How Stackwell makes a match
Stackwell is a focused research desk, not a paid placement directory. The catalog is intentionally smaller than the AI market so every listed tool can be assessed against a consistent set of criteria.
Last updated 24 August 2026
How tools enter the catalog
Tools are assessed for a clear agency or creative-studio use case, practical capability, pricing clarity, adoption effort, and whether they add something meaningfully different from the rest of the catalog.
A tool may be left out when its use case is too vague, pricing cannot be verified, it is no longer available, or it substantially duplicates a stronger option. The homepage’s “What we left out, and why” section shows examples of this editorial filter.
How matches are scored
The matching engine combines five named factors: capability fit, workflow fit, value for money, ease of adoption, and stack compatibility. Your selected business type, constraints, budget band, existing tools, and time drains affect the result.
A match score is a relative fit signal inside Stackwell’s catalog, not a statistical prediction that a tool will succeed for every team.
How time-saved estimates work
Time saved is a directional estimate, not a measured productivity result. It uses the weekly workload you select, the tool’s match score, and a fixed baseline for its category.
There is no separate efficiency percentage for each category. Every category uses the same score-adjusted recovery rate: 55% of the match score, capped at 50% of the stated workload. For example, an 80% match uses a 44% recovery rate before the remaining adjustments.
Monthly category baselines: productivity and automation 10 hours; writing and content 8; HR and recruiting 7; sales and CRM 6; meetings and notes 6; design 5; data and analytics 5; customer support 5; research 5; marketing 4; video and audio 4; finance and operations 4. An unlisted category defaults to 5 hours.
The calculation weights the category baseline at 35% and the stated workload estimate at 65%, then adjusts the result by match strength. These are internal assumptions rather than independently measured vendor claims.
Sanity checks limit each tool to 40 hours per month. A complete recommended stack is also capped at the total monthly workload you stated, calculated as weekly hours × 4.3; individual estimates are reduced proportionally when that cap applies.
Pricing and review checks
Catalog records carry a last-reviewed date. Pricing is checked against public vendor sources where available, with quote-only, unpublished, and retired products shown as such rather than given invented prices.
Read the pricing and review policy for assumptions, regional billing treatment, and how to report an issue.
What independence means here
No vendor can pay to rank higher in Stackwell. Some links may earn a commission, but commercial relationships do not change scoring, inclusion, or recommendation order. Read the affiliate disclosure for the full policy.