1. Where the data comes from
Each of the 513 products has a structured record of up to 150 fields — pricing tiers, contract terms, integrations, compliance statements, regulatory clearances, company signals and category-specific capabilities. Every field that holds a fact is tied to a source: vendor documentation, pricing pages, press releases, FDA and other regulatory databases, and public review platforms. 6,061 sources are cited across the database, and each product page lists its own.
Records are re-verified on a rolling 90-day cycle; every page shows when it was last verified and when the next review is due.
2. Three states, never two
For every yes/no fact we store yes, no or unknown. Unknown is shown as a hatched “?” and is never presented as “no”. Most review sites silently treat a missing checkmark as a missing feature; we think that is the single biggest source of misleading comparisons.
3. The six software axes
| Axis | Weight | What it measures |
|---|---|---|
| Features | 25% | Share of the category-specific capabilities we could confirm in vendor documentation. |
| Integrations | 20% | Number of verified PMS/software integrations (log scale) plus a public/open API. |
| Pricing clarity | 15% | How much of the price is public, plus trials, free plans and contract flexibility. |
| Support | 10% | Support channels, included onboarding, training formats and dedicated account management. |
| Vendor stability | 15% | Years in market, company size and funding; discontinued/legacy status overrides. |
| User rating | 15% | Normalized G2/Capterra rating (0–5 scaled to 0–10); needs at least one public rating. |
Each axis is scored 0–10 from its inputs and carries a coverage value — the share of its inputs we could verify. An axis with no verified inputs is left blank (“no data”) and drawn as a gap on the radar, not as a zero.
4. From axes to one number
The overall index is a weighted mean of the axes. Where an axis is unknown, we substitute the neutral midpoint (5.0) at half weight. That keeps a product with thin documentation from outranking a well-documented one on two lucky axes, without punishing it as if every unknown were a failure. A product needs at least three measurable axes to receive an index; today 478 of 513 products qualify, with an average evidence weight of 69%.
Rankings within a category order products by index, then by data completeness. Discontinued products are never ranked.
5. Hardware is scored differently
| Axis | Weight | What it measures |
|---|---|---|
| Openness | 30% | Open STL/PLY export, no export fee and no mandatory subscription to keep using the device. |
| Regulatory | 20% | Clearances on file (FDA 510(k), CE-MDR, Health Canada, TGA) — more markets score higher. |
| Price clarity | 20% | Whether a list price or price range is publicly documented. |
| Spec disclosure | 15% | Share of type-relevant technical specs the vendor publishes. |
| Accuracy evidence | 15% | Independent accuracy studies with numeric trueness/precision. |
For scanners, printers and mills the questions that matter are lock-in and evidence: can you export open files, do you need a subscription to keep using the device, and is there independent accuracy data? Vendor accuracy claims are not counted as evidence.
6. What the index is not
- Not a hands-on test. Where we have tested a product ourselves the page carries a “hands-on tested” badge; we never imply testing we have not done.
- Not a measure of clinical quality. It measures documented capabilities, openness and evidence.
- Not influenced by money. Affiliate and partnership data live in a separate part of our data model that the scoring code cannot read. See our disclosure.
7. Publication threshold
Product pages below 70% data completeness are published for transparency but marked noindex until we fill the gaps, so search engines only surface profiles we stand behind. Today 349 of 513 profiles (68%) meet the threshold.
8. User ratings
Ratings are snapshots from G2 and Capterra, shown with date, sample size and a link. Capterra, GetApp and Software Advice share one review pool, so we count it once. Samples under 10 reviews are flagged as directional. We do not mark up third-party ratings as our own review stars.
9. Corrections
Vendors and readers can submit a correction with a source. Accepted corrections are reflected in the next build and the page’s “last verified” date.