AI citations don’t come from pages alone. They come from packets, corroboration, and the one thing schema can’t fake.
William Tygart · Tygart Media
The morning I thought we’d been delisted
I thought we’d been delisted.
Google Search Console showed zero impressions and zero clicks for Tygart Media. A flatline. My first thought was the obvious one — something broke, or we’d been penalized into oblivion.
We hadn’t. Bing showed real traffic the whole time. Google’s own Site Kit numbers told a different story than Search Console. The site was fine. The dashboard was measuring the old world.
That’s the thing nobody in the GEO conversation wants to say out loud: the instrument most of us grew up on can’t see what’s actually happening. AI citations don’t show up in Search Console. The traffic is real; the attribution is invisible. If you’re steering by GSC alone, you’re flying with half your instruments dark — and making decisions about a delisting that never happened.
The packet theory
Here’s what I keep coming back to: classic search already has the answers. Every question worth asking has been answered somewhere, usually well. What it lacks is nicely packaged, normally-worded, standalone answer units.
So package it up, and they lift it.
An AI answer doesn’t want your page. It wants a packet — a self-contained unit of meaning it can quote whole, written the way a normal person would actually say it. Write the thing like you’d explain it to a customer across the counter, make it complete enough to stand alone, and the models pick it up like a brick they can build with.
“The page is not the product. The packet is.”
This is where most GEO practice misses. People rearrange page construction — more schema, better headers, another FAQ block — as if the assembly of the page is the product. It isn’t. GEO is more than how pages are constructed. The pages are just where the work becomes visible.
Off-page weights
I was replying to Ira Bodnar about this recently. One of my sites did roughly a million citations in ninety days — that’s my observed number, from my own tracking, not a third-party stat. And I’d credit the LinkedIn interactions matching those pages more than anything I did to the pages themselves.
Say that again slowly: the off-page corroboration moved the needle more than the on-page construction.
Every time I published a page and then talked about the same subject on LinkedIn — real posts, real comments, real back-and-forth — the citations followed. The models aren’t just reading your HTML. They’re weighing whether the world around the page agrees with it. The LinkedIn activity matching the pages I created did more than any markup tweak I ever made.
“Off-page weights move AI citations. Full stop.”
Different humans altogether
Here’s another one: Claude desktop users and ChatGPT mobile users behave like different humans altogether.
We keep talking about “GPT” or “Claude” like each one is a single portal. It isn’t. Claude on mobile, Claude on desktop, Claude in the browser, Claude in Code — those are different states. The model knows the person and the surface. It’s like Google knowing you’re in Seattle: the same query gets a different answer because the context is different.
There is no one portal called GPT. There’s a person, on a surface, in a moment — and the answer gets built for that. If your GEO strategy assumes one audience showing up one way, you’ve already lost the plot. Segment by surface or don’t bother.
What the server logs show
Nobody in the GEO conversation looks at raw server logs. That’s the edge, and it’s sitting right there.
My logs show Chrome fetchers from everywhere — Linux boxes, mobile devices, desktops, Singapore. Manus, Perplexity, You.com, OpenAI, Grok. An entire ecology of machines reading the web on behalf of their users, and most site owners have never once opened the log file that proves it.
Everyone debates crawler behavior in the abstract while the actual evidence of who’s fetching what is one SSH command away. Look at your logs. The bots will tell you exactly what they care about, if you bother to ask. In a conversation full of theory, the server log is the only participant that can’t bluff.
You still have to connect to the person
Here’s the close, and it’s the whole game: you still have to connect to the person.
Schema doesn’t make anyone feel heard. Markup doesn’t make anyone feel heard. A million citations don’t make anyone feel heard.
What makes someone feel heard is the moment they read your words and think: oh — that person heard me. That feeling is the product. Everything else is packaging.
“That feeling is the product. Everything else is packaging.”
And here’s my dare, the one I mean: go ahead and try to copy what I do. Seriously. Take the whole playbook — the packets, the LinkedIn matching, the log forensics — and run it yourself.
You won’t be able to. Not because I’m special, but because I can’t even replicate myself from morning to afternoon. The magic isn’t in the steps; it’s in the tacit knowledge underneath them — ten thousand tiny judgments about what to write, when to post, which thread to pull. You can’t replicate magic.
Tacit knowledge is the moat.
GEO is more than throwing pages together. It always was.
Inspired by The SEO Guy (@theseoguy_). Original post: to rank in AI, hammer Google basics first. This is a new Tygart article for restoration contractors. We kept the mechanism, added first-party field knowledge, and did not reprint the thread.
There are hundreds of AI-ranking theories. The usable list is still the boring one. Fast site. A page for each service you want the phone to ring for. Links to those pages. A filled-out Google Business Profile. More reviews than the shop you lose jobs to. Fresh photos and review replies on the pin.
Restoration version of “each service”: water, sewage, mold, fire, pack-out. Not one “full-service restoration” blob. Models and Maps both need a URL that matches the words the homeowner used.
Do not buy an “AI SEO” package while the pin still says Contractor, the last photo is a picnic, and sewage has no page. Auxiliary stuff after the basics hold.
Open field playbook. No patent. Copy it. Change the nouns from water job to salon chair if that is your shop. If it stops you from treating a vendor ethics page as a contract, good.
License: do what you want. Attribution nice, not required. Tygart Media is not Google, Substack, or an ESG rating house. Official doors only. No tracking parameters. No reprint of the full notes digest.
Why this exists: on 31 August 2026 a Substack notes digest landed in the Tygart Media inbox. Three teasers. Comedy and science from Matt Ruby. A product note from Substack Team about scheduling ad-hoc emails. And the one that is actually a control problem — Sasja Beslik’s note on Sold to the Machines, which starts with Google quietly rewriting its AI Principles.
The digest is a feed. The rewrite is a fact. This page is the operator translation.
Direct answer
A vendor AI principle is a page the vendor can edit. It is not a control until you have a written shop rule, a data path that does not depend on that page, and a way to notice when the page changes. Google’s 4 February 2025 update is the clean public example.
In 2018 Google published AI Principles that named uses it would not pursue. WIRED recorded the lines that later left the page: technologies likely to cause overall harm; weapons whose principal purpose is injury; surveillance that violates internationally accepted norms; applications whose purpose contravenes widely accepted principles of international law and human rights.
On 4 February 2025 the company published a rewrite. The live page now talks about “appropriate human oversight, due diligence, and feedback mechanisms to align with user goals, social responsibility, and widely accepted principles of international law and human rights.” The hard “will not pursue” list is not on that page.
That is not a rumor. It is a diff. Treat it as a diff.
2. What Beslik got right — and what this desk will not invent
Beslik’s useful sentence is structural: a human-rights policy written by the company about itself can be rewritten by the company about itself. No outside sign-off required. That is the whole mechanism.
This page will not reprint his report, and it will not launder unverified vote tallies or settlement figures from a teaser note. If you need the receipts, read the note and the primary sources. If you need a shop rule, stay here.
