Methodology
How we choose and score
The 0-to-10 scores in the charts are ours, not the manufacturer's. They follow the same rule every time, so two models can actually be compared instead of judged by how well their listing is written.
Where the data comes from
Every page starts from the product's real Amazon.com listing: title, price, average rating, review count, features, spec table and the reviews the page itself shows. The photos are the product's own, hosted on our server.
We then cross-check that listing against the manufacturer's official site and, where it is needed, with specialist retailers. It is a necessary step, because Amazon listings often arrive incomplete or flatly self-contradictory: two JBL speakers on this site claim 180 vatios in a device that fits in your hand, and a third from the same family claims 4.8. Not one of the ten publishes its maximum volume in decibels, which is the only figure that tells you whether you will hear it while riding. Where they conflict, the manufacturer's data wins and we write it on the page.
What we do when a figure is missing
We leave it empty. You will see a «—» in the table and a note explaining it. We do not estimate, we do not round to a believable number and we do not copy the figure from a similar model. If a brand does not publish its speaker's battery life at medium volume, that cell stays blank and the matching axis on the chart is marked «no data».
It looks less impressive, but it's the only way the comparison is worth anything.
The six scores
The first four are computed from manufacturer data; the last two carry the editor's judgement, checked against verified buyer reviews. Every score is recalculated whenever a new speaker is added, so they stay comparable. When a manufacturer does not publish a figure — and RMS power is the one they hide most, because peak watts look bigger — that cell stays empty and the axis is not scored. We would rather show the gap than fill it with a guess.
Battery life
From the claimed battery life. Read it carefully: it is measured at 50% volume with the lights off, and of the ten listings on this site exactly ONE says so. 6 h scores 0 and 24 h or more scores 10.
Sound
Number and type of drivers, passive radiator, published frequency response and whether there is an app EQ. This is not an acoustic measurement: it is what can be checked against the listing.
Power
The power the manufacturer publishes, except when it cannot be true: two JBLs on this site claim 180 watts for speakers that fit in your hand, and those cells stay empty. Amazon never says whether its figure is RMS or peak, so the column cannot be used for comparing — and the note on each page says so.
Portability
Weight and size, and above all how it attaches. A 280 g speaker with a moulded carabiner goes on a handlebar; a 998 g one with a strap goes in the backpack and not on the bike.
Durability
Published IP rating, whether it floats, and the warranty. IPX5 (splashes) scores far below IP67, and a rating that appears in the bullets rather than the spec table counts the same: the manufacturer still said it.
Value
The other five axes added up and divided by the price, normalised within its category. An expensive model only scores well if it genuinely gives you more.
How we make money
From Amazon affiliate commissions. If you buy a speaker through our links, Amazon pays us a percentage and you pay exactly what you would have paid going there directly.
That is an obvious conflict of interest, and the only way to handle it is in the open: we write the cons of every model in the same detail as the pros, we summarise the negative reviews as carefully as the positive ones, and we never order anything by what pays us most. If a speaker has a known problem, it is written on its page.
Read the full affiliate disclosure.
Prices go stale
The prices you see are a snapshot from the date shown next to each one. Amazon changes them constantly. Treat them as a guide for comparing models, never as the price you'll pay: you only know that when you open the link.