A line of inquiry
Discovery Science
How buyers find a business now, across search and AI answers.
Discovery has split into three surfaces: classic search, the local map pack, and the AI answers that increasingly stand in for a list of links. This line explains the mechanics of each, how an engine decides whom to name, and what genuinely earns a place in the answer rather than a place in the noise.
39 pieces in this line, across 32 threads.
Thread
The infrastructure honesty gap
Crawlability Is Not One Setting: A Field Guide to the Three Kinds of AI Bots
Training crawlers, retrieval crawlers, and user-triggered fetch agents visit your site for different reasons and obey the rules differently. Why one robots.txt line cannot govern all three, and how to read the difference.
Read · 11 minE-E-A-T Is Not a Ranking Factor: Reading Google's Rater Guidelines Correctly
A large share of commercial search advice sells E-E-A-T as a score a page can optimize into a document. Google's own documentation says it is a human-rater evaluation heuristic. What the primary sources actually state, and why the difference costs money.
Read · 9 minThe Honesty Audit: Which AI-Visibility Claims Are Actually Backed by Evidence
A meta-review of the confident claims sold in AI search, llms.txt, guaranteed rankings, quantified accuracy jumps, structured-data lifts, tested one at a time against the primary sources. Some hold. Some do not.
Read · 11 minThe llms.txt Autopsy: What Happens When 137,000 Sites Actually Get Checked
llms.txt is sold as a way to control how you appear in AI answers. When someone finally measured 137,000 sites, 97 percent of the files had never been read. Here is the evidence, and what the same data says does work.
Read · 10 minThread
The citation gap
Ranking #1 on Google and Still Invisible When Someone Asks ChatGPT
Classic search ranking and AI-answer inclusion are two separate, differently sourced surfaces. For auto repair, the second one now decides a fast-growing share of who-do-I-call decisions, and it is nearly unmeasured by the shops it affects.
Read · 8 minWhy Your Med-Spa Can Rank on Google and Still Be Invisible in ChatGPT
A page-one Google ranking and an AI-answer citation are two loosely overlapping outcomes. In aesthetics specifically, the AI-answer surface is currently dominated by manufacturer brand names, not the practices that deliver the treatment. Here is the evidence, and what it changes.
Read · 9 minYour Patients Are Already Asking ChatGPT Before They Ask You
AI-chatbot use for health information roughly doubled in a year. This is not a hypothetical future channel for a medical practice, it is a present, fast-growing one, and it runs on the same entity-consistency work that already decides the map pack.
Read · 9 minThread
How India's Local Businesses Get Found
The Entity Void: The Knowledge Graph Barely Knows These Businesses Exist
Raveneye Global queried Google's Knowledge Graph for 48 businesses already leading their local Indian markets. One resolved to a matching entity, and it belonged to a national brand, not a local firm.
Read · 13 minWho Actually Answers a Local Search in India: 48 Queries, One AI Overview
In 48 live local searches across six Indian cities, Google's AI Overview answered once. The map pack answered for three of eight trades, and never for the other five.
Read · 15 minThread
Local ranking mechanics
Primary Category Is the Single Biggest Lever in a Service-Area Google Business Profile
For a contractor with no storefront, the category chosen on a Google Business Profile does more work than review count or post frequency. Here is why, and what most service-area profiles get wrong.
Read · 6 minThe Google Business Profile Field Most Auto Shops Set Once and Never Touch Again
Primary category is the single strongest lever on a Google Business Profile, and it is also the field most shop owners set at claim time and never revisit. Here is why that one field caps everything else a shop tries to fix.
Read · 6 minThread
Category language and definitions
Thread
Entities and the machine-readable web
Thread
Entity and knowledge-graph visibility
Thread
Entity identity and the knowledge graph
Thread
Entity resolution and the visibility gap
Thread
How agentic commerce actually selects a product
Thread
Local discovery, decided before the click
Thread
Technical foundation and machine trust
Thread
Technical foundation and structured data
Original research
Studies from this line of inquiry
Machine-Readiness Audit: Africa, July 2026
The first Africa edition of a quarterly, multi-region benchmark. We sampled 11,582 independent African small businesses from the map. 41.4% had a real website; the rest reach customers through social and messaging, not a machine-readable page. About 4.5% of all map-listed African businesses are fully machine-readable.