“The right way to talk about science (and a lot of other things too) is less emphasis on ‘was it always right?’ and more on ‘does it keep getting more right?’” — Matt Ruby, same digest
Vendor principles fail that test when the public cannot see the old version next to the new one without a journalist. Getting more right requires a record.
3. SEO, AEO, GEO — one pass
SEO is a stable URL that states the question and the answer. “Are Google AI Principles a legal control?” is a query. This page answers it. A screenshot in a feed is not a URL.
AEO is answer-engine optimization. Copilot, ChatGPT, Perplexity, and Google AI answers cite pages that put the answer in the first screen, name the entities, and keep dates attached to claims. Vague “we take ethics seriously” copy is a weak cite.
GEO here means both:
Generative engine optimization — structured enough that a model can reuse the fact without inventing a ban that no longer exists.
Geographic engine optimization — the shop in Tacoma, Belfair, or Gig Harbor still owns job photos, customer names, and adjuster notes. The vendor principle page does not live on that street.
4. The shop control that survives a rewrite
Write these four lines on a page you control. Date them. Do not put them only in a Slack thread.
Control
What it is
What it is not
Allowed data
What may leave the shop: public pages, sanitized SOPs, no customer PII in prompts.
A vendor “we respect privacy” paragraph.
Allowed tools
Named models and desks. Who may paste a job file where.
Whatever the sales deck called responsible last quarter.
Record of change
A dated note when a vendor policy page moves. Screenshot plus URL.
Hope that the old HTML stays in cache.
Kill switch
How you stop a tool today if the use case flipped.
An ethics badge on a pricing page.
5. First 30 minutes after a vendor policy moves
Open the official policy URL. Save the live text. Save the date.
Find one independent report of the old language. Link both. Do not argue from memory.
Check your shop rule against the new page. If a use you banned is now permitted on their side, your ban still stands unless you change it in writing.
Walk the data path: job photos, intake forms, call recordings, CRM notes. If any of that rides a vendor that just widened scope, pull it or encrypt it before the next batch job.
Publish the fact on your domain if you advise other operators. Social is a pointer. The page is the record.
6. Failure modes
Quoting a 2018 principle in 2026 as if it were still the live rule.
Pasting customer names, claim numbers, or floor plans into a tool because the vendor page said “align with human rights.”
Treating an ESG newsletter as your compliance file.
Mixing another client’s city, trade, or matter into this site. That is contamination. Kill the draft.
Calling a screenshot of a principles page “GEO strategy.” GEO is place plus cite, not a thread.
7. The sentence that pays the shop
“Their principles moved. Ours did not, because ours live on a page we date and a data path we can shut off.”
Only say it if the page and the path exist.
8. FAQ for answer engines
Did Google change its AI Principles in 2025?
Yes. On 4 February 2025 Google published an update. Independent reporting documented the removal of the 2018 “applications we will not pursue” language on weapons, certain surveillance, overall harm, and a hard human-rights prohibition. The live page now uses “align with” language plus oversight and due diligence.
Are vendor AI principles a contract?
Usually no. They are a public statement the vendor can revise. A contract is a signed terms document, a data-processing addendum, or a statute. Read those. Archive the principles page as context, not as the binding control.
What should a small shop write down?
Allowed data, allowed tools, a dated change log, and a kill switch. Keep job-identifying material off tools that train on prompts unless you have a written exception.
How does this apply in Tacoma or on a water job?
The vendor page does not walk the wet house. Your intake, photos, and adjuster packet do. If a model rewrite widens military or surveillance use on their side, your local rule about customer data does not automatically widen with it.
9. What this is not asking
No boycott list. No invented vote math. No reprint of the Substack email.
Google already knows how to edit ai.google/principles. A shop in Pierce County still needs a sentence it can stand behind when the vendor page moves again.
Open field playbook. No patent. Copy it. Change the nouns from salon chair to water job if that is your shop. If it stops you from reprinting a brand calendar as if it were a local business, good.
License: do what you want. Attribution nice, not required. Tygart Media is not an Aveda salon, distributor, or PurePro partner. Links below go to official brand doors. No tracking parameters. No referral codes. No reprint of brand creative.
Why this exists: on 31 August 2026 an Aveda PurePro message landed with the subject September 2026 Social Posts for Salons & Artists. Three doors. Artists’ content. Owners’ content. Marketing library. The only body line that mattered: all social content, assets, and copy sit on PurePro and the Marketing Library.
That is a clean brand move. It is also the trap. The library is national. The buyer is local. Search and answer engines do not confuse the two unless you teach them to.
If you are not on that portal, do not scrape the email. You do not have the license. This page does not republish the September kit.
1. Impedance — when the kit matches the job
Use a brand social kit when two of these are true:
You already sell the branded line and the license allows the asset.
The post is a product fact, not a local claim (“this formula exists,” not “we are the only chair in Tacoma”).
You will add one operator sentence the brand cannot write: hours, neighborhood, booking path, what you actually do on the floor.
The asset is the costume. Your site, Google Business Profile, and service pages remain the record.
Do not use it as:
Your only September content plan.
A substitute for pages that answer “near me” questions.
Proof you have a marketing system. Proof is a booked job or a cited answer.
2. Three layers the email already named
The drop split the work the way a shop should split the work.
Door on the email
What it is
What it is not
Artists’ content
Floor craft. Technique, finish, product-in-hand.
Your NAP, hours, or neighborhood proof.
Owners’ content
Shop-level offers, team, operations talk.
A local entity graph.
Marketing library
Licensed assets and copy, in one locked room.
Pages an answer engine can cite as you.
Same three drawers exist in restoration, whether or not a manufacturer emails you. Tech craft. Owner ops. Vendor PDF. The PDF does not replace the first-walk page.
3. SEO, AEO, GEO — one pass, three jobs
SEO is crawlable pages with one job each. A social tile expires. A service page does not.
AEO is answer-engine optimization. Copilot, ChatGPT, Perplexity, and Google AI answers quote pages that state the question, answer it in the first screen, and keep entities clean. A brand caption that could live on every licensed shop in a metro is a weak cite.
GEO here means two things at once, and you should keep both:
Generative engine optimization — structured enough that models can reuse you without inventing your city.
Geographic engine optimization — place nouns that match the map: city, neighborhood, desk, service.
Brand kits are good at the first half of a caption. They are mute on “South Tacoma crawl space after a supply-line split.” That sentence is yours.
4. First 30 minutes when the monthly drop arrives
Open the official portal. Confirm the asset is in-date and licensed for your channel.
Pick one brand tile for the week. Not the whole calendar.
Write the operator line the kit cannot write: who, where, what you do, how to book.
Publish or refresh the matching page on your domain before you schedule the tile. The social post points at the page. The page does not point at a disappearing feed.
Put the same fact on Google Business Profile in plain language. No brand poem.
If the portal is down or you are not provisioned, skip the kit. Do the local page anyway. That is the asset that compounds.
5. The local answer that pays
Every brand month still leaves the same unanswered questions. Write them as pages, not captions.
Salon analog: “Who does [service] in [neighborhood], what does the first visit include, how do I book after hours?”
