Buyer's guideNo named competitors

What an SEO tool can honestly tell you, and what needs paid data

Keyword volume, SERP position, and backlink counts all come from data nobody gets for free. Six questions about which parts of an SEO tool are genuinely first-party and which are quietly a guess — with Automix's own honest answer, including the numbers it refuses to show at all.

See what Khoji already does Read the technical docs

What this guide will not do

No named or characterised competitorNo invented statistic or ratingHonest where our own answer is a limit

Before you buy

A rank-tracking number with no data source is not a metric. It is a guess with a decimal point.

Some things an SEO tool tells you come from your own site and require no third party at all: what your page's title tag says, whether your robots.txt blocks a crawler, whether your JSON-LD matches what the page actually displays. Other things — how many people search a keyword each month, where you rank against a competitor, whether Google has actually indexed a page — require paid access to data nobody generates for free: Google's own Search Console, a search-volume index, a backlink crawl of the whole web. A tool that blurs the two categories is the one worth being suspicious of.

The six questions below do not name a specific SEO platform. They separate what any tool can honestly compute itself from what it needs a paid data source to say — and answer each one about Khoji, Automix's SEO agent, including the parts Khoji plainly does not do yet because doing them without real data would mean inventing numbers.

Ask what it refuses to do

A tool with no refusals has not thought about the edges of your business.

These are specific refusals in Automix's own code, named so you can ask the same question of anything else you evaluate.

It will not emit a schema value your page text does not actually contain

Across fifteen schema types, a rating, price or availability figure the page does not display is refused rather than filled in to complete the block.

It will not claim to know whether Google indexed a page without a connected Search Console property

The sitemap-versus-robots reconciliation is real and first-party. Actual index status is stated as unknown, not guessed at, until the owner's verified property is connected.

It will not benchmark your content against competitors it cannot actually see

With no SERP data connected, competitive benchmarking is not offered at all — not approximated with a model's guess about what a top-ranking page probably contains.

Evaluation criteria

6 questions to ask any vendor in this category.

Ask these of Automix and of anything else on your shortlist. The status pill is how the honest answer scores today, not a claim of perfection.

Does a technical audit remember your last check — and does it look at AI crawlers, not just Google's?

Why it matters. 'Your pricing page went noindex on Tuesday' is only a useful sentence if the tool has a before to compare against; a one-off snapshot cannot say what changed, only what is true right now. And in 2026, a site's visibility to AI crawlers — the bots behind AI answer engines, not just search indexes — is a distinct question from its visibility to Googlebot, and worth checking separately.

How Automix answers it. An audit fetches one page through an SSRF guard, stores the result, and reports what changed since the last run over the same page. Seventeen on-page checks run, plus robots access checked per named AI crawler, not just Googlebot. A bounded whole-site crawl now runs alongside it — capped at 100 pages and depth 4, with robots.txt re-checked before every single fetch, not only the first — and closes the gaps a one-page audit cannot see on its own: broken internal links, redirect chains, orphan pages the sitemap lists but nothing links to, and duplicate titles across the whole site. Per-page load time and Core Web Vitals stay a separate, still-gated capability, because that needs a PageSpeed Insights connection nothing here has by default.

Works today

Will generated schema markup ever include a value your page text does not actually contain?

Why it matters. Structured data is a promise to a search engine about what is on the page. A schema generator that fills in a plausible-looking rating, price or review count the page never displays is producing exactly the kind of markup that gets a site penalised for manipulation once discovered.

How Automix answers it. Across fifteen schema types, the generator refuses to emit a value the supplied page text does not actually contain — it will not invent a rating or a price to complete a schema block.

Works today

Can it check schema markup already on your page, not just generate new markup?

Why it matters. Structured data drifts. A page's on-page copy changes and nobody updates the schema block sitting in the template, so the markup ends up asserting a rating, a price or an availability status the visible page no longer shows — which is exactly where an inflated rating a page never actually displays tends to come from.

How Automix answers it. The same rules that stop the generator from inventing a value run backwards over markup already on a page, checking it against what the page text actually contains today.

Works today

Can it tell you whether Google has actually indexed a page, or only whether you submitted it?

