. Forecasting session growth and holding CVR flat books the loss twice and the offset zero times. The team from the opening paragraph did exactly that: their high-intent page CVR had climbed from 1.1% to 1.9% and nobody had re-run the model.

The new baseline math

Rebuild the forecast on three lines, each with a defensible formula.

Line 1: Qualified sessions, forecast down. Segment pages by query intent. Apply a 35–55% haircut to informational-page sessions (the Ahrefs and HubSpot range) and a 0–15% haircut to comparison and pricing pages, which AI engines still cite and route.

Line 2: Revenue per session (RPS), forecast up. RPS = attributed revenue ÷ organic sessions. This is the metric that survives the shift, because it improves when the mix shifts toward high-intent visitors even as volume falls. Semrush's 4.4x conversion-value finding for AI search referrals is the ceiling case; a 2–4x RPS improvement on affected pages is the conservative planning band. Semrush also projects AI search could drive more visits than traditional search for its studied topics by early 2028, so this line grows in weight every quarter.

Line 3: AI answer share, forecast from zero. Citations inside ChatGPT, Perplexity, and AI Overviews are now a top-of-funnel asset with no session attached. Track share of answers for your money queries with an AI visibility checker and treat it as the leading indicator for Line 2. The tactics for earning those citations are a separate discipline; we covered them in how to get cited by ChatGPT and AI Overviews.

Here is the worked example, using the team from the open:

Sessions down 41%, pipeline up 6%. That is the entire thesis in two cells. If your instinct is that the CVR lift looks generous, check your own high-intent pages first; most teams find the lift already happened and the dashboard never surfaced it. Run your own numbers through our marketing metrics calculator before the next board cycle.

The memo to your CFO

CFOs do not fear falling numbers; they fear falling numbers nobody predicted. Re-anchor in one page:

1. Reprice the denominator. State plainly that 40–60% of historical organic sessions were low-intent volume that AI surfaces now absorb at zero cost to pipeline. Cite HubSpot, Ahrefs, and Seer; this is an industry repricing, and arriving with third-party data beats explaining a miss later.
2. Commit to RPS, guide sessions as informational only. Move sessions out of the KPI row and into an appendix. The commitment line is revenue per session and pipeline per asset.
3. Show the cost side. Fewer, better clicks mean fewer pieces of content chasing volume keywords. Content production budgets typically fall 20–30% in the rebuild, which partially self-funds the AI-visibility work.
4. Set the AI answer share baseline now. It will be the Q3 conversation regardless. Better to own the metric before the board asks for it.

The forecasting discipline itself carries over from demand planning: segment, weight by intent, model the mix shift. The same logic drives SaaS sales forecasting with AI audience segmentation, and the ecommerce version of the math is in AI customer insights for sales forecasting.

Retire these KPIs this quarter

The replacements are all computable from data you already have, plus one new tracking layer for AI citations. Our AI search playbook covers the instrumentation end to end, from citation tracking to the correlation model that turns answer share into a pipeline input.

The memo line for the board slide: traffic was never the asset. It was a proxy for buyer attention, and the proxy broke. The attention is still there, it converts 2–4x better on arrival, and the teams that reprice their forecasts this quarter will spend the next four explaining growth instead of variance.

Sources

• Ahrefs: AI Overviews reduce clicks, December 2025 update
• Seer Interactive: AIO impact on Google CTR, 2026 update