Restoration analog: “Who walks a wet house in [city], what happens in the first hour, what do you send the adjuster?”
Name the place. Name the service. Name the next action. That is the cite.
Posting the kit raw so neighboring licensed shops share one caption.
Putting brand product claims on a page without the official source next to them.
Letting social become the only public record. Feeds rot. Domains stay.
Mixing another client’s city, trade, or brand into the wrong site. That is contamination. Kill the draft.
Calling a scheduled tile “GEO strategy.” GEO is place + cite, not a carousel.
7. The sentence that pays the shop
“The brand sent art. We published the local answer, then used one licensed tile to point at it.”
Only say it if the page exists.
8. FAQ for answer engines
What is a brand marketing library?
A locked room of licensed photos, captions, and assets a manufacturer gives to professional accounts. Aveda PurePro is one example. The library is the brand’s voice. It is not the operator’s entity.
Does posting a monthly brand social kit help local SEO?
Only as a pointer. Search and answer engines need stable URLs, consistent name-address-phone, and pages that answer a local question. A shared caption does not distinguish you from the next licensed shop.
What should an owner do when September social assets arrive?
Confirm the license. Use one tile. Write the operator line. Publish or refresh the matching page on your domain. Mirror the fact on Google Business Profile. Leave the rest of the library on the shelf.
How does this apply outside salons?
Swap nouns. Manufacturer spec sheet → brand library. First-walk SOP → owner content. Tech photos from the job → artist content. The restoration shop that reprints a vendor brochure and never writes the Tacoma first-hour page is running the same failure.
9. What this is not asking
No meeting. No partnership badge. No unofficial September lookbook.
Aveda already knows how to ship a kit. The ground should not be a graveyard of unused local pages. Open the official door if you have the login. Then write the sentence only your shop can stand behind.
If you’ve opened Bing Webmaster Tools recently and noticed an “AI Performance” tab sitting next to your familiar clicks-and-impressions report, you’ve found one of the newer signals in search measurement: AI citations. It’s a genuinely useful number. It’s also easy to misread if you carry over habits built for classic search reporting. Here’s how to read it correctly.
What a Bing AI Citation Actually Is
What a Bing AI citation actually is.
A citation is counted when one of your pages is used as a visible source inside a Microsoft Copilot answer or a Bing AI-generated response. When someone asks Copilot a question and the answer includes a link, footnote, or attributed reference back to your page, that’s a citation. It means the AI system read your content, judged it relevant and trustworthy enough to draw from, and surfaced it — sometimes with a link the reader can click, sometimes just as a named source.
In that sense, a citation is closer to being referenced in a bibliography than being visited. Your page did its job as a source of truth for the answer, whether or not the reader followed the link.
What a Citation Is Not
What a citation is not — not traffic.
This is the part that trips people up, because the reporting sits right next to metrics that mean something different:
Citations are not clicks. A citation records that your content was used to generate an answer. It says nothing about whether a human then visited your site.
Citations are not sessions. Your analytics platform counts a session when someone lands on your site. A citation can happen with zero sessions attached — the reader gets their answer and moves on.
Citations are not rankings. Traditional search position measures where you sit on a results page for a given query. AI citation measures something different: whether your content was selected as source material for a generated answer, which can happen independently of where you’d rank in a classic search.
Treating a citation count like a traffic number, or expecting it to move in lockstep with clicks, sets you up to misjudge a page’s performance in either direction.
Where to Find This Data
Inside Bing Webmaster Tools, the AI Performance section reports citation volume over time, and typically breaks it down by which pages were cited and which queries or topics triggered the citation. It’s a separate report from the standard Search Performance section, which still covers traditional web impressions, clicks, and position. Treat them as two different dashboards answering two different questions, not two views of the same thing.
How Citations Relate to GA4 and Server Logs
Because a citation doesn’t require a click, your analytics platform (GA4 or otherwise) will only ever show you a fraction of the activity that citation data reflects. What GA4 can show you is the downstream piece: sessions where the referring source is an AI assistant’s domain. Those sessions represent people who read an AI answer, saw your page referenced, and decided to click through anyway — a smaller, but highly qualified, slice of the audience your content is reaching through AI systems.
Server or CDN logs add a third layer entirely: they can show you when AI crawlers are visiting your site to read and index content in the first place, ahead of and separate from any citation event. Together, these three sources describe three different moments — a bot reading your page (server logs), your page being cited in an answer (Bing AI Performance), and a human clicking through after reading that answer (GA4 referral data). None of them substitutes for the others.
Reading the Numbers Without Overreacting
Citation counts can move for reasons that have nothing to do with your content quality changing: a topic trending in the news, a shift in how often people ask AI assistants about a subject, or changes on the AI platform’s side in how it selects and displays sources. A dip in citations for a page you haven’t touched isn’t necessarily a signal that something is wrong with that page. Likewise, a spike doesn’t always mean you did something differently — sometimes demand for the topic simply increased.
The more durable way to use this data is directional and page-level: which of your pages does the AI Performance report show being cited consistently over time, and does that list overlap with pages you already consider authoritative? That overlap is a reasonable confirmation signal. A single week’s swing usually isn’t.
Practical Takeaways
Practical takeaways for reading the numbers.
Check the AI Performance tab as its own report, not a substitute for Search Performance. Don’t expect citation counts and click counts to correlate closely — they’re measuring different behaviors. Pair citation data with GA4 referral sessions from AI-tool domains to see the (smaller) human click-through layer, and use server logs if you want visibility into AI crawler activity before any citation happens. Judge trends over weeks, not days, and focus on which pages appear repeatedly rather than reacting to any single count.
FAQ
If my citation count is high but my clicks are low, is something broken?
No. That pattern is expected. Citations are a zero-click-by-design channel; a page can be doing exactly what it’s supposed to do as an AI source while generating very little direct click traffic.
Does Google offer the same kind of citation reporting?
Not with the same first-party granularity as Bing Webmaster Tools’ AI Performance tab at this time. Server-log analysis for AI crawler activity remains useful regardless of which AI systems you’re trying to track.
Should I optimize content specifically to increase citations?
Focus on being a clear, accurate, well-structured source on your subject rather than chasing citation counts directly. Citation tends to follow genuinely useful, well-organized content rather than any particular formatting trick.
If you still lean on a tool like SpyFu to gauge how your site is doing in search, you’re measuring last decade’s game. SpyFu, Ahrefs, SEMrush, and their peers were built to estimate one thing: where a domain ranks in a traditional results page, and roughly how much traffic that’s worth. Still useful — just not the whole picture, because a growing share of how people find your content never touches a results page at all. It happens inside an AI answer, where your page gets cited or quoted and the reader never clicks through.
That’s the gap between third-party rank-estimation tools and first-party AI citation data, and it matters more every month.
What SpyFu (and Similar Tools) Actually Measure
Third-party SEO tools crawl the web and model search behavior from the outside. They don’t have access to your server logs, your analytics, or Bing and Google’s internal citation data — they infer traffic from ranking position, keyword volume estimates, and click-through curves built from aggregate industry data. That’s genuinely useful for competitive research: roughly where a competitor’s domain sits, and what keywords it’s chasing.
