Editorial policy & process

Editorial content on Toolradar is written and signed by Louis Corneloup, founder of Toolradar and Dupple, the publisher behind 5 newsletters reaching 550K+ tech professionals. Louis is the editor responsible for every guide, blog post, and comparison published under the Toolradar byline. 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 fetched directly from the vendor's official pricing page within the last 30 days. We re-verify the pricing of the top 500 tools weekly. 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 are derived from actual user reviews on G2, Capterra, Reddit, and our own review system, not generated from feature lists. Patterns must appear in at least three independent sources to be included.
• Comparison verdicts reflect the editor's judgment based on testing or documented evidence. We say "Tool A is the better pick for X" only when we can defend it; otherwise we say the choice depends on Y.
• Media mentions ingested from our 8,971-strong corpus are tied back to the original article URL, the publication, and the publish date. Tier 1 sources (TechCrunch, The Verge, Wired, etc.) are weighted more than tier 3 forum posts. The full list 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 and sponsored placements clearly labeled as such. Editorial coverage is never sold. Vendors cannot pay to appear in a guide's top picks, change their score, or move up a comparison ranking.

When a tool we recommend in a guide also happens to sponsor a Dupple newsletter, the guide carries a disclosure at the top of the relevant section. We track these overlaps explicitly so we can be honest about them rather than hoping no one notices.

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