# SEOTesting

> SEOTesting is a Google Search Console analytics and experimentation platform that runs time-based and split SEO tests to measure whether a specific change caused a measurable performance shift, surfaces CTR, content decay, and keyword cannibalization opportunities from your own GSC data, and reports which pages earn clicks from AI chatbots such as ChatGPT, Claude, Perplexity, and Gemini.

- Category: SEO & Content Marketing (https://saastracker.org/categories/seo-content)
- Website: https://seotesting.com
- Starting price: $50 per month (Single Site)
- Free plan: No
- Free trial: 14 days, fully functional, no credit card required, up to 30 sites during the trial
- Founded: 2017, HQ: Bude, Cornwall, UK, Ownership: Independent and bootstrapped, trading as SanityCheck.io Limited; a minority stake was sold to join the TinySeed accelerator
- Profile last reviewed: 2026-08-22
- Canonical profile: https://saastracker.org/products/seotesting

## Overview

Almost every tool in this category tells you what to do. SEOTesting is the rare one that tells you whether what you did worked. It connects to Google Search Console, keeps the data past Google's sixteen-month retention window, and wraps it in an experimentation layer: you declare a change, mark the pages affected, and the tool compares actual performance against a forecast or against a matched control group to say whether the difference is real or noise.

Two test types do the heavy lifting. Time-based tests compare a page group's performance before and after a change against a forecast of what would have happened anyway. Split tests divide comparable pages into test and control groups so the impact is isolated from seasonality and algorithm updates, which is the only honest way to attribute a lift when Google is changing under your feet. Around those sit the reports a working SEO opens weekly: CTR opportunities, content decay, keyword cannibalization, content groups filtered by RegEx, sitemap and robots.txt monitoring, and a monthly executive summary that correlates traffic shifts with your logged tests and with known Google updates.

On AI search, SEOTesting has taken the most grounded position of any vendor in this category, precisely because it refuses to guess. LLM Traffic identifies pages that already receive clicks referred from AI chatbots and surfaces the likely prompts behind them. LLM Tests measure whether a change increased sessions from ChatGPT, Claude, Perplexity, and Gemini. There is also a Query Fan-Out Generator, a semantic Content Evaluator, and a Persona Page Evaluator. None of it claims to tell you your share of voice inside a model's brain; it measures traffic and prompts you can observe. That is a smaller claim and a more defensible one.

## How it works

1. You connect Google Search Console and SEOTesting begins archiving the data, which immediately solves a problem Google created: GSC only keeps sixteen months, and comparative SEO work needs longer. The tool also removes some of the interface's practical limits on how much data you can pull and filter at once.

2. To run a time-based test, you define a group of pages using RegEx or a URL list, log the date and description of the change, and let the test run. SEOTesting forecasts what the group would have done without the change and reports the difference, with the annotation of what you changed attached permanently to that period so a future analyst is not guessing.

3. A split test works differently and is the more rigorous option. You divide a set of comparable pages into a test group and a control group, apply the change to only one, and the tool compares the two over the test window. Because both groups sit under the same domain and experience the same algorithm updates, seasonality and volatility largely cancel out.

4. The AI side runs on observable signals rather than model scraping. LLM Traffic identifies pages taking clicks referred by chatbots and infers the prompt patterns behind them, and LLM Tests measure whether a specific change moved sessions from named assistants. The Query Fan-Out Generator expands a topic into the related questions and objections an answer engine is likely to resolve, which becomes brief input for whichever writing tool you use.

## Best for

SEO practitioners, consultants, and in-house teams who already produce content and now need evidence, particularly anyone who has to defend an SEO program to a CFO or a client and is tired of pointing at a line chart and asserting causation.

## Not the right fit for

- Teams with too little traffic to test; split and time-based testing need enough impressions and clicks for a difference to be distinguishable from noise, and a site with a few hundred monthly clicks will get nothing useful.
- Anyone looking for a content editor, keyword database, or backlink index; SEOTesting deliberately does none of those and expects you to buy them elsewhere.
- Buyers who want AI share-of-voice dashboards showing brand mentions inside chatbot answers; SEOTesting measures AI-referred traffic and prompts, not mentions.
- Very small sites where $50 per month per site is a large share of the marketing budget relative to the number of tests you would actually run.
- Teams unwilling to change one thing at a time; the whole value depends on disciplined experiment design, and a team that ships five changes at once will learn nothing from any tool.

