Every correction we have made to this site, with the date, what was wrong, and what a reader would have seen before it was fixed.
Perplexity publishes three things it looks for before it labels a site a trusted source: whether the site corrects its mistakes, whether it identifies its authors, and whether it keeps what it sells separate from what it states. This page is the first of those three, kept as a log rather than a promise. A page that says "we correct our mistakes" and lists nothing is a claim with no evidence behind it.
Newest first. Each entry says what was wrong, what a reader would have seen, and what is there now. Nothing here is a typo or a broken link; those get fixed without a note. These are the ones where the site told somebody something that was not true.
What was wrong. Every article except the newest carried the byline "AgencyKit Team". A reader who wanted to know who had written the piece, and whether that person had run an agency, had no way to find out.
What it is now. All 86 articles now carry a named byline at the top, a biography box at the bottom, and a link to an author page. The author is a person in the page's structured data, not an organisation.
What was wrong. The machine-readable copy of each FAQ answer had been written once and then left alone while the visible prose was rewritten. Twenty-nine of them, across seven pages, had drifted far enough that a search engine quoting the markup would have quoted something the page no longer said.
What it is now. All 985 answers were rewritten from the visible copy, and a check now refuses to publish the site if a single one drifts again.
What was wrong. Twenty-eight guides carried step-by-step markup for search engines. Of its 135 steps, only 7 had text a reader could find anywhere on the page. The rest had been written beside the page rather than from it.
What it is now. Forty-three were rewritten from the visible copy. The other 92 have no visible step to pair with, so they were left exactly as they are and listed by name for a person to write properly. Inventing a plausible sentence for a step the page never shows would have been the same mistake in a tidier shirt.
What was wrong. Articles carried a "Reading this in Google?" heading with Google's Preferred Sources button under it. The button only paints for a reader Google recognises, so everybody else saw the heading above 218 by 40 pixels of nothing.
What it is now. The block now hides itself until the button actually appears.
What was wrong. A file read by AI crawlers on 185 pages gave the plan prices as $29 and $49. The real prices are $9, $39 and $69. Nothing billed from that file, which is exactly why it drifted, and any assistant quoting it was giving a figure ten to twenty dollars away from what a customer pays.
What it is now. The file reads the live prices. A check compares it against them before the site can be published.
What was wrong. The site's structured data described three offers named "Starter Lifetime Deal", "Agency Lifetime Deal" and "Agency Pro Lifetime Deal" at $49, $99 and $199. AgencyKit is a monthly subscription and has never sold a lifetime licence. Those three prices appeared nowhere else on the site. Two trial emails made the same offer, three times before expiry and once after.
What it is now. Removed from the structured data and from both emails. The emails now point at the upgrade page, which is the page written for the person reading them.
What was wrong. No page named the company behind AgencyKit, not in the footer and not in the privacy policy, which opened with "we" and left it there.
What it is now. Every page now names Win Market Agency, Cairo, Egypt, in the footer, and the privacy policy names it as the data controller.
Every article on this site is written by a named person with a link to their page, and the author is a person in the page's structured data rather than a company. Google strongly encourages accurate authorship, Microsoft's publishing rules refuse pseudonyms and "Admin" outright, and one of Perplexity's three criteria is whether a site identifies its authors. It is also simply the honest thing: you should be able to see who is telling you this and decide what that is worth.
Thirty pages on this site compare AgencyKit with HoneyBook, Dubsado, Bonsai, Moxie, 17hats, ClickUp, Asana, Notion, Trello or Clockify. Here is the standard they are held to, so you can check it rather than take our word for it.
Every competitor price and feature is read off that company's own published pricing page, with the date it was read shown on the page, not taken from a review site or a round-up. Where a competitor does something we do not, the page says so and names it. Where a price has changed since we last checked, the date tells you how stale the figure is, which is why the date is there at all. And if you find a figure that is wrong, the next section tells you how to say so and it gets corrected here with its date.
The two worst entries in the log above were both the same kind of mistake: the site described something the software did not do. A lifetime licence that was never sold, and a price that had moved. Both were true once and became false without anybody touching the page.
So a feature claim on this site is now checked against the code that implements it before it is written, and the prices and plan limits are read from the live store rather than typed. Several of those checks refuse to publish the site at all when they disagree with what is on the page. It is not a guarantee of being right. It is a guarantee that being wrong gets noticed by something other than a customer.
Tell us at hello@agencykit.tech and say which page. A correction to a factual claim gets made and listed here with its date. A correction that changes what a reader should have concluded gets a note on the page itself, not only here.
Last reviewed 18 September 2026. This page is reviewed whenever an entry is added to it, and the date above changes then rather than on a schedule, because a review date that moves on its own says nothing.