Blog
Strategies and research for the new era of AI-powered discovery.
AI Search Strategy
Article age is a weak reason to rewrite a page. Here is a method and a copyable register for ranking refreshes by outdated facts, source changes, citation trends and buyer importance.
15 min read
AI Search Strategy
Pricing, features, support and announcement pages drift apart. A manual reconciliation register shows which statements conflict, which page wins, and what to fix before AI engines repeat the mess.
13 min read
AI Search Strategy
Split quarterly AI search goals into deliverables you control and outcomes you only observe, then pin each to a population, baseline and decision rule. A copyable goal sheet included.
16 min read
AI Search Strategy
Separate what AI engines say about you, what visits you can measure, and what you cannot, then use a copyable decision memo to ask for a funded first program.
13 min read
AI Search Strategy
A null result is still a finding. Learn how to write up an AI search experiment that showed no clear change: hypothesis, method, uncertainty and learning, with a copyable template.
14 min read
AI Search Strategy
Assign who verifies claims, who checks the evidence, who resolves disagreements and who approves. A copyable review rubric and sign-off log for AI-drafted marketing content.
15 min read
AI Search Strategy
Every product, market, persona and language multiplies your tracked prompts. A slot-ledger worksheet shows how to spend a fixed prompt budget on the buying decisions that matter most.
14 min read
Technical GEO
Decide which product facts belong on a public page and which stay behind login. A decision sheet scores each fact by buyer need and risk, with a worked example.
14 min read
Technical GEO
Some crawlers read only the raw HTML your server sends. A five-step check compares that HTML with the rendered page, so you know what a non-rendering crawler can see.
13 min read
Technical GEO
A robots.txt that allows AI crawlers does not mean your CDN or firewall does. Use status codes, verified identities and a troubleshooting log to find out, without disabling protections.
16 min read
AI Engines
A saved AI answer with no search on record can mean three different things. Learn to tell an observed no-search from missing telemetry, with reporting language for each.
15 min read
AI Engines
Adding “for a small team” to a tracked prompt can quietly turn it into a different question. A decision tree separates harmless rewording from a new constraint or audience.
15 min read
AI Engines
A link in an AI answer can be a source, a brand's own site, or a product card. Code each one separately, so your report says what the answer actually showed.
14 min read
AI Engines
A copyable change log for the moments an AI engine, its collection method or your own setup changes, so you know which before-and-after comparisons need a note.
17 min read
AI Engines
Gemini (app) and Gemini (API) can answer the same prompt differently. Align the question, date and collection context first, then compare answer evidence with a worksheet.
15 min read
AI Engines
You asked ChatGPT and saw your brand. Your monitoring report says otherwise. Record seven variables, then compare like with like using this discrepancy log.
14 min read
Agency & Team Operations
Turn four weekly AI visibility reports into one page leadership can act on: decisions, open uncertainties and resource requests, with the weekly detail kept as an appendix.
14 min read
Agency & Team Operations
A quarterly review deck outline for AI search work: what was done, what was observed, what is unresolved and what to choose next, with an evidence checklist.
15 min read
See how AI search engines talk about your brand.
Free to start. No credit card required.