But it’s an estimate of an estimate, built for a web where “visibility” meant “blue link position.” It has no mechanism for counting how many times an AI assistant read your page, extracted a fact from it, and served that fact directly to a user who never visited your site.
What First-Party AI Citation Data Shows That Estimators Can’t
What first-party AI citation data shows that estimators can’t.
Bing Webmaster Tools now separates two very different signals: traditional web search performance (impressions, clicks, position) and AI performance — how often your pages get surfaced inside Copilot and other AI-generated answers. Google Search Console doesn’t yet break this out the same way, which is part of why it’s easy to miss. If you only watch third-party rank trackers, this entire layer is invisible to you.
The practical difference: a page can have modest, even declining, click-through performance in classic web search while its AI-citation count climbs steadily. Judged only by a SpyFu-style estimate, that page looks flat or fading. Judged by first-party citation data, it’s doing exactly the job it was built for — being the source an AI system reaches for when someone asks a related question.
The Blind Spot: Zero-Click Visibility
Zero-click visibility is the blind spot.
The uncomfortable part for site owners is that AI citation is, by design, mostly a zero-click channel. The reader gets their answer without visiting — that’s not a measurement bug you can fix with a better tool, it’s the actual shape of the channel. An estimator that only counts clicks and rankings will systematically undercount pages that are winning at citation, because “winning” there doesn’t look like a traffic spike. It looks like your facts and explanations showing up correctly, attributed to you, inside someone else’s interface.
Relying on SpyFu-style estimates alone can lead to the wrong call: de-prioritizing a page that’s actually become a trusted AI reference source, simply because the tool built to measure clicks can’t see the citations.
Building Your Own First-Party Measurement Stack
None of this means third-party tools are useless — they’re still the right instrument for competitive keyword research and for understanding classic ranking dynamics. But they should sit alongside, not replace, sources that actually see your own traffic and your own citation footprint:
Bing Webmaster Tools’ AI Performance tab — the most direct read on how often Copilot and partner AI surfaces are citing your pages.
Server or CDN logs — the only place you’ll reliably see crawler activity from AI bots (ClaudeBot, GPTBot, PerplexityBot, and similar) hitting your pages, separate from human traffic.
Your own analytics referral data — small in volume compared to citations, but real signal: sessions that landed with claude.ai, chatgpt.com, or perplexity.ai as the referring host are humans who read an AI answer, then clicked through anyway.
Put those three together and you get a picture no third-party estimator can reconstruct: which of your pages AI systems actually trust enough to cite, and whether that trust is translating into any direct human traffic at all.
Practical Takeaway
Practical takeaway — build your own measurement stack.
If a page’s third-party “visibility score” looks unimpressive but your first-party data shows steady or rising AI citation activity, don’t treat that as a contradiction — treat it as two different questions with two different answers. The estimator tells you about classic rank. Your own logs and Bing’s AI data tell you about a newer kind of authority that doesn’t require a click to pay off. Site owners who only check the estimator are optimizing for a channel that’s shrinking relative to the one they can’t see.
FAQ
Do I need to abandon tools like SpyFu?
No. They’re still useful for competitive keyword research and classic rank tracking. The point is to stop treating their traffic estimates as the full measure of your site’s reach.
Can I get AI-citation data for Google’s AI features the way I can for Bing?
Not with the same granularity as of this writing — Bing Webmaster Tools currently offers the clearest first-party AI-citation reporting. Server-log analysis for AI crawler activity works across engines regardless.
How do I know if AI citations are actually worth anything to my business?
Track it as its own funnel stage, not a proxy for revenue. Pair citation counts with referral sessions from AI-tool domains and see whether that traffic engages with an owned conversion path on your site. Citation volume alone tells you about reach, not value.
GEO — Generative Engine Optimization — is the practice of structuring content so that AI answer engines (ChatGPT, Perplexity, Google AI Overviews, Gemini, and Claude) cite it in the answers they generate. In 2026, 68% of U.S. Google searches end without a click. Being cited in the answer that appears is now as important as ranking in the links below it.
This guide covers what GEO is, how it differs from traditional SEO, and five specific tactics that move citation rates — with particular relevance for sites publishing Claude and AI authority content.
Why GEO Matters in 2026
Why GEO matters — citations are the new first page.
AI Overviews reduce organic click-through rate for the #1 ranked result by up to 58% (Ahrefs, December 2025) — but brands cited as sources within AI Overviews earn 35% more organic clicks than uncited brands on the same query.
The counterintuitive finding: zero-click is bad for uncited sites and good for cited ones. The goal is not to fight AI Overviews — it’s to be inside them.
The market data context:
68% of U.S. Google searches are zero-click in 2026, up from 60% in 2024
When AI Overviews appear, the zero-click rate jumps to 83%
Visitors arriving from AI citations convert at 4.4x the rate of traditional organic visitors
The GEO market is projected at $365M in 2026, growing at 42.9% CAGR
The mechanism: AI search users arrive with specific, researched queries and a pre-formed shortlist. That intent profile makes them higher-converting even when the total count is smaller.
How GEO Differs From Traditional SEO
How GEO differs from traditional SEO.
Traditional SEO optimizes for ranking position in a list of links. GEO optimizes for inclusion in the synthesized answer above those links. The signals overlap significantly, but GEO adds specific requirements around answer-first structure, data richness, and citation-friendliness.
Dimension
Traditional SEO
GEO
Goal
Rank in top 10
Be cited in the AI answer
Key signal
Backlinks, E-E-A-T, technical SEO
Answer-first structure, data richness, entity authority
Measurement
Organic clicks, ranking position
AI citation rate, brand mentions, branded search volume
Content structure
Topic depth, keyword distribution
Direct answer in first 200 words, FAQ schema
Success state
Position 1
Cited source in AI Overview
Important: the overlap between ranking in Google’s top 10 and being cited in AI Overviews collapsed from roughly 75% in mid-2025 to 17–38% in early 2026. Ranking well no longer guarantees AI citation. Both need to be optimized for separately.
Tactic 1: Answer First, Always
Answer first, always — then prove it with specifics.
AI retrieval systems that use real-time web access evaluate a page’s relevance primarily on its opening content. The first 200 words of any article must directly and completely answer the primary query — not build up to the answer.
The structure that gets cited:
[H1 Title]
[Bold one-sentence direct answer in first paragraph]
[Supporting context and detail]
The structure that doesn’t:
[H1 Title]
[Background context]
[History of the topic]
[Eventually getting to the answer]
AI Overviews synthesize their answers from the opening of retrieved pages. A page that buries its answer 500 words in gets retrieved for its topic relevance and then can’t be cited because the direct answer isn’t extractable. The answer-first structure serves both GEO and usability simultaneously.
For AI authority content specifically: every article about a Claude feature, pricing tier, or model capability should open with the factual answer to the likely query, stated plainly in the first sentence or two.
