# 37 terms a buyer meets, each with a sourced example

Glossary, read 8 Nov 2026

**The short answer.** Thirty-five terms a judge uses at ringside, from weighted total, margin and tie-break to approval before action, read-only access and fact line, each defined in two sentences and followed by an example from a vendor page read on 8 Nov 2026 with its source. Every term names the criterion it belongs to.

- [Agent audit log](/glossary/agent-audit-log/): A log that records what an agent did, on which record, when, and who approved it, so that an auditor can tell a human change from an agent change.
- [Agents under control](/glossary/agents-under-control/): The bout criterion weighted 25: whether an agent on the platform runs a People process end to end, and whether a person approves before it goes live, sees what it did in a log, can switch it off and can keep it from sensitive fields.
- [Approval step](/glossary/approval-step/): The point at which an agent stops and a person says yes before a change is written to the record or a message leaves the system.
- [Bake-off](/glossary/bake-off/): A bout a buyer referees: two vendors, one anonymized data file, the same script, and a score for what each did on that file during the session rather than on a slide.
- [Base fee](/glossary/base-fee/): A fixed monthly charge added to the per-person price, so that the bill at ten people is far higher per head than at a hundred.
- [Bout](/glossary/bout/): A bout is one pair of platforms scored against each other on six criteria with published weights, producing two weighted totals, a winner and a margin.
- [Bout weights](/glossary/bout-weights/): The six weights a bout uses: build 30, agents 25, openness 15, time to value 10, pricing 10 and fit 10, totaling 100.
- [Build](/glossary/build/): The bout criterion weighted 30: what a People admin can build on the platform without the vendor, scored from 0 to 10 on reports, dashboards, tools and workflows made from a plain-language description on live data, and on whether the thing built can be shared, edited and kept current.
- [Demo Script](/glossary/demo-script/): The fifteen questions on this site that turn a vendor demo into a bout: each maps to a bout criterion or to the exit, and each question page carries every vendor's published answer with a status and a source, so the buyer knows before the call which answers should take ten seconds.
- [Employer of record](/glossary/employer-of-record/): A vendor that employs a worker in a country where the customer has no entity and leases the worker back, priced per employee per month far above an HRIS seat.
- [Fit](/glossary/fit/): The bout criterion weighted 10: how well the platform's shape matches company shapes, scored on multi-entity records, per-office policies, country payroll, size bands and modules such as hiring and performance.
- [Flat rate](/glossary/flat-rate/): A single monthly price for a company up to a stated size, regardless of headcount.
- [Implementation fee](/glossary/implementation-fee/): A one-time charge for setting the platform up, separate from the subscription, that may be published, mentioned in the terms without a number, or absent.
- [Index order](/glossary/index-order/): The ranking of the fourteen platforms on the shared AI-Era HRIS Index, used on this site as the tiebreaker when two weighted totals are equal.
- [Margin](/glossary/margin/): The difference between the two weighted totals in a bout, to one decimal place.
- [MCP server](/glossary/mcp-server/): A server a vendor publishes so that an outside AI client such as Claude, ChatGPT or Gemini can read, and sometimes write, the platform's data through the Model Context Protocol, under the user's own permissions.
- [Minimum subscription](/glossary/minimum-subscription/): The smallest number of seats, or the smallest annual amount, a customer pays for during a term whether or not the seats are used.
- [None found](/glossary/none-found/): The phrase this site uses when a capability was looked for on a vendor's pages on the read date and not found.
- [Off switch](/glossary/off-switch/): An account setting that turns an AI feature off for the company, a role or a department.
- [Openness](/glossary/openness/): The bout criterion weighted 15: how far outside tools, including AI agents, can reach the platform's data.
- [Per-office policy](/glossary/per-office-policy/): A time-off, working-time or payroll policy that applies to one location or entity and not to another, assigned to a person by where they sit.
- [Pricing transparency](/glossary/pricing-transparency/): The bout criterion weighted 10: whether the price per seat, the plan contents, the add-on structure, the minimum and the implementation fee are published.
- [Read date](/glossary/read-date/): The date on which a vendor page was read for the score or fact that cites it, printed in the eyebrow of every page and beside every source.
- [Read-only MCP](/glossary/read-only-mcp/): An MCP server whose tools return data and never change it.
- [Record](/glossary/record/): A platform's wins and losses across its thirteen bouts, written wins first.
- [Release Ledger](/glossary/release-ledger/): The dated list of product releases this network reads for each vendor, tagged for AI, builder, agent and analytics content.
- [Sandbox](/glossary/sandbox/): A copy of the account where an admin can try a workflow, an agent or an import without touching live records.
- [Scorecard](/glossary/scorecard/): The fourteen by fourteen grid on this site where each cell holds the winner and margin of the bout between the row platform and the column platform.
- [Sensitive fields](/glossary/sensitive-fields/): Fields such as salary, medical notes and performance ratings that a platform can hide from an agent, a view or a user even when the rest of the record is visible.
- [Situation verdict](/glossary/situation-verdict/): A bout re-scored on the situation's own weights, so a first HRIS, a second HRIS, a multi-entity company, an AI-native company or a professional services firm sees its own margin.
- [Time to value](/glossary/time-to-value/): The bout criterion weighted 10: how quickly a company is running on the platform, scored from the published timeline, the import route, the presence of a free trial or sandbox, and whether an implementation fee exists and is published.
- [Trust](/glossary/trust/): The Index criterion for certificates, audit logs, sensitive-field controls, data retention after termination and AI governance, scored 0 to 10.
- [Trust fact line](/glossary/trust-fact-line/): The line on every bout page that prints both platforms' trust scores and the certificates behind them without adding them to the total.
- [Vendor claim](/glossary/vendor-claim/): A statement on a vendor's own page about speed, scale or outcome that the site could not check against a second source.
- [Vibe View](/glossary/vibe-view/): Shapes's name for a dashboard, report, tool or workflow built from a plain-language description on live, permission-aware data.
- [Weighted total](/glossary/weighted-total/): The number printed large on a bout card: each criterion score from 0 to 10 multiplied by its bout weight, summed and divided by 100, to one decimal place.
- [Winner mark](/glossary/winner-mark/): The triangle in the center column of a bout card that points at the platform with the higher score on that criterion.

Last reviewed 10 Nov 2026. Canonical: https://hriscomparison.com/glossary/
