Viclaro / Blog

Viclaro field notes

How AI assistants decide who to recommend.

Research from Viclaro Atlas on AI recommendations, prompt shape, and the website fixes that make service firms easier for ChatGPT, Claude, Gemini, and Perplexity to cite.

GEO tool comparison 15 min read

The 14 GEO Tools We Compared Against Viclaro, Ranked by Whether They Publish Their Methodology

Every GEO dashboard shows a chart. Almost none of them shows the data behind the chart. This comparison ranks 14 leading tools plus Viclaro on a single axis: can a stranger, without a login, inspect the prompts, responses, citations, and version history behind the numbers. Fourteen dashboards. Zero fully public methodologies. One awkward finding.

Read post
Recommendation diversity 5 min read

Which AI Assistant Has the Broadest Local Shortlist?

Some assistants repeat a compact set of familiar providers. Others distribute extracted firm citations across a long tail. In reconciled NYC cosmetic-surgery data, Perplexity named 296 resolved firms while GPT-4o named 69—but breadth is not the same as quality or fairness.

Read post
Model consensus 5 min read

The AI Recommendation Consensus Gap: High Visibility Can Depend on One Model

A firm can accumulate recommendations on one assistant and remain absent from the other three. Aggregate share measures volume; model consensus measures breadth. Across eight NYC markets, those two signals repeatedly tell different stories.

Read post
Buyer journey research 6 min read

Where Firms Disappear from AI Answers: 7,637 Responses Across the Buyer Journey

AI assistants do not name providers at the same rate throughout a buyer’s decision. Across 7,637 reconciled Atlas responses, decision prompts led four case-driven legal categories, while estate planning and medical markets followed different patterns.

Read post
Atlas model comparison 10 min read

ChatGPT, Claude, Gemini, and Perplexity Do Not Recommend the Same Local Leaders

A blended AI visibility ranking can look like one market. Underneath it are four different shortlists. After reconciling the live extraction corpus across eight NYC categories, none produced the same model-level leader on ChatGPT, Claude, Gemini, and Perplexity.

Read post
AI search measurement 9 min read

Google and Bing Now Report AI Visibility. Here Is What Their Dashboards Still Cannot Tell You.

Google and Bing have finally started separating AI search visibility from ordinary search reporting. That is a major improvement. It still does not answer the commercial question most service firms care about: when a buyer asks who to hire, did the assistant recommend you?

Read post
Atlas index 6 min read

The 2026 NYC AI Recommendation Index: 13,368 Extracted Firm Citations Across 8 Markets

Viclaro reconciled 13,368 ranking-eligible firm citations across 3,172 ranked firm records in eight NYC service markets. The result is not one AI search economy: category concentration, model consensus, and shortlist depth differ too much for one universal benchmark.

Read post
Source provenance 7 min read

Most AI Local Recommendations Do Not Show Their Work

A recommendation can name a firm without displaying a source. In the latest eight-category Atlas scan corpus, only 63 of 7,637 raw responses contained an explicit URL. Source analysis based only on visible links therefore observes a small, non-random slice of the evidence system.

Read post
Atlas benchmarks 10 min read

AI Visibility Benchmarks by Industry: What 8 NYC Markets Show in 2026

There is no universal “good” AI share of voice. After reconciling live extraction across eight NYC markets, top-five concentration ranges from 23.6% to 57.1%, while top-ten four-model coverage ranges from two firms to nine.

Read post
AI search strategy 9 min read

Query Fan-Out Changes AI SEO: One Buyer Question Is Now a Search Portfolio

AI Mode does not retrieve against one neat keyword. Google says it fans a question into multiple searches, and AI Mode queries now average three times the length of traditional searches. The content unit for GEO is becoming a portfolio of verifiable answers.

Read post
Product validation 8 min read

Atlas Rankings Track Viclaro Audit Visibility

Looking back across the ten most visible NYC divorce firms in Atlas, Viclaro found a strong match between Atlas visibility and audit visibility. The value is not just the correlation; it is that the audit turns the Atlas list into an artifact a firm can act on.

Read post
Measurement 8 min read

AI Mentions vs. Citations: One Gets You Recommended. The Other Gets a Footnote.

A firm can be named without its website being cited. A website can be cited without the firm being recommended. Collapse those events into one “visibility” score and the dashboard becomes impressively precise about the wrong thing.

Read post
Prompt methodology 9 min read

How to Build an AI Visibility Prompt Set That Is Not Just Your Keyword List in a Fake Mustache

A useful prompt set samples how people choose, not how marketers label pages. Start with buyer situations, vary the constraints and wording, keep branded prompts out of the score, and version the panel so a trend means something.

Read post
Research design 9 min read

Why AI Recommendations Change Between Runs (and When a Change Is Real)

An assistant is not a laminated directory. Search results move, generated wording varies, and a tiny prompt edit can change the shortlist. The answer is not to give up on measurement. It is to stop treating one answer like a closing stock price.

