Articles
Resource library
Ahrefs’ AI Share of Voice blends observed brand appearances with modeled demand estimates. Learn what its formula captures, where the uncertainty enters, and how to report it beside direct visibility tests.
AI Training Is Not AI Search: What the New Robots.txt Controls MeanA practical guide to the shared AI-training signal and the separate crawling, search and grounding controls announced by OpenAI, Google and Microsoft.
ChatGPT Ads Expand to 26 More Countries: How to Measure Paid and Earned VisibilityOpenAI has expanded ChatGPT Ads to 26 more countries. Learn who can advertise, who may see ads, and how to measure paid placements separately from organic mentions and citations.
ChatGPT Ads Readiness Checklist for US and Canadian BusinessesA practical ChatGPT Ads checklist for US and Canadian businesses: confirm advertiser eligibility, define an outcome, prepare the landing page and measurement, and set a test budget you can evaluate.
ChatGPT Wallet & Agent Access: How to Prepare Your Website for AI-Led Leads and PurchasesChatGPT Wallet is OpenAI’s payment feature for user-approved purchases with a tokenized credential. ChatGPT can recommend a business while its agent still fails to use the website. Learn how signed access, forms, checkout and merchant confirmation work, and why agent readiness is separate from search rankings.
ChatGPT’s Missing UTM Tag: What the October 2026 Referral Dip MeansA reported ChatGPT link-tagging bug may have shifted some October referrals into Direct. Here is what the evidence shows, what it cannot prove, and how to audit analytics without overstating the impact.
Gemini 4 Argon in AI Mode: How to Measure Search VisibilityGoogle opened Gemini 4 Argon to AI Pro and Ultra subscribers in Gemini and AI Mode. Learn how to test each surface separately and track whether visibility changes.
How AI Search Data Voids Let a Single Page Shape an AnswerA 2026 experiment found that one attacker-controlled page could enter answers from web-searching AI systems in a substantial share of synthetic tests. Here is what the study means—and what it does not.