Before a product enters the lab
We begin with the decision, not the product. What problem is a household trying to solve? Which attributes are meaningful? What does credible welfare or safety guidance say? Which claims can be measured, and which depend on individual preference?
A test plan is written before a verdict. It identifies the intended user, comparison set, conditions, measurements, observation period, safety limits, and known blind spots.
Every hands-on test record will include
- How the item was acquired: purchased, borrowed, or provided by a manufacturer.
- The exact model, size, and version tested.
- Test dates, duration, setting, and relevant household conditions.
- Applicable cat characteristics without exposing private household information.
- Measurements, original photographs, observations, failures, and uncertainty.
- Who may benefit, who should skip it, and what alternatives deserve consideration.
Category-specific methods
A fountain and a cat tree should not be judged by the same checklist. Before the first review in each category, Splintercat will publish a category method defining the tests that matter.
Climbing furniture
Usable platform dimensions, base footprint, assembly, fasteners, anchoring, movement under controlled load, surface wear, and cleaning access.
Water fountains
Usable capacity, sound under defined conditions, access, assembly, disassembly, cleaning burden, replacement parts, and power-loss behaviour.
Litter equipment
Interior dimensions, entry access, tracking, leakage, cleaning, noise where relevant, consumables, and maintenance time.
Carriers
Opening access, structure, closure security, usable interior space, handling, vehicle fit, cleaning, storage, and manufacturer-rated limits.
Content labels
Every substantive page carries one of these labels near the top:
Direct evaluation
Physical testing under a published method, with original evidence and stated limits.
Evidence synthesis
Credible guidance, manuals, specifications, and cited sources. No rating or test badge.
Everyday context
A clearly labelled, longer-term observation. Useful, but not universal proof.
Marketplace Watch is a separate research label for dated retailer-popularity snapshots. It records what is selling and adds practical buying checks; it is not a Splintercat rating, endorsement, or hands-on review.
Research Shortlist and Exact Product Record are Research Guide formats. A shortlist compares distinct use cases using published facts. A product record identifies the exact model, fit boundary, cautions, maintenance, dated sources, and retailer destinations. Neither label claims hands-on experience.
Ratings and “best” claims
Splintercat will not launch a numerical score merely to make a page look complete. A category receives ratings only after a consistent method and a meaningful comparison set exist. The page will explain what the score represents.
“Best” requires a defined use case. A stable tower for a large, active cat in a permanent room is a different decision from a compact scratcher for a renter with limited floor space.
Money and editorial control
Manufacturers cannot pay for a ranking or edit a verdict. Sponsorship, samples, discounts, affiliate relationships, and other material connections will be disclosed beside the relevant content. Affiliate compensation may affect which retailer links are offered; it does not change the published testing method.
See the affiliate disclosure and editorial policy for the operating rules.
Updates, failures, and corrections
Products change. Prices, availability, specifications, and long-term performance change too. Material updates will carry a real revision date and a note explaining what changed. If an error affects a conclusion, the correction will be visible rather than silently buried.
The lab is being built. Pages marked Hands-on Tested will not appear until an actual test record exists. Current publishing is limited to clearly labelled Research Guides—including research shortlists and exact product records—and Marketplace Watch snapshots.