Read post
AI search rankings 9 min read

How to Read AI Search Rankings (Beyond Vanity Metrics)

AI search rankings measure how often assistants like ChatGPT, Claude, Gemini, and Perplexity name a business when a buyer asks who to hire. Read for share of voice, model coverage, and the gap between positions — not the rank number alone.

Read post
AI SEO tools 10 min read

The State of AI SEO Tools in 2026: What Actually Works

AI SEO tools measure how brands show up inside AI assistants like ChatGPT, Claude, Gemini, and Perplexity. In 2026 the mature ones can reliably measure per-model citation rate, share of answer, and category shape. They still cannot measure assistant usage share, conversion attribution, or click-through when no URL is returned. This guide walks through what the category can honestly do, what it cannot, and how to evaluate a tool.

Read post
AI search optimization 9 min read

How to Optimize Your Site for AI Search: A Step-by-Step Playbook

Optimizing for AI search means increasing how often AI assistants like ChatGPT, Claude, Gemini, and Perplexity name your brand in their answers. The workflow is a loop: measure your baseline citation rate on a defined prompt panel, diagnose which buyer prompts you lose and on which models, ship targeted content changes with structured data, and re-measure to prove the delta. This piece walks through each step with the numbers, sample sizes, and pitfalls that separate a real result from a lucky snapshot.

Read post
AI visibility metric 9 min read

AI Share of Voice: The Metric That Replaces Market Share in AI Search

AI share of voice is the portion of AI-generated recommendations in a category that name your brand. It is measured across a fixed prompt library, a fixed panel of assistants, and a fixed sampling protocol. It is not market share — a firm can be a category leader by revenue and hold near-zero AI share of voice, or vice versa. A good number depends entirely on the shape of the category: a monopoly might hit 60%, a competitive leader 20%, a fragmented market 8%. This piece defines the metric, walks through how to calculate it, and shows what different share levels actually mean.

Read post
Measurement 8 min read

How to Measure AI Visibility (Without Fooling Yourself with One Screenshot)

To measure AI visibility, you need a panel — multiple assistants, multiple samples per prompt, per-model breakdowns preserved. Not a screenshot. Not a single-model score. The four major assistants disagree with each other by 5x or more on the same firm in the same category, so any measurement built on one model at one moment is roughly that far off from the actual market. This piece walks through the sample size, model coverage, and reporting design that make AI visibility measurement defensible.

Read post
GEO fundamentals 6 min read

Why Schema and llms.txt Will Not Fix Vague Positioning

Technical signals can make evidence easier to discover and parse. They cannot manufacture the evidence an assistant needs to recommend your firm.

Read post
Personal injury 8 min read

AI SEO for Personal Injury Lawyers: Why the Market Is Still Open

AI SEO for personal injury lawyers looks different from AI SEO in other legal verticals because the market is still fragmented. In our NYC audit, the top PI firm holds only 8.0% share of AI recommendations, only two of four major assistants cite it, and the entire corpus is 15x smaller than NYC divorce. That combination — no dominant firm, partial model coverage, thin overall citations — describes an unusually cheap window for a mid-market or boutique firm to break into the top five before consolidation closes it.

Read post
AI recommendations 10 min read

How ChatGPT (and Other AI Assistants) Choose Which Businesses to Recommend

ChatGPT and other AI assistants pick businesses by combining training data, live web retrieval, and their own citation preferences. The result is not one answer but a probability distribution across a handful of firms, and it varies by assistant. Here is how the mechanism works, how to check where your business sits, and why the answer differs across models.

Read post
GEO vs SEO 10 min read

GEO vs SEO: What Changes When Buyers Ask AI Instead of Google

Generative engine optimization (GEO) is the practice of getting a business named by AI assistants like ChatGPT, Claude, Gemini, and Perplexity. It shares foundations with SEO — authority, structure, evidence — but the retrieval unit is a citation in a generated answer, not a link on a SERP. Here is what actually changes.

Read post
AI-friendly content 9 min read

How to Write Content That AI Assistants Actually Cite

Content that AI assistants cite looks different from content designed for humans or Google. It answers the buyer's exact question in a single quotable sentence, in the buyer's vocabulary, marked up with FAQPage schema, and reachable without a login. Get those four things right and ChatGPT, Claude, Gemini, and Perplexity can lift your page verbatim into a recommendation. Miss them and even a well-written site stays invisible.

Read post
Legal AI SEO 9 min read

How ChatGPT Recommends Divorce Lawyers: What We Learned Auditing NYC

ChatGPT recommends divorce lawyers by retrieving named firms from a shortlist that its training data and retrieval layer have built up over months of web crawling — not from lawyer directories, and not from Google search rankings. In NYC divorce, three firms hold 53.5% of every AI recommendation. This piece walks through how AI actually decides which lawyer to name, using data from 7,485 citations across four assistants, and what a firm outside the top ten can do about it.

Read post
AI search strategy 7 min read

Why Your Law Firm Can Rank in Google and Still Be Invisible to ChatGPT

Google rankings reward pages built for keywords. AI recommendations reward pages that answer a buyer situation clearly enough to be reused in an answer.

Read post