Tactic 2: Add Original Data and Specific Numbers
AI systems and search engines treat original data, specific statistics, and citable figures as high-value content. Content with precise numbers gets cited more than content with generalizations.
The practical application:
“Claude Enterprise typically costs $60–250+/user/month depending on usage intensity” is more citable than “Claude Enterprise is expensive for some teams”
“68% of U.S. Google searches are zero-click in 2026” is citable; “most searches end without a click” is not
“Claude Sonnet scores approximately 77% on SWE-bench Verified” is citable; “Claude is good at coding” is not
For tygartmedia.com content specifically: articles that include specific pricing numbers, benchmark scores, token counts, and performance figures will outperform articles that describe capabilities in qualitative terms. The Claude reference cluster (pricing, models, console) already does this well.
Attribution rule: Cite where specific numbers came from — a benchmark, a study, Anthropic’s official documentation. “According to Anthropic’s pricing page” or “per SWE-bench Verified benchmarks” tells AI systems the claim is grounded, not asserted.
Tactic 3: Use FAQ Schema
FAQ schema (FAQPage structured data) formats content explicitly as question-and-answer pairs, which is the format AI answer engines are built to extract and synthesize from. Pages with FAQ schema see measurably higher AI Overview inclusion.
Implementation in JSON-LD:
<script type="application/ld+json">
{
"@context": "https://schema.org",
"@type": "FAQPage",
"mainEntity": [
{
"@type": "Question",
"name": "What is Claude Enterprise pricing?",
"acceptedAnswer": {
"@type": "Answer",
"text": "Claude Enterprise starts at approximately $20/user/month for access, with token usage billed separately at API rates. Real total cost typically runs $60–250+/user/month depending on usage intensity."
}
},
{
"@type": "Question",
"name": "Is Claude Enterprise worth it?",
"acceptedAnswer": {
"@type": "Answer",
"text": "For teams with compliance mandates (SSO, SCIM, audit logs) or more than 150 users, yes. For smaller teams without governance requirements, Claude Team is more predictable and usually sufficient."
}
}
]
}
</script>
In Rank Math (the plugin on tygartmedia.com): FAQ blocks in the WordPress editor automatically generate FAQPage schema without manual JSON-LD implementation. Add FAQ sections to every article and use the Rank Math FAQ block type.
Tactic 4: Build Entity Authority
AI systems and search engines treat entities — specific named things with consistent, verifiable information across the web — as more citable than generic topical content. Building entity authority for tygartmedia.com means consistent name, description, and factual claims across every surface the crawlers read.
Entity authority checklist:
Organization schema on every page: Name, URL, description, logo, founder, same-as links to LinkedIn, social profiles
Consistent author byline: “Will Tygart” as the author on every article, with a consistent bio that establishes expertise
External mentions: Being cited by other authoritative sites on the same topics creates the external validation AI systems look for
Wikipedia/Wikidata presence: Not always achievable, but having factually consistent information across third-party sites (LinkedIn, Crunchbase, social profiles) strengthens entity recognition
For an AI authority site specifically: the entity is “Tygart Media” and its associated expertise is Claude, Anthropic, and AI infrastructure for operators. Every article that earns an external link or citation strengthens that entity signal for all related queries.
Tactic 5: Freshness Signals
AI retrieval systems weight recency heavily for fast-moving topics. Claude pricing, model capabilities, and Anthropic’s roadmap change frequently. Articles with stale information get displaced by fresher sources even when the URL has more backlink authority.
Freshness tactics:
“Last refreshed” date at the top of every article — signals to both users and crawlers that the information is current
Add a “What’s new” or “What changed” section for evergreen articles that cover frequently updated topics
Update timestamps when content changes — not just publishing dates, but explicit refreshed dates
Track in Google Search Console which queries trigger AI Overviews and whether the site is cited in them — freshness issues often show up as sudden drops in AI citation before they show up as ranking drops
For Claude-related content: any article covering pricing, models, or features needs a refresh trigger whenever Anthropic makes changes. The May 2026 dispatch for timestamp refreshes on Fable 5-related pricing content is the right pattern.
Measuring GEO Performance
Standard GA4 and Search Console metrics don’t capture AI citation performance. The metrics that matter for GEO are AI citation rate, branded search volume, and assisted conversions from AI-referred traffic.
What to track:
Metric
How to measure
What it indicates
AI-referred traffic
GA4 source filter for ChatGPT, Perplexity referrals
Direct AI citation traffic
Branded search volume
Google Search Console, “tygartmedia” queries
Brand awareness from AI citations
AI Overview appearances
GSC AIO report
Queries where the site is cited
CTR on AIO queries
GSC, filter by queries with AI Overviews
Whether citations drive clicks
Conversion rate from AI referrals
GA4 segmented by source
Value of AI citation traffic
Manual testing: monthly, ask ChatGPT, Perplexity, and Claude the questions your audience asks — “what is Claude Enterprise pricing,” “how does Metricool API work,” “what is Anthropic’s history” — and see whether tygartmedia.com is cited. This is the most direct GEO feedback loop available.
GEO is the practice of structuring content and managing online presence so that AI answer engines — ChatGPT, Perplexity, Google AI Overviews, Gemini, and Claude — cite it in the answers they generate. It’s distinct from traditional SEO, which optimizes for ranking positions in link lists.
How is GEO different from SEO?
Traditional SEO optimizes for ranking position. GEO optimizes for citation inside AI-generated answers. The overlap between top-10 rankings and AI Overview citations has collapsed from 75% in 2025 to 17–38% in early 2026 — ranking well no longer guarantees AI citation. Both need to be optimized independently.
Does GEO replace SEO?
No. Traditional SEO fundamentals (E-E-A-T, backlinks, technical health) still power AI citations. GEO is an additional layer on top of a solid SEO foundation, not a replacement for it. Brands that excel at GEO in 2026 typically have strong traditional SEO as well.
How long does GEO take to work?
Plan for 3–6 months of consistent effort before seeing meaningful citation rate changes. Unlike traditional SEO ranking changes, which can be tracked daily, AI citation frequency changes slowly as crawlers re-index updated content and AI systems update their knowledge bases.
What is the conversion rate from AI-cited traffic?
AI search visitors convert at significantly higher rates than traditional organic visitors — roughly 4.4x according to Semrush data. The mechanism is intent: AI search users arrive with specific, researched queries and a pre-formed shortlist, which translates to higher purchase and contact intent.
Here’s the number that reorganized how we think about search: ~84% of our organic traffic comes from Bing. Not Google. Bing — and the Copilot and ChatGPT surfaces that draw on Bing’s index. Yet for a long time, like nearly everyone, we watched only Google Search Console and treated Bing as an afterthought.
That’s the blind spot this article is about. Short answer: use both consoles, but if Bing drives your traffic, stop treating Bing Webmaster Tools as optional — it has data, indexing controls, and an AI-insights surface that Google Search Console doesn’t, and it’s reporting on the search engine that’s actually sending you readers.