Why it matters. A sitemap entry is a request, not a fact. Whether Google's own index actually contains a page — and why not, if it does not — is data that only Google itself has, through Search Console, and no amount of first-party crawling can substitute for it.

How Automix answers it. The sitemap is parsed and reconciled against robots.txt, so a URL that is both submitted and blocked is surfaced, and a real sitemap.xml can be built straight from one of your own crawls — both need no third party. Whether Google has actually indexed a page needs a connected Search Console property: inspecting a URL or submitting a sitemap to Google without one returns a 409 naming the missing connection, not a guess. Once connected, coverage is real arithmetic over every page you have actually asked it to inspect — not Google's own whole-site index count, which the API this deployment uses does not expose.

Needs your setup

Does it scan for spammy ranking-manipulation tactics, wherever the advice came from?

Why it matters. Search engines publish spam policies naming specific tactics — hidden text, link schemes, scaled content abuse — and a lot of SEO advice circulating in the wild, including AI-generated advice, recommends exactly those tactics without saying so. A tool that can flag that pattern in advice you are about to follow is doing something genuinely useful with no third-party data at all.

How Automix answers it. SEO advice from any source is scanned for ranking promises and tactics named in the search engines' own published spam policies — a first-party check that needs nothing beyond the text itself.

Partial today

Can it benchmark your draft against what is actually ranking for that query today?

Why it matters. 'Optimise this against the competition' is a claim that only means something if the tool has real SERP data to compare against. Without it, 'competitor analysis' collapses into a model guessing what a top-ranking page probably contains — which is not analysis, it is a plausible-sounding story.

How Automix answers it. Benchmarking a draft against top-ranking pages needs SERP data nobody has connected in this deployment, so it is not offered. What does work with nothing connected: scanning any draft or any piece of advice for the specific tactics search engines' own spam policies name.

Partial today

Where we do not measure up yet

In one place: which of the answers above is not a green light.

Pulled straight from the criteria above, not a separate list — a guide that scores itself full marks on every criterion is not one you should believe, ours included.

Can it tell you whether Google has actually indexed a page, or only whether you submitted it?

The sitemap is parsed and reconciled against robots.txt, so a URL that is both submitted and blocked is surfaced, and a real sitemap.xml can be built straight from one of your own crawls — both need no third party. Whether Google has actually indexed a page needs a connected Search Console property: inspecting a URL or submitting a sitemap to Google without one returns a 409 naming the missing connection, not a guess. Once connected, coverage is real arithmetic over every page you have actually asked it to inspect — not Google's own whole-site index count, which the API this deployment uses does not expose.

Needs your setup

Does it scan for spammy ranking-manipulation tactics, wherever the advice came from?

SEO advice from any source is scanned for ranking promises and tactics named in the search engines' own published spam policies — a first-party check that needs nothing beyond the text itself.

Partial today

Can it benchmark your draft against what is actually ranking for that query today?

Benchmarking a draft against top-ranking pages needs SERP data nobody has connected in this deployment, so it is not offered. What does work with nothing connected: scanning any draft or any piece of advice for the specific tactics search engines' own spam policies name.

Partial today

Questions

Common questions.

Does Khoji do keyword research or rank tracking?

Not yet. Keyword search volumes and SERP rank positions come from paid third-party indexes nobody has connected in this deployment, and building that feature against invented numbers would produce a tool that looks like an SEO suite while reporting fiction — so it is left out rather than faked.

Can Khoji check backlinks?

No — a backlink index is one of the paid, third-party data sources this deployment does not have connected, for the same reason rank tracking is not offered.

What can Khoji actually do with zero setup?

A one-page technical audit with change-over-time comparison, schema generation and validation in both directions, sitemap-versus-robots reconciliation, and a scan for ranking-manipulation advice — all first-party, all working with no credential connected.

If I connect Google Search Console, does more become available?

Index status specifically — whether Google has actually indexed a given page and why not — becomes real Search Console data instead of an unknown. That connection is not built into this deployment yet.

Why would an SEO tool refuse to generate a schema value instead of just filling one in?

