Editorial policy & process

Editorial content on Toolradar is written and signed by Louis Corneloup, founder of Toolradar and Dupple, the publisher behind 6 newsletters reaching 950K+ tech professionals. Louis is the editor responsible for every guide, blog post, report and hand-written comparison published under his name. When external contributors write under their own byline, their bio and verifiable profiles (LinkedIn, X, GitHub) appear at the top and bottom of the article.

Editorial sign-off is not a content-mill rubber stamp. It means the editor has read the piece end to end, checked sources, and confirmed it represents what we would tell a friend asking the same question. Pages that fail that bar do not ship.

Editorial team: 1 named editor (Louis), 0 anonymous staff writers, 0 sponsored bylines. We will add named contributors in 2026 and disclose them on this page when they join.

• Pricing is read from the vendor's own pricing page and every price shows the date it was last read (a tool whose page was never read shows no date). Pages are re-read on a rolling schedule, oldest first. If a vendor changes a tier, we update the page and log the change in the tool's pricing history.
• Feature claims come from the vendor's product docs, hands-on testing, or first-party user reviews, never aggregated marketing copy without verification. If a vendor claims an integration on their landing page that we cannot find in their integrations directory or changelog, we leave it out.
• Pros and cons on hand-written pages come from user reviews and documentation. On tool pages they are drafted by software from the tool's stored data and public reviews and are not reviewed one by one; the ratings shown are the published ones, with their source and review count.
• Comparison verdicts on hand-written comparisons reflect the editor's judgment based on testing or documented evidence. Generated comparisons name no overall winner: they state where each tool leads on stated criteria (user rating, entry price on the same billing basis, free plan) with the numbers.
• Media mentions (15,600+ so far) link to the original article and show the publication and the publish date its feed gives. Only independent press, blogs and newsletters count: a vendor's own blog, review sites, forums and sites that republish other outlets do not, and one story carried by several outlets counts once. Each outlet has a tier; in the ranking score a tier 1 story weighs 3 times a tier 3 one, and further stories from the same outlet count less. The live feed is on the /in-the-news page.

Every numerical claim, pricing, user counts, founding year, funding rounds, is checked against a primary source before publication. Vendor names, product names, and capitalizations follow the vendor's own usage. If a claim cannot be verified to a primary source, we either leave it out or flag it as "reported" with the source linked inline.

For new tools added through the public submission flow, the editor reviews the submission against the vendor's site, app store listing (if any), and at least one third-party signal (G2, Capterra, Product Hunt, GitHub) before publishing. Submissions that look generated, low-effort, or self-promotional get rejected with a reason.

We use large language models for three things: drafting first passes of structured fields (TL;DRs, feature lists), normalizing data fetched from vendor sites (pricing tables, capability matrices), and surface-area tasks like categorization and slug generation. Every AI-assisted output is reviewed by the editor before publication and corrected where the model got it wrong, which happens often enough that we treat the AI as a fast intern, not a co-author.

We do not publish AI-generated long-form content (guides, blog posts, comparison verdicts) without substantive human rewriting. Pages that read like raw model output have no place on Toolradar. If you find one, email us, it's a process failure and we'll fix the page or unpublish it.

When we make a substantive correction, a wrong price, a misattributed quote, a claim a vendor disputes with evidence, we update the page, change the "Updated" date in the byline, and (for material errors) add a brief correction note at the bottom of the page. We do not silently rewrite history.

Pages also get refreshed proactively: pricing weekly, feature lists when a vendor ships a major release, and editor's takes when our own opinion on a tool changes after extended use. The byline date reflects the most recent meaningful update, not just a cron job touching a timestamp.

Typical SLA: 48 hours for vendor-flagged factual disputes with a primary source attached. Pricing corrections that affect rankings get re-scored on the next nightly job.

Toolradar makes money from sponsorships in the Dupple newsletter network, paid placements and affiliate links. Some offers on our pages may be paid placements or contain affiliate links. Editorial content (reviews, pros and cons, pricing data) is written independently. Money does not change a tool's ranking score, which is computed only from public review volume and rating (G2, Capterra, Trustpilot, app stores) and media mentions, and no vendor can pay to change the editor's written take.

A tool we recommend in a guide can also sponsor a Dupple newsletter. We track these overlaps internally so they never change a score or an editorial take.

Our full revenue and disclosure policy lives at /how-we-make-money.