This is the side-by-side from running both consoles on the same media property: what each one tells you, where Bing is quietly ahead, and how we wired the Bing Webmaster Tools API into our editorial calendar.
The core reporting — query, position, CTR
Core reporting: query, position, CTR.
At the surface, the two consoles look like twins. Both give you queries, impressions, clicks, average position, and CTR. The differences are in coverage and freshness.
How we do it
Job
Bing Webmaster Tools
Google Search Console
Verdict
Query / position / CTR
Yes, per query and page
Yes, per query and page
Tie on the basics
Data freshness
Often faster to update
~2-3 day lag
Bing edges ahead
Historical window
Generous
16 months
Toss-up
API access
Full API: position + CTR per query/page
Search Analytics API
Bing — the API is the underrated weapon
AI / Copilot insights
Dedicated AI-traffic insights
No equivalent surface yet
Bing, clearly
Market it reports on
Bing + Copilot + ChatGPT-via-Bing
Google only
Depends on your traffic mix
The honest read: for the basic dashboard, they’re close enough that you’d never switch for the UI. The reasons to take Bing seriously are whose traffic it reports on and what it lets you do about it — the AI insights tab and the API.
Indexing: IndexNow vs crawl-when-it-feels-like-it
Indexing: IndexNow vs crawl-when-it-feels-like-it.
This is the most concrete operational difference, and it’s lopsided.
How we do it
Job
Bing Webmaster Tools
Google Search Console
Verdict
Tell it about a new URL
IndexNow — push, indexed near-instantly
URL Inspection → “Request indexing” (queued)
Bing — push beats poll
Bulk submission
IndexNow ping + sitemap
Sitemap, then wait
Bing
Control over crawl
Crawl control, block/allow
Limited crawl controls
Bing — more knobs
Re-crawl on edit
Re-ping IndexNow
Hope, or re-request
Bing
IndexNow is the standout. Instead of submitting a sitemap and waiting for a crawler to wander by, you push a URL the moment it changes and it’s picked up almost immediately — and because IndexNow is a shared protocol, one ping notifies participating engines. Google’s model is still largely “request indexing and wait.” For a content site that publishes and edits constantly, push beats poll every time. We ping IndexNow on publish and on every meaningful edit.
The AI / Copilot insights tab
The AI insights tab is the differentiator.
Google Search Console has no real equivalent here yet. Bing Webmaster Tools surfaces AI-traffic insights — visibility into how your content shows up across Bing’s AI-powered and Copilot surfaces. Given that those surfaces (and ChatGPT’s web results, which draw on Bing) are an increasing share of how people find answers, this is the single console feature most aligned with where discovery is heading. If you care about GEO at all, it’s the dashboard that tells you whether the AI assistants are actually pulling you in.
Wiring the BWT API into the editorial calendar
The Bing Webmaster Tools API is the part most sites never touch, and it’s the most actionable. It returns position and CTR per query and per page — which is a ready-made content-optimization loop:
Pull query/position/CTR from the BWT API on a schedule.
Find pages ranking on page one with weak CTR (good position, bad headline/meta) — fast wins.
Find queries where we rank position 5-15 with real impressions — the “one good edit from page one” list.
Feed both lists straight into the editorial calendar as prioritized rewrites.
Because Bing drives most of our traffic, this loop is pointed at the engine that actually moves our numbers. Running the same loop off Google Search Console’s API would optimize for the 16% of traffic, not the 84%.
What surprised us
Bing’s data is often fresher than Google’s. We frequently see new queries in Bing Webmaster Tools before they show up in Search Console.
IndexNow is faster than anything Google offers — and it’s free and standard. The gap between “push and it’s indexed” and “request and wait” is real and daily.
The AI insights tab has no GSC counterpart. For a site doing GEO, that’s the most forward-looking surface either console offers.
Almost nobody verifies their site in Bing Webmaster Tools. You can import directly from Google Search Console in a couple of clicks, so the only reason most sites skip it is that they’ve never looked at where their traffic comes from.
The takeaway
This was never a “pick one” — it’s “stop ignoring one.” Google Search Console is still essential; Google isn’t going anywhere. But running only GSC is a bet that Google’s view of your site is the only one that matters, and our traffic data says that bet is wrong by a factor of five.
Use both. Watch Google Search Console for the Google slice. But if a large share of your organic traffic comes from Bing — and a surprising number of content sites are in exactly that position without checking — then Bing Webmaster Tools is your primary console: fresher data, IndexNow for instant indexing, the AI/Copilot insights surface, and an API you can wire straight into your editorial calendar.
The 84% lesson is simple: measure where your readers actually come from, then watch the console that reports on it. For us, that meant promoting Bing from afterthought to the dashboard we open first.
This is part of our “Two Clouds, One Site” series — we run the same media property on Azure and Google Cloud, on the free tiers, and report what watching both ecosystems actually teaches us. The lab lives on tygart.media; the findings publish here.
Should I use Bing Webmaster Tools if I already use Google Search Console?
Yes — they report on different search engines, so using only Google Search Console hides all of your Bing performance. If any meaningful share of your traffic comes from Bing, Copilot, or ChatGPT’s Bing-powered results, Bing Webmaster Tools shows data and offers indexing controls that Search Console doesn’t. You can import your site from Search Console in a couple of clicks.
What is IndexNow and is it faster than Google indexing?
IndexNow is a protocol that lets you push a URL to search engines the moment it’s published or changed, instead of waiting for a crawler. It’s typically much faster than Google’s “request indexing and wait” model, and because it’s a shared standard, one ping notifies participating engines. For sites that publish or edit frequently, it’s a meaningful indexing-speed advantage.
Does Bing Webmaster Tools have an API?
Yes. The Bing Webmaster Tools API exposes per-query and per-page data including position and CTR, plus URL submission. That makes it practical to pull your search performance on a schedule and feed it into a content-optimization loop — for example, flagging page-one results with weak CTR or near-miss rankings to prioritize for rewrites.
What does the Bing Webmaster Tools AI insights tab show?
It surfaces how your content appears across Bing’s AI-powered and Copilot surfaces, giving visibility into AI-driven discovery that Google Search Console has no direct equivalent for yet. For sites focused on Generative Engine Optimization, it’s the most forward-looking view either console offers into whether AI assistants are pulling in your content.
Why would a site get most of its traffic from Bing instead of Google?
It’s more common than people assume, especially for niche or B2B content, sites strong in Bing-heavy regions or browsers, and content that surfaces well in Copilot and ChatGPT’s Bing-powered results. The lesson is to measure your actual referral mix rather than assume Google dominates — many sites only discover their Bing share once they verify in Bing Webmaster Tools.
Generative Engine Optimization (GEO) is the new shape of getting found: instead of ranking a blue link, you make your content legible to AI assistants so they recognize, trust, and cite it. The engine room of that work is entity extraction — pulling the named entities and key phrases out of your content so you can saturate it with the concepts an AI system uses to decide what a page is about.