## Features

### SEO testing

The reason the product exists: attributing outcomes to changes.

- **Time-based SEO testing**: Compares a page group's actual performance after a change against a forecast of what would have happened without it, so the result accounts for trend rather than just before-and-after totals.
- **SEO split testing**: Divides comparable pages into test and control groups so seasonality and algorithm updates affect both equally, which is the only rigorous way to isolate impact on a live site.
- **LLM tests**: Measures whether a change increased sessions referred from ChatGPT, Claude, Perplexity, and Gemini, applying the same before-and-after discipline to AI traffic.
- **Annotations and changelog**: Every logged change is pinned to a date and visible on the performance charts, so six months later somebody can still explain the step change instead of inventing a reason for it.
- **Content groups with RegEx**: Define page and query groups using regular expressions, which is how you build a meaningful test cohort out of a messy URL structure.

### Search Console reporting

The reports GSC should have and does not.

- **Data retention past sixteen months**: Archives your Search Console data beyond Google's retention window, which matters enormously for year-over-year comparison and for any test running longer than a quarter.
- **CTR optimization report**: Finds queries with high impressions and weak click-through, which is the cheapest available traffic in most accounts because it needs a title rewrite rather than a new page.
- **Content decay detection**: Surfaces pages losing clicks over time so refresh work is prioritized by evidence rather than by which post someone remembers.
- **Keyword cannibalization**: Identifies where multiple pages compete for the same query, a problem most teams suspect they have and cannot easily prove.
- **Monthly SEO report**: An executive summary correlating traffic shifts with your logged tests and with known Google updates, which is the artifact that survives contact with a leadership meeting.
- **Rank tracking from GSC data**: Positions derived from your own Search Console data rather than from a third-party scraper, so the numbers match what Google reports to you.
- **Daily email alerts**: Key metric changes pushed by email so the dashboard does not need a daily visit.

### AI search and LLM tooling

Grounded in observable traffic rather than scraped model output.

- **LLM Traffic report**: Identifies which pages already earn clicks from AI chatbots and surfaces the likely prompts behind that visibility, which is a concrete starting point for AEO work.
- **Prompt pattern analysis**: Reveals conversational prompt patterns your content appears to be surfacing for, rather than a keyword list.
- **Query Fan-Out Generator**: Expands a topic into the related questions and objections an answer engine is likely to resolve on its way to an answer, useful as brief input for whichever writing tool you use.
- **Content Evaluator**: Measures how well a page covers a topic using semantic similarity rather than keyword matching.
- **Persona Page Evaluator**: Assesses whether a page actually addresses a defined target audience, a check most optimization scores skip entirely.

### Monitoring and workflow

The unglamorous checks that catch expensive mistakes.

- **Sitemap monitor**: Watches XML sitemaps for unexpected changes, which is how you catch a deploy that dropped half the site from the index.
- **robots.txt monitoring**: Alerts when robots rules change, the single most common way a site accidentally deindexes itself.
- **Internal linking credits**: Metered per plan at 25 credits per month on Single Site, 75 on Medium, and 300 on Large, for internal link recommendations.
- **Chrome extension**: View Search Console data for a page while browsing it, without switching tabs.
- **Filtered dashboards**: Views segmented by country, brand versus non-brand, or any custom filter, saved for repeat use.

### Integrations and access

How the data gets out to where decisions are made.

- **API access**: Connect SEOTesting data to internal dashboards and reporting pipelines.
- **MCP server**: An MCP server for wiring the data into AI workflows and assistants, which is a genuinely forward-looking inclusion for a three-person company.
- **Looker Studio connector**: Renders tests and reports inside Looker Studio dashboards for teams that already report there.
- **Unlimited users on every plan**: Pricing is based purely on site count with no per-user fee, so adding the whole team or a client's stakeholders costs nothing.

