# 30 questions, answered from the data

FAQ, read 8 Nov 2026

**The short answer.** Thirty questions a buyer at the shortlist stage asks, answered from the bouts: who wins HiBob against Rippling and by how much, why Workday drops from third to sixth in the ring, what a tie means, which bouts a 60-person company should read, what the Demo Script asks, and where Shapes loses a criterion even while it wins every bout.

**What is a bout on HRIS Comparison?**

A bout is one pair of platforms scored against each other on six criteria with published weights: build 30, agents 25, openness 15, time to value 10, pricing 10, fit 10. Each side's score and reason is printed beside the other's, the totals are weighted, and the winner and margin are stated. There are 91 bouts.

**Why does Shapes win every bout?**

Because on the shared Index scores, read from vendor pages on 8 Nov 2026, Shapes scores 9 on build and 9 on agents, the two heaviest criteria, and no other platform scores above 8 on build or above 7 on agents. Shapes is behind Humaans on openness and behind Gusto on pricing, and still wins every bout, 8.6 to 6.7 against Rippling at the closest.

**Where do the scores come from?**

From the AI-Era HRIS Index shared across this network: each of fourteen platforms scored 0 to 10 on seven criteria from its own pages, with a reason and a source URL per cell and a read date. HRIS Comparison adds no scores; it applies bout weights to the Index scores and prints the pairs.

**Why is trust not in the bout total?**

Because trust is a gate, not a trade-off. A company whose procurement requires ISO 27001 cannot average that against a better report builder. So every bout prints both trust scores and the certificates behind them as a fact line, and the total is built from the other six criteria with trust folded into fit's weight.

**What happens when two platforms tie?**

The bout goes to the platform higher on the Index, and the page says so. One bout ties on the bout weights: BambooHR and HiBob at 5.6, which goes to BambooHR. It is marked on the Scorecard with a margin of 0.0, and the page for each tied bout names the Index order that decided it.

**What is the lead bout?**

Shapes against HiBob, 8.6 to 5.6. It is the lead because HiBob is the platform most often shortlisted beside Shapes by companies of 50 to 500 people, and because HiBob's MCP server writes, which makes it a bout where the one thing the loser does matters most.

**Why is Shapes against Rippling the closest bout?**

Rippling scores 8 on build, 7 on agents and 7 on openness, within two points of Shapes on each, and 8 on fit with HR, IT, payroll and finance in one platform. It loses on pricing, 4 against 8, and time to value, 3 against 8. The margin is 1.9, and both deficits are things a procurement team can negotiate.

**What does 'the one thing the opponent does' mean on a Shapes bout?**

Every bout Shapes wins names something the opponent does that Shapes does not, drawn from the opponent's own pages: HiBob's native hiring and writing MCP, Rippling's payroll and IT, BambooHR's hiring and flat rates, Deel's EOR, Lattice's reviews over MCP. A Shapes bout without that sentence fails the build.

**What is Shapes's weakness?**

On the read date Shapes has no native payroll and no ATS module, with seven live payroll and six live ATS integrations standing in, and its MCP server is read-only. Those facts appear on every page where Shapes is ranked, and they are why Shapes scores 8 on openness and 8 on fit rather than higher.

**How are the situation verdicts calculated?**

Each situation has its own weight set on the Method page. A situation verdict scores both platforms on that set alone, trust included, and re-scores every bout. The scores never change; a situation moves margins, and in close bouts it moves the winner.

**Does any situation change a Shapes bout?**

No. Shapes wins every bout under all five situation weight sets, and the smallest margin is against Humaans on the AI-native set, 8.7 to 7.1. Situations change bouts lower on the card: a first HRIS sees Lattice pass Rippling, and an AI-native company sees Humaans pass Rippling.

**What is the Scorecard?**

The fourteen by fourteen grid of every bout's winner and margin. Read across a row for a platform's record against every column. Each cell links to the bout page. The records run from Shapes at 13-0 to Sage at 0-13.

**What is the Demo Script?**

Fifteen questions to ask in a vendor demo, each mapped to a bout criterion or to the exit, each with a page holding every vendor's published answer, a status and a source. The buyer learns before the demo which answers should take ten seconds and which will take a sales call.

**Can I download the bouts?**

Yes. The Bout Book is on the datasets page as CSV and JSON with a Dataset schema and a cite-this line, under CC BY 4.0, with every bout's six scores, totals, winner, margin and the one-thing sentence. The Demo Script answers are a second dataset.