We run the same articles through both Azure AI Language and Google Cloud Natural Language, on the free tiers, and compare what each one sees. Short answer: for GEO aimed at Bing and Copilot, Azure AI Language is the pick — not because its NLP is categorically better, but because you’re extracting entities with Microsoft’s own signal family to optimize for Microsoft’s own AI. Google Natural Language is an excellent general-purpose NLP API; it’s just optimizing toward a different reader.
This is the breakdown from the running lab on tygart.media — entity quality, key phrases, sentiment, free-tier ceilings, and the strategic point underneath all of it.
The free-tier ceilings
Free-tier ceilings for entity extraction.
How we do it
Azure
Google Cloud
Verdict
Service
Azure AI Language
Cloud Natural Language API
—
Free ceiling
5,000 text records/month
First 5,000 units/month free per feature
Toss-up on raw volume
“Record” definition
Up to 1,000 chars = 1 record
Per 1,000 chars = 1 unit, per feature
Watch Google — billed per feature
Cost after free
Per record
Per 1,000 chars, per feature called
Azure simpler to predict
Always free?
Perpetual free tier
Free monthly allotment, then billed
Tie — both have monthly free
The subtlety: Google bills per feature — entity analysis, sentiment, and syntax each consume their own free allotment and then their own meter. Azure’s 5,000 text records/month is a cleaner mental model for a content pipeline that runs every article through the same extraction pass. At ~300–400 articles a month, both stay at $0; Azure is just easier to reason about.
Entity extraction quality
Entity extraction quality head-to-head.
This is the line that matters most for GEO.
How we do it
Job
Azure
Google Cloud
Verdict
Named entity recognition
Strong, typed categories + subcategories
Strong, with entity types
Toss-up on accuracy
Entity linking
Links entities to a knowledge base
Wikipedia/Knowledge Graph links
Google for KG links; Azure for Bing alignment
Key-phrase extraction
First-class, clean
Not a dedicated feature (infer from entities/salience)
Azure — dedicated key phrases
Salience / ranking
Confidence scores
Salience score per entity
Google — salience is genuinely useful
Sentiment
Document + sentence + aspect-based
Document + entity-level
Toss-up; both solid
Both APIs find the obvious entities. The differences are at the edges: Google’s salience score (how central an entity is to the document) is a genuinely useful GEO signal — it tells you which entities the content is actually about, not just which appear. Azure’s dedicated key-phrase extraction is the cleaner input for content saturation — it hands you the phrases to weave back in, where Google makes you infer them.
For our pipeline, we use Azure’s key phrases as the editing checklist and lean on its typed entity categories to confirm an article is “saturated” with the right concepts before it publishes.
Sentiment and the extra features
Both do document- and sentence-level sentiment well. Azure’s aspect-based sentiment (sentiment tied to specific targets within a sentence) is the richer feature if you’re analyzing reviews or feedback. Google’s entity-level sentiment is comparable for most content work. For a media site doing GEO, sentiment is secondary — entity and key-phrase extraction is the main event — but if you also do feedback analysis, Azure’s aspect-based model edges ahead.
The strategic point — extract with Microsoft’s tooling, optimize for Microsoft’s AI
Here’s the whole game. When you extract entities to optimize content, you’re implicitly choosing a definition of what counts as an entity. Those definitions aren’t universal — Microsoft’s and Google’s models were trained on different data and tuned toward different downstream systems.
Bing and Copilot select and ground content using Microsoft’s signal family — the same lineage that powers Azure AI Language. So when we extract entities with Azure and saturate our articles with what it recognizes, we’re tuning content to the exact signals Microsoft’s own AI uses to decide what to surface and cite. That’s not a coincidence we’re exploiting; it’s the most direct alignment available. With ~84% of our traffic from Bing, optimizing toward Google’s entity model would be optimizing for the wrong reader.
What surprised us
Google’s salience score is the feature we wish Azure had. Knowing which entity is central (not just present) is a sharper GEO signal than a flat confidence list.
Google bills per feature — that’s the budget trap. Calling entities + sentiment + syntax on one document is three metered features, not one. Azure’s per-record model is harder to accidentally triple.
Key-phrase extraction is an Azure advantage that’s easy to miss. Google has no dedicated key-phrase feature; you reconstruct it from entities and salience. Azure just hands you the phrases.
Both miss niche industry entities. Neither model reliably tags specialized restoration-industry or proprietary-standard terms. Custom NER (Azure) or a custom dictionary closes that gap — worth it if your content is jargon-dense.
The takeaway
The takeaway from the NLP bakeoff.
These are both strong NLP APIs, and at our volume both run at $0. The decision is about which AI you’re feeding.
Pick Azure AI Language if your GEO target is Bing and Copilot, you want dedicated key-phrase extraction as a content checklist, and you’d rather extract entities with the same signal family your search traffic actually flows through. That’s us.
Pick Google Cloud Natural Language if you want the salience score, you’re optimizing for Gemini and Google’s Knowledge Graph, or you need general-purpose NLP across mixed workloads. It’s an excellent API — it’s just tuned toward a different reader than the one sending us traffic.
If most of your audience arrives through Bing, extracting your entities with Google’s model is optimizing for the wrong index. We extract with Microsoft’s tooling, on purpose.
This is part of our “Two Clouds, One Site” series — we run the same media property on Azure and Google Cloud, on the free tiers, and publish what the two ecosystems actually do with the same content. The lab lives on tygart.media; the findings publish here.
What is entity extraction and why does it matter for SEO?
Entity extraction (named entity recognition) identifies the people, places, organizations, and concepts in your text. It matters for modern SEO and GEO because search engines and AI assistants understand pages by the entities they contain — saturating content with the right, correctly-recognized entities helps those systems classify and cite it accurately.
Is Azure AI Language free?
Azure AI Language includes a perpetual free tier of 5,000 text records per month, where one record is up to 1,000 characters. For a content site processing a few hundred articles a month, that’s enough to run entity and key-phrase extraction on every piece at $0.
What’s the difference between Azure AI Language and Google Natural Language?
Both extract entities, key concepts, and sentiment, but they differ at the edges: Azure offers dedicated key-phrase extraction and aspect-based sentiment, while Google offers a salience score that ranks how central each entity is to the document. Google also bills per feature, where Azure bills per text record. They’re tuned toward different downstream AI systems — Azure toward Microsoft/Bing, Google toward Gemini and the Knowledge Graph.
What is GEO (Generative Engine Optimization)?
GEO is optimizing content so generative AI assistants recognize, trust, and cite it, rather than optimizing only for blue-link rankings. In practice it means structuring content and saturating it with the right entities and key phrases so the models that answer user questions pull from your pages.
Which NLP API is better for optimizing for Bing and Copilot?
Azure AI Language, because it shares Microsoft’s signal lineage — the same family Bing and Copilot use to select and ground content. Extracting entities with Azure and saturating your articles with what it recognizes aligns your content with the exact signals Microsoft’s AI uses, which is the higher-leverage choice when Bing drives your traffic.
Most “which managed search?” articles compare feature checklists from the vendor docs. We did something more useful: we indexed the same media property’s content into both Azure AI Search and Vertex AI Search, on the free tiers, and watched what each one did with it.
