How decisions get made here.

How topics are chosen

We start with a distinct problem someone is trying to solve: plan a system, understand a constraint, complete a project, troubleshoot a failure, or compare genuinely different paths. Search-result research can help identify the language and competing coverage, but it does not prove demand, difficulty, or correctness. Unverified keyword metrics remain blank.

We avoid pages created only by swapping product names or locations. A new page needs a materially different intent, decision, workflow, or evidence set.

Selection starts with the job

We define the use case, meaningful constraints, compatibility requirements, and attributes that can change a decision before selecting products. There is no universal “best.” A product belongs in a guide or project only when it helps answer the stated problem.

What each evidence label means

• Manufacturer-stated: a current product page, technical catalog, or instruction document supports the value.
• Retailer-stated: a merchant supplies the claim; it is not treated as primary technical evidence.
• Independent measurement: a disclosed test method and result exist. The launch catalog currently makes no such claim.
• Original editorial analysis: criteria, calculations, or tradeoff reasoning derived from identified inputs.
• Unverified: evidence is missing or insufficient; the claim is excluded, labeled, or kept in review.

Sources and specification verification

Published product records use primary sources such as manufacturer pages and current technical catalogs, store source URLs, and carry a verification date. Missing facts stay missing. The initial catalog contains real Kichler low-voltage landscape products; manufacturer photography is not copied, so abstract editorial visuals remain in place until reuse permission or a licensed feed exists.

Testing language

We do not describe a product, build, or tutorial as hands-on tested unless it was actually performed and the method can be explained. Research-based analysis is labeled as research-based. Manufacturer claims are not restated as independent findings. Review drafts explicitly say when no execution record exists.

Publication workflow

Content moves through draft, needs-review, verified, published, or archived states. Only useful, source-supported work in the published state is indexable and sitemap-eligible. An unassigned author or editor stays null; the site does not invent a byline or review event.

How matching works

The outdoor-lighting matcher considers only catalog products with documented system identity, VA load, operating-voltage range, transformer output taps, and capacity. A category or brand match alone never produces a transformer candidate. Results are a planning screen, not an installation approval.

Affiliate independence

Affiliate eligibility does not determine inclusion, compatibility, or rank. Merchant offers remain inactive until a complete destination and attribution are approved. Product pages can be useful with zero offers, and editorial pages do not require affiliate modules.

Corrections and freshness

Material updates change the editorial record date. If a specification, source, compatibility statement, calculation, or product relationship becomes unreliable, the record should be corrected or moved out of published state. Readers can use the contact page to report a specific issue and source.