SEO Metrics Beyond Pageviews: Agency Performance KPIs

Pageviews no longer prove SEO value. Learn the metrics we use to evaluate agency performance, from qualified sessions to pipeline and AI citation share.

## The dashboard that lied A UK B2B software client came to us last spring with a familiar problem. Their organic traffic had climbed 40% year over year. Their board deck said so. Their agency said so. Then their head of sales pulled the CRM and asked a simple question: why did demo requests stay flat? The answer took an afternoon to find. Most of the new sessions were landing on two glossary pages, ranking for broad informational queries, and bouncing in under 20 seconds. The pages that actually drove pipeline, comparison pages, integration pages, pricing-adjacent content, had lost rankings in the previous core update and never recovered. The traffic chart looked great. The business did not. Google's recent core updates have made this pattern harder to ignore. Helpful content systems now reward pages that satisfy a real intent, not pages that accumulate clicks. For anyone evaluating an SEO agency in 2026, that shift changes the scoreboard. Pageviews are a lagging vanity number. Here is what we would put in their place. ## Why pageviews broke as a performance metric Pageviews were a reasonable proxy when search was a volume game. Rank for a head term, collect clicks, monetize with ads or a broad funnel. That world is gone for most B2B and ecommerce teams. Three things killed the metric: **AI-generated answers absorb informational clicks.** A user asking "what is a data clean room" now gets a paragraph in the results page. They never visit. Your pageview count drops even though your brand may still be cited. **Search intent has fragmented.** The same query means different things to a CTO, a procurement lead, and a developer. A pageview tells you nothing about which one arrived. **Attribution got better, which exposed the gap.** Once you can trace a session to a closed deal, traffic without conversion looks like what it is: noise. None of this means traffic is irrelevant. It means traffic is an input, not an outcome. Judge an agency on what the traffic produces. ## The metrics we actually track now ### 1. Qualified organic sessions Raw sessions count everyone. Qualified sessions count the people who match your ICP, measured by firmographic enrichment (Clearbit, RB2B, or your CRM's reverse-IP layer) or by behavior signals like pricing page visits and multi-page depth. A realistic benchmark for B2B SaaS: 15 to 30% of organic sessions should qualify. If your agency reports 100,000 sessions and 4% qualify, you are paying for reach you cannot sell to. ### 2. Pipeline influenced per content cluster Stop reporting traffic by page. Report pipeline by topic cluster. Group pages into themes (integrations, comparisons, use cases, technical docs) and map closed-won revenue back to the cluster where the first touch happened. This is the single most useful view we build for clients. It usually reveals that 20% of clusters drive 80% of revenue, and that the agency has been optimizing the wrong half. ### 3. Share of AI citations If buyers are getting answers without clicking, you need to know whether your brand is the one being quoted. Tools like Profound, Peec AI, and Otterly track how often your domain appears in AI Overviews, [ChatGPT](https://chat.openai.com/) responses, and [Perplexity](https://www.perplexity.ai/) answers for your target queries. This is early, and the tooling is imperfect. Track it anyway. In 18 months it will be as standard as rank tracking. ### 4. Assisted conversion rate by intent tier Split your keyword set into three tiers: informational, commercial, transactional. Then measure conversion rate separately for each. | Intent tier | Typical session share | Typical conversion rate | |---|---|---| | Informational | 60 to 75% | 0.1 to 0.5% | | Commercial | 15 to 25% | 1.5 to 4% | | Transactional | 5 to 15% | 4 to 12% | If an agency grows informational sessions while commercial share shrinks, they are padding the top of the funnel. Ask why. ### 5. Branded search volume growth This is the quiet one. When SEO works, people start searching for your name. Branded query volume in Google Search Console and Google Trends is a durable signal that content and PR are compounding. We like to see branded impressions grow 10 to 20% quarter over quarter in a healthy program. Flat branded search plus rising non-branded traffic usually means you are renting attention, not earning it. ### 6. Cost per qualified session vs. paid Your SEO agency retainer is a customer acquisition cost. Divide it by qualified sessions and compare against your paid search CPC for the same intent tier. If SEO costs $40 per qualified session and paid costs $25, something is wrong with the content strategy or the targeting. If SEO costs $8 and paid costs $25, you have a compounding asset. Report it that way in the board deck. ## What we recommend If you are briefing an agency or rebuilding in-house, here is our take on the stack. - **Analytics:** GA4 for baseline behavior, but pipe everything into a warehouse (BigQuery or Snowflake) and join it to CRM data. GA4 alone will not show you pipeline. - **Attribution:** HockeyStack or Dreamdata for B2B multi-touch. For ecommerce, Triple Whale or Northbeam. Self-serve GA4 attribution models are not worth defending in a QBR. - **AI visibility:** Profound for share-of-voice in AI answers. Pair it with manual spot checks monthly; the tools miss things. - **Rank and intent tracking:** Ahrefs or Semrush for keyword tiers, plus a custom intent classifier you build once and reuse. - **Reporting cadence:** Monthly on qualified sessions and pipeline, quarterly on cluster-level ROI. Kill the weekly pageview email. What to hire an agency for: technical SEO at scale, content production in a niche you do not have writers for, and link acquisition where relationships matter. What to keep in-house: intent classification, CRM joins, and the definition of a qualified lead. Those are strategy, not execution. ## How to rewrite the agency contract Most SEO statements of work still promise traffic growth. Ours do not. We write them around three commitments: 1. Qualified organic sessions in target ICP segments. 2. Pipeline influenced by priority clusters, measured at 90 and 180 days. 3. Share of AI citations for a defined query set. Traffic targets can stay in the appendix as a diagnostic. They should not be the headline. If an agency pushes back on this, ask them what they would measure if pageviews disappeared tomorrow. The answer tells you whether they sell rankings or revenue. ## FAQ ### Are pageviews completely useless? No. They are a diagnostic, not a KPI. Use them to spot broken pages, cannibalization, and sudden ranking losses. Do not use them to justify a retainer. ### How long before these metrics show movement? Qualified sessions and intent mix shift within 60 to 90 days. Pipeline attribution needs two full quarters to be statistically useful. AI citation share moves fastest of all, often within weeks of publishing. ### What if our CRM data is a mess? Fix that before you hire anyone. An agency cannot attribute pipeline you cannot track. Start with UTM discipline, a single source of truth for lead status, and a reverse-IP tool on the site. That foundation is worth more than any content sprint.

SEO Metrics Beyond Pageviews: Agency Performance KPIs

Frequently asked questions

Related articles