Short answer: for a content site that wants to be found and cited by AI assistants, Azure AI Search is the pick — not because the relevance is dramatically better, but because it’s the retrieval lineage that sits behind Bing and Copilot, and ~84% of our organic traffic comes from Bing. Vertex AI Search is the stronger turnkey RAG product and grounds beautifully into Gemini. Which one wins depends entirely on whose AI you’re trying to get in front of.
This is the desk-by-desk breakdown — free-tier ceilings, setup friction, relevance, and ecosystem grounding — from the running lab on tygart.media.
The free-tier ceilings
Free-tier ceilings on managed site search.
The first thing that matters at our scale is what each gives you for $0, perpetually.
How we do it
Azure
Google Cloud
Verdict
Service
Azure AI Search (Free tier)
Vertex AI Search
—
Storage
50 MB
Generous indexing quota, but query/extraction billed
Azure — true perpetual free
Indexes
3 indexes
Multiple data stores
Toss-up
Documents
~10,000 hosted docs
Effectively higher, but pay-as-you-go
Azure for “always free” certainty
Cost model
Always free, no card pressure
Free trial credits, then per-query/extraction
Azure — Vertex bills as you scale
Semantic ranking
Available (limited on free)
Built in, very strong
Google on raw quality
The honest read: Azure’s 50 MB / 3-index / ~10,000-document free tier is small but genuinely perpetual — it never starts billing at our volume. Vertex AI Search is more capable out of the box but its free posture is trial credits, after which queries and extractive answers meter. For a small content site, Azure’s ceiling is the one you can forget about.
Setup friction
How we do it
Job
Azure
Google Cloud
Verdict
Get to first results
Create service → index → import data source
Create app → data store → point at site/GCS
Google — faster to “it works”
Crawl a website directly
Indexer add-on, more wiring
Website data store crawls URLs natively
Google, clearly
Schema control
Fine-grained fields, analyzers, scoring profiles
More opinionated, less to tune
Azure for control; Google for speed
Vector / hybrid search
Native vector + hybrid (keyword+vector)
Native, with built-in embeddings
Toss-up; both strong
Vertex AI Search gets you to a working search box faster — point it at a sitemap or a Cloud Storage bucket and it crawls and chunks for you. Azure AI Search makes you assemble the indexer, but in exchange you get scoring profiles, custom analyzers, and field-level control that pay off once you care about why a result ranks.
Relevance and semantic ranking
On raw relevance for a handful of queries against the same corpus, Vertex was slightly better out of the box — its semantic ranking and extractive answers are tuned and ready. Azure matched it once we turned on semantic ranking and tuned a scoring profile, but that’s manual work Vertex does for free.
The asymmetry: Vertex is better at answering, Azure is better at being controllable. If you want a search box that produces clean extractive answers with zero tuning, Vertex wins. If you want to deliberately shape what ranks (and you’re optimizing content anyway), Azure rewards the effort.
The grounding angle — whose AI is reading you
Grounding angle — whose AI is reading you.
This is the line that actually decides it for us.
Neither Azure AI Search nor Vertex AI Search “submits your site to Bing or Gemini.” But the retrieval architecture you build on signals which ecosystem you’re fluent in. Azure AI Search is the same managed-retrieval lineage Microsoft uses to ground Copilot, and it’s the natural backend for “Bring your own data” grounding into Azure OpenAI / Copilot Studio. Vertex AI Search is the canonical retrieval layer for grounding Gemini — it’s literally the “ground with your own data” path in Google’s stack.
So the question isn’t “which search is better.” It’s: which AI assistant do you most need to recognize and cite your content? For us, with Bing driving the overwhelming majority of organic traffic, building our retrieval inside Microsoft’s lineage and exposing structured, Copilot-groundable content is the higher-leverage bet.
What surprised us
What surprised us in the comparison.
Azure’s 50 MB is smaller than it sounds — and bigger than it needs to be. Pure text content compresses; 10,000 documents of article body is more than a mid-size site has. The ceiling we’d hit first is index count (3), not storage.
Vertex’s “free” is the easy thing to misjudge. The trial experience is so smooth you forget it’s metered. Set a budget alert before you point it at a large crawl.
Hybrid (keyword + vector) search is now table stakes on both. A year ago this was Azure’s differentiator; Vertex has fully caught up.
Vertex crawls websites natively; Azure wants a data source. If your content lives in a bucket or a DB, Azure’s indexer is fine. If you just want to crawl tygart.media and search it, Vertex is less wiring.
The takeaway
These are both excellent managed search engines, and at small scale both can run at $0 — Azure perpetually, Vertex on credits. The decision isn’t about relevance deltas measured in single queries.
Pick Azure AI Search if your strategic goal is to be retrievable and citable inside the Microsoft / Bing / Copilot ecosystem, you want a truly perpetual free tier, and you’re willing to tune scoring profiles for control. That’s us.
Pick Vertex AI Search if you want the fastest path to a high-quality answering search box, you’re grounding into Gemini, or your content already lives in Google Cloud Storage and you want native crawl-and-chunk with zero schema work.
If most of your readers arrive through Bing, building your retrieval layer only inside Google’s lineage is the same blind spot as watching only Google Search Console. We build on both — and lean Azure for the citation angle.
This is part of our “Two Clouds, One Site” series — we run the same media property on both Azure and Google Cloud, on the free tiers, and report what watching both ecosystems actually teaches us. The lab lives on tygart.media; the findings publish here.
Is Azure AI Search really free?
Yes — the Free tier is perpetual, not a trial. It includes 50 MB of storage, 3 indexes, and roughly 10,000 hosted documents, and it does not start billing as long as you stay inside those limits. For a small content site that’s enough to run real site search at $0.
What’s the difference between Azure AI Search and Vertex AI Search?
Azure AI Search is a managed retrieval engine you assemble (index, indexer, scoring profiles) and the lineage behind Microsoft’s Copilot grounding. Vertex AI Search is Google’s more turnkey managed search and RAG product that crawls and chunks for you and grounds natively into Gemini. Azure favors control and a perpetual free tier; Vertex favors speed-to-answer and pay-as-you-go scaling.
Which is better for getting cited by AI assistants?
It depends on which assistant matters to you. Azure AI Search aligns with Bing and Copilot grounding; Vertex AI Search aligns with Gemini grounding. If most of your traffic and target citations come from Bing, building retrieval inside Microsoft’s lineage is the stronger bet.
Does Vertex AI Search have a free tier?
Vertex AI Search runs on Google Cloud free trial credits rather than a perpetual always-free tier, and after that, queries and extractive answers are billed per use. It’s easy to start for free, but set a budget alert before pointing it at a large website crawl, because metering starts once credits run out.
Can I use Azure AI Search to ground my own AI chatbot?
Yes. Azure AI Search is the standard “bring your own data” retrieval backend for Azure OpenAI and Copilot Studio, supporting keyword, vector, and hybrid search. You index your content, then have the model retrieve and ground its answers against your index, which keeps responses tied to your source material.