## Use cases

- **In-house SEO defending a budget**: Rewrote 40 title tags last quarter, traffic went up 12 percent, and the CFO wants to know how much of that was the rewrite and how much was seasonality. Outcome: A split test on comparable pages separates the two, and the monthly report ties the result to the logged change, which converts an assertion into evidence and generally converts a budget conversation as well.
- **Agency running SEO for twenty client sites**: Clients keep asking what the retainer bought this month and screenshots of Search Console are not persuasive. Outcome: The Large plan covers 20 sites at $375 per month with unlimited users, so every client stakeholder gets access, and each engagement ships a report correlating specific changes with specific outcomes.
- **Content lead investigating an AI traffic question**: Analytics shows a growing trickle of referrals from ChatGPT and Perplexity and nobody knows which pages are responsible or why. Outcome: The LLM Traffic report names the pages already earning chatbot clicks and the prompt patterns behind them, which turns AEO from a theory into a shortlist of pages worth replicating.
- **Consultant with one flagship client and a hypothesis**: Believes adding FAQ sections will lift CTR but has been burned before by changes that coincided with a core update. Outcome: A time-based test with a forecast baseline, run on the $50 Single Site plan, either confirms the hypothesis or kills it cheaply before it becomes a site-wide project.

## Pricing

Flat monthly tiers priced purely on the number of websites, with unlimited users on every plan and month-to-month billing.

- **Single Site**: $50 per month. 1 website; Unlimited users; 25 internal linking credits per month; Month-to-month, no contract.
- **Medium**: $125 per month. 5 websites; Unlimited users; 75 internal linking credits per month; 1:1 onboarding call.
- **Large**: $375 per month. 20 websites; Unlimited users; 300 internal linking credits per month; Quarterly training. Works out to under $19 per site per month, which is the tier agencies should be looking at.
- **Enterprise**: Custom quoted. Increased site allowance; Unlimited users; Unlimited training; Customer success support and custom terms.

Billing notes:

- Billing is month-to-month with no contract; you can switch plans or cancel at any time.
- Annual plans are available but are not published on the pricing page; you have to email the company to arrange one.
- The 14 day trial is fully functional, requires no credit card, and allows up to 30 sites, which is unusually generous and lets an agency test the whole portfolio before paying.
- After a trial or cancellation there is a 7 day grace period before data is deleted, so an accidental lapse does not immediately destroy your archive.
- There is no per-user fee at any tier, so cost scales with sites rather than headcount.

Value assessment: At $50 per month for one site, SEOTesting is priced like a niche utility, and per site at the Large tier it drops under $19, which is close to trivial for an agency. The value question is not the price but the fit: it is worth every dollar to a team with enough traffic to run a real test and disciplined enough to change one thing at a time, and worth nothing at all to a site with 300 monthly clicks or a team that ships ten changes in a sprint. The Search Console data archive alone, past Google's sixteen-month window, justifies the entry price for many working SEOs before the testing features are even considered.

## Strengths

- The only tool in this comparison set that establishes causation rather than correlation, through both time-based forecasting and genuine split tests.
- Archives Google Search Console data past the sixteen-month retention limit, which is quietly one of the most valuable things any GSC tool can do.
- Unlimited users on every plan with pricing based solely on site count, so team size never affects the bill.
- The AI features are grounded in observable data (chatbot-referred clicks, prompt patterns, session tests) rather than in scraped model output presented as certainty.
- Ships an API, a Looker Studio connector, and an MCP server, which is unusual technical reach for a company with three or four people.
- A 14 day no-card trial covering up to 30 sites, plus a 7 day post-trial grace period before data deletion.
- Annotations turn institutional memory into a permanent record, which survives staff turnover in a way that a shared spreadsheet does not.

## Limitations

- Useless below a traffic threshold; testing requires enough impressions and clicks for a difference to be statistically distinguishable, and the site never tells you what that floor is.
- Not a content tool at all: no editor, no scoring, no keyword database, no backlink data, so it is always an addition to a stack rather than a consolidation of one.
- Rigor demands discipline. The tool cannot stop a team from shipping six changes at once and then reading the result as if it proved something.
- AI coverage measures referred traffic and prompts, not brand mentions or share of voice inside AI answers, so it does not replace a dedicated AI visibility tracker.
- A three or four person company means a small support surface and real key-person risk, which buyers standardizing on it for a large agency should think about honestly.
- Annual pricing is not published, so getting a discounted rate requires a sales email rather than a checkout button.

