How we choose picks

What we do

Four inputs, in order of how much weight each carries.

1. Published specifications

Capacity, port count, materials, dimensions, mounting, what is in the box, and the warranty. These come from the manufacturer’s own product page or a major retailer listing, and every figure we publish is cited on the page it appears on. Where a manufacturer does not publish a figure, we leave the spec out rather than filling it in.

2. How the feeder comes apart

The criterion we weight most heavily, and the one most roundups skip. We read what each maker says about disassembly and cleaning — whether the lid lifts or threads, whether the base separates from the reservoir, whether the ports come out, whether a standard bottle brush fits through the opening.

Why this outranks almost everything else: nectar needs changing every three to five days in moderate weather and daily in very hot weather, with a full break-down wash at least weekly, all season. Cornell Lab describes the purpose of that schedule as preventing mold and deadly fermentation.

That schedule is fixed. Whether it actually gets kept is not — it depends on friction. A two-part saucer washes in about two minutes; a sealed decorative bottle takes ten and involves a fight. Across sixty-odd washes in a season, that difference decides whether the feeder stays safe to drink from. So we treat cleaning access as a safety specification rather than a convenience, and we score it on every product. The cadence in full.

3. Port geometry and insect exclusion

Whether the design physically prevents insects reaching the nectar, rather than merely discouraging them. A saucer holds nectar below the ports, out of reach of a bee’s short tongue while a hummingbird’s long one still reaches it. A fitted guard such as a Nectar-Guard tip closes behind a bill. Both are mechanisms. Yellow decoration and vague “bee-resistant” marketing are not. The bees page.

4. Live pricing

Prices render from a live data layer and expire after 48 hours. Where there is no in-date price, the page says “check price on Amazon” rather than showing a remembered, estimated or list number.

What we explicitly do not do

• We do not test these feeders. We own none of them. There is no lab, no yard, no trial and no test data behind any recommendation here.
• We hold no ornithological credentials. Scooter M. writes this site as an enthusiast, not an expert.
• We do not accept free products. Nothing is sent to us, so nothing is owed.
• We do not sell placements. No brand can pay to appear here or to move up a list.
• We do not publish invented figures. No fabricated prices, ratings, review counts or test results, ever.

Nearly everyone in this category claims to have tested a dozen units in a backyard. We have not tested any, and saying so plainly is the point — a claim you cannot check is worth less than a method you can repeat.

How a pick earns its category

Every roundup awards specific categories rather than a single ranking, because the right feeder genuinely depends on the yard.

• Best overall — the best balance of cleaning access, insect exclusion, no-drip design and sensible capacity. Cleaning access dominates.
• Best budget — the cheapest feeder that still genuinely comes apart. We will not recommend a cheap feeder that cannot be cleaned properly, at any price.
• Best for bees — a physical exclusion mechanism, not a deterrent claim.
• Best glass — a glass reservoir plus a base that actually detaches. Sealed decorative glass does not qualify.
• Best camera feeder — judged on camera hardware, then penalized honestly for the cleaning friction that electronics adds.

The score out of ten

Every pick carries an editorial score. It reflects our reading of the published specifications against the criteria above, weighted toward cleaning access.

It is not a customer rating and not a test result, and it never appears in this site’s structured data — marking an editorial opinion up as a product review rating would contradict the plain statement that we have not tested these feeders.

What we would skip

Every roundup names at least one product or product type we would not buy, with the reason. Every single-model review names the buyer who should walk away. A page with no negatives is an advertisement, and this is the standing rule that stops that happening.

How to challenge us

If a specification here is wrong, a link is dead, or a pick looks indefensible, tell us and we will fix it or explain why we disagree. The address is info@wingandwindow.com, and corrections are logged on the page they affect. Our full standards are on the editorial policy page.

How we are funded

Affiliate commission on qualifying purchases through our links, at no extra cost to you. That is the whole model, and it does not change what we recommend — our top overall pick is a $25 plastic dish. The full disclosure.