About & Methodology

What PowerMatchLab is

PowerMatchLab compares portable power stations for the United States market and helps visitors size their needs. It is reader-supported: as an Amazon Associate I earn from qualifying purchases made through our links, at no cost to you. We do not accept payment for placement, ratings or reviews.

Who runs PowerMatchLab

PowerMatchLab is run by Bilal Siali, founder and editor. He researches manufacturer-published specifications and oversees the development of the transparent tools on this site. PowerMatchLab does not perform laboratory testing, and information that cannot be confirmed is identified as “Not verified” rather than estimated.

Three kinds of statement

Every number on the site is one of these, and we label which:

• Manufacturer claim — a specification published by the maker, recorded with a source label and a “last checked” date on each product page. We have not independently measured it. Direct manufacturer links are added when available.
• PowerMatchLab calculation — a figure we compute from your inputs or from published specs using the documented assumptions below (for example the Power Calculator result and runtime estimates).
• Editorial assessment — our judgement, such as the PowerMatch Score or the pros and cons framing. Explainable, but a judgement.

When a value is unknown we keep it empty internally and show “Not verified” where the interface needs something visible. We never substitute 0, false or “N/A” for a genuine unknown.

Power Calculator

The calculator is deterministic. For each device:

• continuous watts = running watts × quantity
• daily energy (Wh) = continuous watts × hours per day

Then across all devices:

• total energy demand = sum of daily energy × days of autonomy
• required usable capacity = total energy demand ÷ usable efficiency
• recommended minimum capacity = required usable capacity × (1 + reserve)
• required continuous output = sum of continuous watts (assumes simultaneous use)
• required surge = total running watts + the single largest startup surge above its own running watts

The example appliances are editable starting points, not universal wattages. Invalid entries (negative or non-numeric) are clamped and rows with zero power or quantity are ignored in the totals.

Default assumptions (all adjustable in the calculator)

Recommendation engine

After the calculation, every catalog product is classified in two separate steps — compatibility, then proportionality — because clearing the minimum bar is not the same thing as being a sensible recommendation. A 6,000 Wh station technically has enough capacity for a 400 Wh need, but recommending it first would bury genuinely well-matched, smaller options under enormous overkill.

Step 1 — Compatibility

A product fails compatibility, and is labeled Not Suitable with the exact reason, when it does not meet a verified hard requirement:

• Capacity below the usable energy your devices need.
• Continuous output below your simultaneous load.
• A verified surge rating below the estimated startup spike (when a real spike is expected).
• 240V, TT-30, or expansion capability explicitly absent when you marked it required.

An unverified field is never treated as a pass — it is shown as a named caveat instead (“240V capability is not verified for this unit”), and the product is labeled Possible Match when core capacity or output data is missing entirely, since compatibility itself cannot be confirmed either way.

Step 2 — Fit (proportionality)

Among compatible products, a sizing-fit function — documented in full in src/lib/size-fit.ts — classifies how proportionate each one is to your stated need, using the ratio of considered capacity to your recommended minimum, and rated output to your required continuous output:

• Best Fit — capacity within 1.0-1.5× your recommended minimum (already includes reserve headroom) and output within a normal range for that capacity tier.
• Good Fit — capacity 1.5-3.0× the recommended minimum: meaningful extra headroom, still proportionate.
• Oversized — capacity beyond 3.0× the recommended minimum (or output beyond 20× the requirement). Fully compatible, but disproportionate for the stated need — shown as a clearly labeled alternative, never mixed into the primary results.

These thresholds are deliberately round, stated numbers rather than a fitted model — there is no external standard for “how much extra capacity is too much,” so the reasoning is documented in code instead of hidden behind an opaque score. Ranking within a status never considers price, popularity, rating, or whether a product has an affiliate link — only verified specification fields from products.json.

PowerMatch Score

The score is an editorial assessment, normalised against the other products currently in the catalog — it answers “how does this compare with the others we list”, not “how good is this in absolute terms”. Each dimension is scored only when the underlying data exists for that product. If fewer than 3 dimensions can be scored, no overall number is published. A score of 100 means a product leads the catalog on every dimension that could be scored for it — it does not mean the product is objectively flawless, that every one of its specification fields has been independently verified, or that it outperforms products outside this catalog.

Noise Level and Value for Money are deliberately not scored: the catalog carries no noise measurements and no verified prices, so scoring them would mean inventing data. Radar charts are only drawn when every plotted series has at least three real dimensions.

Runtime estimates

Product pages show an estimated runtime table using:

• estimated runtime (h) = capacity_wh × 0.85 ÷ device watts

This is a calculation with an assumed 85% usable efficiency, not a measurement. Real runtime depends on the device, its duty cycle, temperature and the unit’s age. We label these “estimated”, never “guaranteed”.

Amazon links

Each product record can store a direct Amazon product URL and, separately, an Amazon Associates affiliate URL. Only the affiliate URL is ever used as a clickable “Check Price on Amazon” destination. The direct product URL exists purely for our own research — confirming which real Amazon listing a catalog entry corresponds to — and is never rendered as a purchase link, no matter how confident that match is. When a product has no affiliate URL yet, no “Check Price on Amazon” link is shown at all; the page says so plainly instead of substituting the research URL or any other link. Amazon Associates commissions fund the site, but they do not affect which products are listed, the PowerMatch Score, or any recommendation — every product is scored and classified from the same dataset whether or not it currently carries an affiliate link. See the Affiliate Disclosure.

How products are added and removed

A product joins the catalog once it has a working, direct Amazon.com listing for that exact model and manufacturer specifications we can attribute to a named manufacturer source. A product is removed if its exact listing stops being purchasable for that model — for example the listing goes inactive, or the link now redirects to a different model or a generic search page. We do not substitute a similar model, variant, or bundle in a removed product’s place, and we remove its references everywhere on the site (related products, comparisons, guides, calculator, sitemap) rather than leave a broken link or an orphaned page. Full detail is in the Editorial Policy.

How specs and links are reviewed

Each product record carries an official_source label and a last_verified date. A field with no confirmed manufacturer value is shown as “Not verified” rather than estimated or borrowed from a similar model. Reviews happen when a product is added, when a reader flags something, or when we revisit a listing for another reason — never on a fixed schedule we could misrepresent as continuous monitoring.

Corrections

If a specification here is wrong or out of date, that is a bug. The canonical dataset is products.json; corrections there flow through the whole site. Report one to contact@powermatchlab.com — see Contact for what else that address is for, and the Editorial Policy for how we handle a report.