## Comparisons

- **SEOTesting vs Ranktracker**: Ranktracker measures positions from outside using its own scrapers across a large keyword set; SEOTesting measures outcomes from inside using your Search Console data. Ranktracker answers where you rank today, SEOTesting answers whether last month's change is why. Agencies reporting positions to clients want Ranktracker, teams trying to learn what actually works want SEOTesting, and the two overlap surprisingly little.
- **SEOTesting vs Surfer SEO**: Surfer decides what to publish and how to optimize it; SEOTesting decides whether the optimization mattered. Surfer costs three to four times as much and covers a far larger surface area, but it will never tell you that a change was noise. Run them together if content is a real budget line, and if you can only afford one, buy the one matching whether your bottleneck is production or proof.
- **SEOTesting vs Clearscope**: Both build on your own Search Console data and both are made by small, focused teams, but they sit at opposite ends of the workflow. Clearscope shapes the content before it ships and reports AI topic visibility; SEOTesting evaluates the result afterward with controlled tests. At $50 per month SEOTesting is cheap enough to add alongside Clearscope rather than instead of it.

## Implementation

- Setup time: Under thirty minutes to connect Google Search Console and see reports. Meaningful testing takes longer to start because someone has to design a valid experiment, define page groups, and agree not to change anything else during the window.
- Learning curve: Low for the reports, moderate for the testing. The interface is straightforward, but interpreting a test result correctly requires understanding what a control group does and does not prove, and that is a skill, not a setting.
- Onboarding: Self-serve with a 14 day no-card trial. A 1:1 onboarding call is included from the Medium tier, quarterly training on Large, and unlimited training on Enterprise, which is a sensible ladder for a product where methodology matters more than features.
- Migration: Read-only against Google Search Console with no publishing or sending dependencies, so adopting or dropping it carries no operational risk. The cost of leaving is the archived GSC history beyond Google's sixteen-month window, which does not come back.

## Platform, API & security

- Platforms: Web app, Chrome extension, Looker Studio connector, API, MCP server
- API: API access is available for connecting test and report data to internal dashboards, alongside an MCP server for AI workflows and a Looker Studio connector.
- Compliance: UK-registered company (SanityCheck.io Limited); detailed certifications not published
- Data residency: Not publicly specified; the company is based in Cornwall, UK.
- SSO: Not publicly advertised; unlimited users are included on every plan.
- Security notes: Requires read-only OAuth access to Google Search Console properties. The tool does not publish content, modify sites, or require write permissions anywhere.

## Support

- Channels: Email support, 1:1 onboarding call from the Medium tier, Quarterly training on Large, Unlimited training on Enterprise
- Documentation: Documentation written largely by the founder, plus an unusually practical blog on SEO testing methodology that functions as the product's real onboarding material.
- Community: No dedicated forum; the founder is publicly active in bootstrapper and SEO communities and support is vendor-direct.

## Company

- Founded: 2017
- Founders: Nick Swan
- Headquarters: Bude, Cornwall, UK
- Ownership: Independent and bootstrapped, trading as SanityCheck.io Limited; a minority stake was sold to join the TinySeed accelerator
- Employees: Approximately 3 to 4 (Nick Swan, plus developer and customer experience staff)
- Funding: Bootstrapped, with a small stake sold as part of joining TinySeed. No conventional venture rounds disclosed.

Timeline:

- 2016: Launched as SanityCheck in November, initially solving the problem that Google Search Console discards data after sixteen months.
- 2017: Parent company SanityCheck.io Limited incorporated in October in Bude, Cornwall.
- 2020: Rebranded to SEOTesting in April as the product moved from data archiving into SEO experimentation.
- 2022: Joins the TinySeed accelerator, selling a small minority stake while remaining founder-controlled.
- 2024: Reporting suite broadens with CTR optimization, content decay, cannibalization, and RegEx content groups; API and Looker Studio connector ship.
- 2025: LLM Traffic and LLM Tests launch, measuring clicks and sessions referred from ChatGPT, Claude, Perplexity, and Gemini.
- 2026: AI SEO tooling expands with the Query Fan-Out Generator, semantic Content Evaluator, and Persona Page Evaluator, alongside an MCP server.