41.4% of 11,582 map-listed independent African businesses had a real website (4,793) Read the study · 20 min ResearchMachine-Readiness Audit: Europe, July 2026
The first Europe edition of a quarterly, multi-region benchmark. We sampled 52,076 independent European small businesses from the map. 76.3% had a real website, and among those, 77.9% had no type an engine can read as a local business. About 12.7% of all map-listed European businesses are fully machine-readable.
76.3% of 52,076 map-listed independent European businesses had a real website (39,722) Read the study · 20 min ResearchMachine-Readiness Audit: India, July 2026
The truer picture of the Indian market. We measured 110,312 independent Indian businesses in all, the majority across the twenty-two trades that actually make up the high street, kirana grocers, tailors, and more. Only 18.4% of those had a real website, and just 1.2% are fully machine-readable, far below what the Western-professional trades suggest.
18.4% of 66,452 map-listed independent Indian businesses had a real website (12,227) Read the study · 20 min ResearchMachine-Readiness Audit: Japan, July 2026
The truer picture of the Japanese market. We measured 60,059 independent Japanese businesses in all, the majority across the eighteen trades that actually make up the high street, izakaya and ramen counters, soba and udon shops, and more. Only 43.4% of those had a real website, and just 2.3% are fully machine-readable, far below what the Western-professional trades suggest.
43.4% of 34,831 map-listed independent Japanese businesses had a real website (15,107) Read the study · 20 min ResearchMachine-Readiness Audit: Southeast Asia, July 2026
The truer picture of the Southeast Asian market. We measured 55,020 independent Southeast Asian businesses in all, the majority across the eighteen trades that actually make up the high street, warung and warteg eateries, sari-sari and toko kelontong grocers, and more. Only 11.4% of those had a real website, and just 0.9% are fully machine-readable, far below what the Western-professional trades suggest.
11.4% of 33,664 map-listed independent Southeast Asian businesses had a real website (3,840) Read the study · 20 min ResearchMachine-Readiness Audit: US MSMEs, July 2026
The first edition of a quarterly benchmark. We measured 43,362 independent US small businesses; one in four was invisible to a machine reader, a majority had no type an engine can read as a local business, and readiness rose steadily with how established the business was. We will repeat it every quarter, using the same method, to track how the machine-readable web evolves.
43,362 independent small-business homepages read, drawn from Google Business Profile listings across 12 trades and 88 US metros Read the study · 24 min ResearchThe Machine-Readiness Atlas: How the World's Small Businesses Compare
We ran the same audit across six economies, 205,103 independent small businesses in all. The share a machine can fully read ranges about 7.6-fold, from 32.1% in the US to 4.2% in India, and each economy fails in its own way.
205,103 independent small businesses measured across six economies with one fixed method, July 2026 Read the study · 18 minKeep reading
The other lines of inquiry
The Macro Shift
The once-in-a-generation change in how people find a business.
The Attention Landscape
Where a market’s attention actually sits, and how it keeps moving.
Choice Science
Why a buyer picks one business over another once both are found.
Conversion Science
Turning a visit you already earned into a booked customer.
Demand & Paid Media
Buying attention accountably, when you need customers this quarter.
AI Operations
Putting the routine work of a business on evaluated AI systems.
Vertical Playbooks
How each kind of business gets found and chosen, on its own terms.
MSME & Global Commerce
The economics of the small and mid-size business online.
Trust, Ethics & Regulation
The line between persuasion and manipulation, and the rules that hold it.
Measurement & Honesty
Measuring visibility truthfully, and refusing the numbers that lie.
See where you stand in your own market.
A specialist-reviewed read of your visibility across search and AI answers, for your business and your market. No guaranteed number, and no obligation.