How we evaluate
SaaS Tech Watch is buyer-side analysis of the SaaS and platform stack. We write for the person who will be on call when the integration breaks, not for the person who signs the purchase order — though we try to give them something they can forward.
What we publish
- Landscape notes — what a category contains, and the two or three axes on which products in it genuinely differ. No rankings, because a ranking needs a method.
- Comparisons — criteria and weighting stated before any product is scored, with the bounds of the evaluation made explicit.
- Deep dives — how a product is built and where it breaks. Failure modes get equal space with capabilities.
- Signals — short analysis of a release, price change or acquisition: what changes for existing customers, and what is not yet knowable.
How we handle numbers
Every figure carries its source and date in the sentence. We distinguish three verbs and use them precisely: measured means we ran it, reported means a named party published it, derived means we calculated it from named inputs. A vendor’s own benchmark is never presented as an independent finding.
Where a vendor declines to publish pricing, that is the finding. We write it rather than estimating.
What we will not do
- Score a product on a criterion we did not apply to every product in the comparison.
- Compare against an unnamed “legacy vendor”.
- Reproduce an architecture diagram from vendor marketing without saying where it came from.
- Accept payment for coverage, placement or a rating. Nothing here is sponsored.
Corrections
Corrections sit at the top of the affected piece, dated, with what changed. If we got a version number, a limit or a price wrong, that is worth saying loudly — those are the details readers came for.
How this publication is produced
Research, drafting and production are AI-assisted under editorial direction. We say so plainly. The standards above are what make the output worth reading, and they apply regardless of how a draft was produced.