## Integrations

Google Search Console, Looker Studio, Chrome extension, API, MCP server

## FAQ

### What does SEOTesting do?

It connects to Google Search Console and lets you run controlled SEO experiments, either time-based tests comparing performance against a forecast or split tests using a control group, so you can tell whether a change actually caused a result. It also archives your GSC data past Google's sixteen-month limit and reports CTR opportunities, content decay, cannibalization, and AI chatbot traffic.

### How much does SEOTesting cost?

Single Site is $50 per month, Medium covers 5 sites at $125, Large covers 20 sites at $375, and Enterprise is quoted. Every plan includes unlimited users, billing is month-to-month with no contract, and annual plans are available by email but not published.

### What is the difference between a time-based test and a split test?

A time-based test compares a page group's performance after a change against a forecast of what it would have done otherwise. A split test divides comparable pages into test and control groups so both experience the same seasonality and algorithm updates, isolating the change's impact. Split testing is more rigorous but needs enough comparable pages to build two valid groups.

### Do I need a lot of traffic for SEO testing to work?

Yes, and this is the honest limitation. Statistical testing needs enough impressions and clicks for a difference to stand out from natural variance, so a site with a few hundred monthly clicks will not produce readable results. The reporting features still have value at low volume; the testing features do not.

### Does SEOTesting track AI search visibility?

It measures AI-referred traffic rather than brand mentions. LLM Traffic identifies pages already earning clicks from chatbots and surfaces likely prompts behind them, and LLM Tests measure whether a change increased sessions from ChatGPT, Claude, Perplexity, or Gemini. If you want share-of-voice inside AI answers, you need a dedicated tracker like the ones in Surfer or Clearscope.

### Why archive Google Search Console data?

Because Google deletes it after sixteen months, which makes true year-over-year comparison impossible and quietly destroys the baseline for any long-running test. SEOTesting keeps the history indefinitely while you are a customer, which for many working SEOs justifies the subscription before any testing feature is considered.

### Does SEOTesting have a free trial?

Yes, 14 days, fully functional, with no credit card required and up to 30 sites allowed during the trial, which is enough for an agency to evaluate its whole portfolio. After the trial there is a 7 day grace period before your data is deleted.

### Is SEOTesting a replacement for a rank tracker?

Partly. It provides rank tracking derived from your own Search Console data, which is more accurate for your own site than any scraper, but it cannot track keywords you do not rank for or monitor competitor positions. Teams that need competitive rank data still want a dedicated tracker alongside it.

### Who runs SEOTesting?

Nick Swan founded it and still designs features, supports customers, and writes the documentation. The company operates as SanityCheck.io Limited from Bude, Cornwall, with a team of roughly three to four people, bootstrapped apart from a small stake sold when joining the TinySeed accelerator.

### Does SEOTesting need write access to my website?

No. It connects read-only to Google Search Console and never publishes, edits, or modifies your site. You implement changes yourself and log them in the tool, which is also why the annotation feature matters: SEOTesting only knows what changed if you tell it.

## Editorial verdict

SEOTesting is the most intellectually honest product in this category and the one most likely to change how a team works rather than just what it produces. Everything else here helps you make content; this tells you whether the content did anything. Split testing on a live site, a Search Console archive that outlives Google's own retention, unlimited users at $50 per month, and AI features that measure observable traffic instead of inventing a share-of-voice metric all point at a team more interested in being right than in being impressive. The constraints are equally clear: you need real traffic, real discipline about changing one thing at a time, and another tool to actually produce the content. Add it to the stack. Do not expect it to be the stack.

---

Source: SaaSTracker (https://saastracker.org), an independent editorial project. This profile is compiled from public information, carries no peer reviews or paid placement, and was last reviewed 2026-08-22. Awards are judged on published criteria: https://saastracker.org/methodology