The AI visibility check costs us real money to run — it makes live model calls. The other fourteen tools make none: they are arithmetic over a file or numbers you already have, and all fourteen run in your browser, so what you give them never reaches us and there is no server work behind them to pay for. All fifteen are free, because showing you something true is a better pitch than describing it.
We ask one engine — Claude, with live web search — four of the buying questions your customers type, then show you whether you get recommended and exactly which competitors get named instead. Every figure carries a 95% Wilson interval, four samples is four samples, and you get the full prompt list.
Three a day, no account, and a completed result is cached for seven days per domain.
Export your Search Console performance with Compare selected — the range before your change against the range after it — and list the pages you changed. On a day-by-day export we choose comparison pages from your own data, using days outside the change windows; a Compare export has no such days, so every other page with at least 10 clicks across the two periods is used, and the result says so. You get the lift versus what those pages predict, with its 95% interval and p-value — or a plain statement that your file cannot support an answer.
Your file never leaves your browser: it is read by JavaScript in your tab, so there is no upload, no account, no model calls and no rate limit. It is arithmetic over your own numbers, and running it costs us nothing because there is no server work behind it.
A day-by-day export, without the change list. Every page is classified against the rest of your own site over identical calendar days, so a seasonal dip that hit everything does not read as decay — and a page with too little history is told "can't tell" with the reason rather than handed a number.
It needs a table with a Date column and a Page column — one row per page per day — which the Search Console download button does not produce. The tool page names the three exports that do.
Your file never leaves your browser: no upload, no account, no model calls, no rate limit — the same as the one above, because both tools run the same code on the same kind of file, in your tab.
Google's Generative AI performance report — AI Overview and AI Mode impressions — is not its own search type in the Search Console API, has no clicks column and a 1,000-row cap, so an export is how you keep it. Upload it and get the trend over whole weeks, how concentrated the impressions are across your pages, and, with your regular Performance export beside it, each page's share of its web impressions. It says when the table may have been cut off, when the newest day is unreliable, and refuses to print a share above 100%.
Your file never leaves your browser: no upload, no account, no rate limit. You can download a baseline CSV to keep.
Google documents data anomalies that silently corrupt before-and-after comparisons, but not in your export, and its annotations do not show in Compare mode. Upload an export and see every documented anomaly that overlaps your dates and what is safe to compare instead — and whether your own file still carries Google's 2025–26 impressions logging error, measured as impressions moving when clicks did not.
Your file never leaves your browser: no upload, no account, no rate limit.
Before you change a page, find out whether a test on it is worth running. It runs Penlid's own estimator thousands of times on pages shaped like yours and tells you how often it would catch a real change of each size, how big a change has to be to be caught four times in five, and when the answer arrives. Generic A/B calculators assume visitors are split at random between two versions, which a page edit is not.
Everything you type, and any export you drop on it to estimate how differently your own pages move, stays in your browser: no upload, no account, no rate limit.
Type the clicks of the page you changed, before and after, and of three to eight pages you left alone. It runs Penlid's own estimator, not a copy, and returns the lift versus what those pages predict with its 95% interval and p-value, or a plain statement that the numbers cannot tell. It explains that comparison pages must be chosen before you look, and that the windows are 28 days.
Nothing you type leaves your browser: no upload, no account, no rate limit. For the same test on a Search Console file, see did it work?; to size a test before you run it, see the SEO test calculator.
List the queries whose average position sits in a band you choose, 8 to 20 to start, from your Search Console Queries export, ranked by impressions with their clicks, CTR and position. It says what the file cannot: a position is an average, the table stops at 1,000 rows, rare queries are left out, and a query is not a page, with the two clicks that find the page.
Your file never leaves your browser: no upload, no account, no rate limit. You can download the list as a CSV.
Builds your site's own click-through curve by position from your Queries export, weighted by impressions and smoothed so it never rises, then lists the queries sitting well below it with the click gap. Every gap has to survive a binomial noise check, brand terms can be left out, and a file too thin to draw a curve is told so rather than handed an industry average.
Your file never leaves your browser: no upload, no account, no rate limit.
Type your brand name and its misspellings and see clicks, impressions and CTR for branded against non-branded queries from your Queries export. You choose the terms, matching is by whole word with case and accents ignored, and the biggest queries on each side are listed so you can check it. It says that anonymized queries are in neither share, and how it differs from Search Console's own branded filter.
Your file never leaves your browser: no upload, no account, no rate limit. You can download every query with its group as a CSV.
Google lists its confirmed ranking updates on the Search Status Dashboard. Upload your Search Console daily export and see how your clicks and impressions changed around each one inside it: whole weeks before and after, the rollout and a settling week left out, and the change placed among the changes your own file makes at ordinary times. Overlapping updates and Google's documented data errors are named, and the page says plainly that this is timing evidence, not proof.
Your file never leaves your browser: no upload, no account, no rate limit.
Search Console's Queries table leaves out anonymized queries and stops at 1,000 rows, while the chart total includes both. Upload your Performance export and see what share of your clicks and of your impressions is in no query row. It is given as an upper bound when the table is at its row limit, and refused when a query, page or search appearance filter would make the figure mean something else.
Your file never leaves your browser: no upload, no account, no rate limit.
Search Console shows up to 16 months of data. Drop every daily export you have onto this tool and get one deduplicated daily CSV: a stated rule for which value wins where exports overlap, conflicts beyond rounding listed with both values, gaps listed and never filled, and exports taken under different filters refused.
Your files never leave your browser: no upload, no account, no rate limit.
Set your sitemap beside a Search Console Pages export taken over a long range. You get the addresses with no impressions, but only when the table is under Google's 1,000-row limit and was not filtered, and the lowest-traffic pages that did earn something. Each is a page to review (improve, merge, redirect or noindex), never one this tool tells you to remove, and it names the canonical, noindex and new-page checks to make first.
Your file never leaves your browser: nothing is fetched, and there is no upload, no account, no model calls and no rate limit. You can download the review list as a CSV.
Upload a table with a Query column and a Page column and see the queries where two or more of your pages each hold a real share of the impressions, with the split and the positions, and, with dates, how often the top page changed. It gives the four routes to that table as Google documents them: Data Studio (formerly Looker Studio), the Search Analytics API, a Sheets add-on and the BigQuery export. It says plainly that the list is a heuristic: two pages for one query can be right.
Your file never leaves your browser: no upload, no account, no model calls, no rate limit.