Because a schema block asserting a rating or price the page does not display is the kind of structured-data manipulation search engines specifically penalise once detected — refusing to invent the value is a safety decision, not a missing feature.

Is Khoji available to use today?

Not yet — its status is Coming soon, shown the same honest way next to every agent still in build on this site.

Apply it

Apply these six questions to Khoji, and to whatever you use today.

Khoji's implemented surface is deliberately narrow — four capabilities, each first-party. Ask any SEO tool you evaluate to separate its first-party features from the ones that quietly need paid data it may or may not actually have.

See what Khoji already does Read the technical docs

Khoji's own status today:Coming soon

More buyer's guides

Buyer's guide

How to choose a WhatsApp business tool

Six questions that matter more than the feature list any vendor hands you — plus the one constraint every tool built on Meta's Cloud API shares, whether its marketing page mentions it or not.

Read the guide

Buyer's guide

What to look for in a CRM for a small sales team

The feature list on a CRM's homepage is never the reason a required field gets skipped six months in. Six questions about what actually holds under an import, a bulk edit, or a rushed Friday afternoon — plus Automix's own honest answers.

Read the guide

Buyer's guide

Choosing a social media scheduling tool

The word 'scheduled' hides more than it reveals. Six questions about approvals, invented engagement data, and what actually happens when a post's date arrives — with Automix's own honest answer, including the one place nothing actually publishes yet.

Read the guide

What an SEO tool can honestly tell you, and what needs paid data

What this guide will not do

No named or characterised competitorNo invented statistic or ratingHonest where our own answer is a limit

What an SEO tool can honestly tell you, and what needs paid data

A rank-tracking number with no data source is not a metric. It is a guess with a decimal point.

A tool with no refusals has not thought about the edges of your business.

It will not emit a schema value your page text does not actually contain

It will not claim to know whether Google indexed a page without a connected Search Console property

It will not benchmark your content against competitors it cannot actually see

6 questions to ask any vendor in this category.

Does a technical audit remember your last check — and does it look at AI crawlers, not just Google's?

Will generated schema markup ever include a value your page text does not actually contain?

Can it check schema markup already on your page, not just generate new markup?

Can it tell you whether Google has actually indexed a page, or only whether you submitted it?

Does it scan for spammy ranking-manipulation tactics, wherever the advice came from?

Can it benchmark your draft against what is actually ranking for that query today?

In one place: which of the answers above is not a green light.

Can it tell you whether Google has actually indexed a page, or only whether you submitted it?

Does it scan for spammy ranking-manipulation tactics, wherever the advice came from?

Can it benchmark your draft against what is actually ranking for that query today?

Common questions.

Apply these six questions to Khoji, and to whatever you use today.

Evaluate the rest of your stack.

How to choose a WhatsApp business tool

What to look for in a CRM for a small sales team

Choosing a social media scheduling tool

What an SEO tool can honestly tell you, and what needs paid data

A rank-tracking number with no data source is not a metric. It is a guess with a decimal point.

A tool with no refusals has not thought about the edges of your business.

It will not emit a schema value your page text does not actually contain

It will not claim to know whether Google indexed a page without a connected Search Console property

It will not benchmark your content against competitors it cannot actually see

6 questions to ask any vendor in this category.

Does a technical audit remember your last check — and does it look at AI crawlers, not just Google's?

Will generated schema markup ever include a value your page text does not actually contain?

Can it check schema markup already on your page, not just generate new markup?

Can it tell you whether Google has actually indexed a page, or only whether you submitted it?

Does it scan for spammy ranking-manipulation tactics, wherever the advice came from?

Can it benchmark your draft against what is actually ranking for that query today?

In one place: which of the answers above is not a green light.

Can it tell you whether Google has actually indexed a page, or only whether you submitted it?

Does it scan for spammy ranking-manipulation tactics, wherever the advice came from?

Can it benchmark your draft against what is actually ranking for that query today?

Common questions.

Apply these six questions to Khoji, and to whatever you use today.

Evaluate the rest of your stack.

How to choose a WhatsApp business tool

What to look for in a CRM for a small sales team

Choosing a social media scheduling tool