How we rank

Recommendations come from published rules, not payments. Vendors cannot buy a better position.

Where the data comes from

• Vendors' own public websites, integration pages and API docs.
• Official launch feeds such as the Product Hunt API.
• Vendors' answers to our integration questionnaire, and edits vendors make in the vendor portal.

AI reads each vendor's public pages and writes the listing from what they say; listings are published automatically, without a person checking them first. If you spot an error, report it on the product page: each report triggers a fresh check against the vendor's own pages. Listings are also re-checked on a rolling basis, oldest first.

Every integration shows its source and when we last checked it. Lines on the map show how sure we are:

• Solid — confirmed by the vendor.
• Dashed — found on the vendor's website or docs.
• Dotted — inferred or unknown; check before you rely on it.

How a tool is picked

1. Your checked needs map to capabilities. Anything a tool you already use covers is not filled again.
2. A candidate must cover the capability, fit your company size and work in every country you named (global tools always qualify).
3. Candidates are scored: 20 points per need covered, 40 for a native or middleware connection to your core HR system, 10 for each other tool in your stack it connects to, 15 if the vendor confirmed the listing, and 10 if it covers more than one of your needs, plus 5 for each technical requirement you gave that its listing shows it meets (a certification, an open API, or a connection to a system you named). Records not verified in 12 months lose 20.
4. The highest score is the pick; the next 3 are alternatives. Ties go to the tool that covers more of your needs, then the one with more known connections, then the most recently verified.

Compatibility score

Every tool in your stack, and every alternative, shows a match from 0 to 100% for your requirements. It is a weighted average of:

• 55% the needs you named in the tool's category that it covers;
• 20% how many of your countries it works in;
• 10% whether it is built for a company of your size;
• 15% its connection to your core HR system (half credit when the connection isn't confirmed); for the core HR system itself, how many of your other tools it connects to;
• 10%, only when you gave technical requirements: how many its listing shows it meets. A requirement our data doesn't mention gets half credit and appears in the questions to ask the vendor; we never assume a tool lacks it.

Tools you already use are scored the same way and come with alternatives, so you can see whether something fits you better. The score uses only what a tool covers; it says nothing about price or quality.

The "AI agent" label

A tool gets this label when its own website describes an AI agent that does HR work itself, such as sourcing and screening candidates, answering employee questions or running reviews. Tools that only add AI features, such as AI writing help, don't get it. The label doesn't change how a tool is ranked. See all HR AI agents.

Privacy

• No account and no tracking cookies.
• Your answers stay in your browser. They are stored only if you create a share link, and never with your name or email.
• If you describe your situation in your own words, the text is sent once to fill in the checklist (using an AI model when available) and is not stored.
• We count anonymous events (a stack built, a link shared) to see whether the site is useful.
• Vendor outreach uses public business contacts and every email has an unsubscribe link.

Spot something wrong? Use "Report an error" on any product page, or submit a product.