How the radar is built

What full and partial mean

Each platform gets one mark per department. The mark is about what the vendor sells, not about how well it works.

Full

The vendor sells a product for that department's core job, described on its own site.

Partial

It covers a slice of the job. The usual case is a request desk for HR or Finance teams, which sits on top of an HR system or a ledger rather than replacing it.

None

Nothing for that department.

The core jobs we look for:

• HR: the employee record, payroll, benefits and hiring.
• IT: either the service desk (tickets, incidents, requests) or devices and access (laptops, device management, sign-in and offboarding). Covering either in full counts as full, and each row says which.
• Finance: either the books (an accounting ledger) or spending (corporate cards, expenses, bills). Again, each row says which.

Where the radar puts each platform

The radar on the home page is drawn from the same marks. Each department pulls a platform toward its sector: a full mark pulls twice as hard as a partial one. A tool that does one job lands far out in that sector. A suite that covers everything is pulled equally in three directions and ends up in the middle, which is why Rippling sits at the centre and Ramp and Brex sit on the Finance rim.

Prices

We show the lowest price the vendor publishes, in the vendor's own unit: per employee, per user, per agent, or a base fee plus a fee per person. Those units are not comparable with each other, so the matrix deliberately has no "sort by price". If a vendor needs a sales conversation before it names a number, we write quote only.

Two prices did not come straight from the vendor, and both rows say so. gusto.com refused our automated check, so Gusto's figure comes from independent trackers that agree with each other. Zoho loads its all-employee price by script, so that figure comes from a tracker's July 2026 check; Zoho's page itself shows the US$90 flexible price.

AI features

We list only AI features the vendor names on its own pages, with their price when there is one (Freshservice's Freddy AI Copilot, for example, is an extra US$29 per agent a month). We do not rate how well any of it works until we have used it. Where we have not checked a vendor's AI pages yet, the cell says "Not checked yet" rather than guessing.

Who pays us

No one, as of 28 September 2026. The matrix has a Pays us column so you never have to wonder. If we join a vendor's affiliate programme, that cell changes the same day and the vendor's links are marked. Rows are ordered by coverage, then alphabetically. A commission never moves a row or changes a mark. The disclosure page has the details.

What we have not done

We have not run hands-on tests. There are no scores or star ratings on this site, and there will not be any until we have used each platform on a real test account. When a review goes up, it will say what we tested and for how long.

Sources

The page behind every row, with the date we read it. Vendors change prices often, so treat anything older than a few months as a starting point and confirm on the vendor's page before you buy.