How do you model expected revenue per query without first-party conversion data per query?
Three steps: map queries to landing pages via Search Console and analytics, take URL-level conversion rate and AOV from an analytics/CRM join, then project back to query level with Bayesian smoothing. Paid conversions for the same query provide a calibration factor. As a rule of thumb, forecasts become reliable from about 10,000 monthly sessions per cluster.
When is keyword cannibalization acceptable, and when does it hurt?
Acceptable when a head term serves two distinct intents and both pages rank in different SERP features. Harmful when the intent is the same, because signals split between two URLs. Run a SERP overlap analysis over 30 days: above roughly 70 % URL switching, consolidate; below 40 % with different SERP features, two URLs are fine.
How do you handle queries dominated by AI Overviews and the loss of clicks?
Three paths. Informational queries: optimize passage-level citability so you are cited in the overview, and track it with AAR. Transactional queries with an AI Overview on top: structured snippets, product markup, shopping feeds. Purely informational queries without purchase intent: check the ROI critically. Citations and high-intent clicks now count for more than ranking positions alone.
Which internal linking pattern fits when: siloed, hub and spoke, or interlinked?
Hub and spoke for authoritative knowledge clusters (one pillar plus 6-15 sub-pages, linked both ways). Flat silos for e-commerce categories, which avoids diluting PageRank. Interlinked for evergreen knowledge bases. Mixing patterns within the same site section weakens topical authority and crawl prioritization.
How do you weigh refreshing old URLs against creating new ones on a limited budget?
In our experience a refresh usually beats a new page for URLs that are at least six months old, have five or more referring domains and still fit the topic, because historical signals and link equity are kept. Create new pages only for a fundamental intent shift, a new sub-topic without an existing URL, or a change in URL strategy.
What does sustainable traffic generation mean in 2026?
Intent-driven reach across the customer journey instead of ranking for volume: topical authority over one-off articles, hub-and-spoke architecture over isolated URLs, and SEO, paid search, AI Overviews and zero-click surfaces orchestrated together. The metric is pipeline contribution per session.
How does the hub-and-spoke model build topical authority?
A pillar (hub) covers the main topic broadly and links both ways to 6-15 spoke pages that go deep on sub-topics. Search engines read this as coherent authority, not a pile of articles. It needs entity-based cluster definitions, controlled cannibalization and a clear internal authority flow. In our experience it takes effect from about 8-12 spokes per hub.
How do you deal with zero-click searches and AI Overview traffic?
Value shifts from clicks to citations and high-intent conversions. Build passage-level citability, use structured data for entity consistency (Google requires no special AI markup) and focus organic clicks on transactional queries. Track AAR and branded-search lift.
How do you use paid search to complement organic traffic generation?
As a tool for measurement and acceleration, not competition. Paid delivers conversion data per query that steers organic priorities, better landing pages lift the landing-page part of Quality Score, and joint bidding rules prevent brand cannibalization. A joint playbook with a shared keyword map is essential.
Which traffic KPIs matter beyond sessions?
Revenue per session per cluster, commercial-intent share, conversion fit per URL, citation rate in AI answers, topical authority against competitors, click depth and time to pipeline conversion. Report per cluster, not per page, since that reflects authority and intent coverage.