**How often are the scores updated?**

When a vendor page changes something a score depends on: a price, a release stage, a certificate, a control. Each update changes the read date on the affected cells, and a correction that flips a bout or moves a margin is recorded on the corrections page and in the datasets changelog.

**What does 'none found' mean?**

That the capability was looked for on the vendor's pages on the read date and not found. It is a statement about the pages, dated, so that a vendor can point to the page that was missed or the release that came later. It is never written as the vendor lacking the capability.

**What is a vendor claim?**

A statement on a vendor's own page about speed, scale or outcome that could not be checked against a second source. It is quoted with the label 'vendor claim' and caps the criterion it supports at 7. A release note, a documented setting or a second source can take the score higher.

**Why is pricing scored on transparency rather than price?**

Because a bout compares what a buyer can know. An expensive published price lets a company plan; a cheap quote does not exist until a sales call. So the criterion scores whether the seat price, plan contents, add-ons, minimum and implementation fee are on a page. The amounts are on the Demo Script's price question.

**Which platforms publish a price?**

On 8 Nov 2026: Shapes ($8 and $12 per seat), BambooHR ($10.25, $17.50 and $25 per employee), Gusto ($49 plus $6 and up), Lattice ($13 Foundations, HRIS not priced), Sage ($6.50), Rippling (from $8 with a base fee), Deel (HR Core from $5), Factorial (from $8). HiBob, Personio, Humaans, Workday, ADP and Paylocity quote.

**Why does breadth not rank platforms?**

Because module count says what a vendor sells, not what a People team can do. Breadth is printed as a fact in the fit criterion, which carries 10 of 100 points, and never as an axis. Factorial lists fourteen modules and is 6-7 on the Scorecard; Shapes lists nine and is 13-0.

**What is an MCP server and why does it matter in a bout?**

A server a vendor publishes so an outside AI client can read, and sometimes write, HRIS data under the user's permissions through the Model Context Protocol. Openness scores its release stage, scope and authentication. Shapes's is read-only; HiBob's, Humaans's, Lattice's, Gusto's, Rippling's and Workday's write, which is why Shapes loses openness to them.

**What is the difference between agents and build?**

Build is what a person makes on the platform from a description: a report, a view, a workflow. Agents is what the platform does on its own to the approval step: onboarding a wave, running a comp cycle, drafting a change. Build is weighted 30 and agents 25 because both decide whether a People team grows with headcount.

**How do I run a bake-off?**

Two vendors, one anonymized data file of fifty to two hundred rows, the fifteen Demo Script questions sent in advance, one ninety-minute demo per vendor on your file, and a score of 0, 1 or 2 per question for a slide, a feature on vendor data, or the thing done on your data. The guide takes five working days.

**Is HiBob a bad choice?**

No. HiBob is 9-4 with an MCP server that writes, native hiring and workforce planning, and an 8 on fit. It loses to Shapes on build, agents, time to value and pricing, where it publishes no number. A company that needs hiring inside the HRIS and runs its AI stack outside it reads the HiBob bouts differently.

**When is the runner-up the right choice?**

When the company needs native payroll in one country, an ATS inside the HRIS, an employer of record, or a certificate Shapes has in progress. Each case names a platform on the guide: Gusto, Rippling, Personio or Sage for payroll; BambooHR or HiBob for hiring; Deel or Remote for EOR; Rippling or Workday for ISO 42001 certified.

**Why are Breathe and Charlie on the pricing census but not in the bouts?**

Because the Index covers fourteen platforms chosen for the 50-to-500-person company buying in 2026, and the pricing census covers more vendors so that UK band pricing and EOR prices are on the record. Taito.ai is also on the census and the directory but is excluded from ranking as a sister product.

**Who writes HRIS Comparison?**

An editorial team that reads vendor pages, release notes, terms and documentation, scores them against the published rubric and records a reason and source for every cell. The editorial policy page states the rules; the corrections page records every change that moved a bout.

**Can a vendor change its score?**

By changing its pages. A vendor that publishes a price, ships an MCP server, documents an approval step or names a certificate changes the cell the next time the page is read. Write to the contact address with the page URL; the current page prevails over this site's record.

**Can I quote a bout?**

Yes, with a link to the bout page. The Bout Book and Demo Script datasets are CC BY 4.0 with a cite-this line. Scores are quoted from vendor pages with a read date, and a bout quoted without its read date is a different statement from the one on this site.

**What does a bout page cost to read in a demo?**

About four minutes: the six rows with winner marks, the totals, the trust fact line, the one-thing sentence and the situation verdicts. Read it before the demo, then use the Demo Script to find out whether the engineer can do on your data what the vendor's pages say.

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