Skip to content
ChainringClub

How we research

Everyone in this category says they tested twenty bikes. We have not tested any, and we say so. Here's what we do instead — and why we think it's better.

A notebook, calculator and pen on a desk under a work lamp.

1. What we don’t do

Chainring Club has no test lab, no workshop full of loaner bikes, no dynamometer and no staff. We have not ridden the bikes we rank. We have not worn the bib shorts. We have not run a single chain to destruction.

The publishers we compete with have. Some field seven named testers; one owns a dynamometer and has bought over a hundred e-bikes; another says it tested 130 e-bikes in twelve months. That is real work and we are not going to pretend we have done anything like it. Anyone in this category claiming to have ridden twenty bikes this year has spent money we haven’t.

We could paper over this. The phrasing is easy and the field is full of it: “in our testing”, “our experts”, “we put it through its paces”. It would take about ten minutes to salt every page with lines like that, and you would have no way of catching us. That is exactly why we don’t. A claim you can’t check is worth nothing, and a category where everyone makes the same uncheckable claim is a category where the claim has stopped meaning anything.

2. What we do instead

Four things, all of which you can verify without trusting us at all:

  • We compile published specificationsand attribute them to whoever published them. When we say a battery is 624Wh, that is the seller’s claim on their own listing, and we say so in those words.
  • We do the arithmetic in public.Every cost figure on this site shows its inputs, its rate, and the date the rate was read. You can re-run it. If we’ve fumbled a decimal, you’ll find it.
  • We cite third-party test data — with its conflicts disclosed. See section 5. This is the part the rest of the category gets wrong.
  • We say when nobody has published the number.“Not published by the manufacturer” appears on this site wherever it’s true. A publisher willing to print what it doesn’t know is more useful than one that never has a gap.

3. Why there are no scores on this site

Every competing roundup gives each product a number out of ten, or five stars, or a weighted rubric across six dimensions. We don’t, and we want to be direct about why: a score is a measurement, and we measure nothing.

To score a helmet 8.4 we would need data we do not have. We could generate the number anyway — pick some weights, apply them to specs, publish the output. It would look rigorous. It would be arithmetic performed on judgements we invented, and dressing a judgement up as a measurement is a way of lying that survives a fact-check.

So our roundups are ranked and argued instead. Each pick says who it’s for, what the seller publishes about it, and — mandatory on every product — who should not buy it. You can disagree with a ranking and see our reasoning. You cannot disagree with an 8.4.

4. How we compute cost-to-run

This is the thing we do that nobody else in the category does. Not one of the publishers we researched publishes e-bike cost-per-mile or drivetrain running costs, so we built it.

The rule is enforced in the code itself: a number cannot enter a calculation on this site without a source URL and the date it was read. There is no way to pass in a figure we “know”. If we don’t have a source, the page renders “not published” instead of a number.

Worked example — the electricity rate every e-bike page uses:

Note what that example does notdo: it doesn’t apply a charging-efficiency factor. Charging is not 100% efficient, so the true wall-socket cost is somewhat higher than the energy that lands in the pack. We don’t have a published efficiency figure for any specific charger, so rather than invent a plausible-sounding 85%, we compute the energy in the pack and tell you that’s what we’ve done. That is the whole methodology in one decision.

5. Our sources, and their conflicts

In order of how much weight we give them:

  1. Manufacturer specifications and service manuals — for weight, capacity, torque and claimed range. Linked at source, never relayed.
  2. Standards bodies — CPSC, EN, ISO. For any safety claim we cite the standard, never a marketing page.
  3. Published third-party test data — only with its conflicts disclosed.
  4. Our own arithmetic — always with the inputs shown.

The Zero Friction Cycling disclosure

Almost every chain and lube recommendation on the internet traces back to Zero Friction Cycling, which has run over 300,000km of controlled wear testing and publishes the results publicly. It is the most extensive dataset in the category and we rely on it.

Zero Friction Cycling also runs a shop that sells chains and lubricants, including products that test well in its own results. We verified that by reading their site directly rather than taking anyone’s word for it. That is not an accusation — a shop that only stocks what it rates is a coherent way to run a business. But a testing operation that retails the winners has a commercial interest in its own results, and you should know that before you weigh the data.

The publishers ranking above us describe ZFC’s data as independent lab testing. It is rigorous and it is public. It is not independent, and we will not repeat that description. We cite ZFC, we link its raw data so you can judge it yourself, and we say this every single time. There is also a public methodological critique of ZFC published by Hambini Engineering; we could not retrieve that page to verify its claims, so we neither characterise nor adjudicate it — we only note that the dispute exists.

6. Prices, and why they sometimes disappear

Prices on this site come live from Amazon’s API. We never type a price into an article, because a typed price is wrong within a week and stays wrong forever — which is the single most common rot in this category.

Right now we’re showing live prices for 77 products, last verified Jul 17, 2026. If our price data goes more than 48 hours without a refresh, every number on the site vanishes automaticallyand the buttons fall back to “Check price on Amazon”. We would rather show you nothing than show you a number that has quietly gone stale.

7. How we make money

Amazon Associates. If you buy through a link here, we earn a commission at no extra cost to you. That is the entire business model, and the full disclosure is here.

Three things it does not buy: commission rates never affect a ranking; we accept no sponsored placements; we accept no free products. The last one is easy for us to promise because nobody is sending free bikes to a site that admits it doesn’t test them — but it’s true, and it stays true. Every roundup on this site names at least one product to skip, which is the cheapest way to prove a recommendation is a recommendation and not an advert.

8. What would change a pick

New published data. A sustained price move. A spec change or a new model. A safety recall. A reader showing us we’re wrong.

Every page carries the date it was last updated, and that date only moves when something actually changed. Bumping dates to look fresh is standard practice in this category and it is a small, constant lie.

Corrections

If we’ve got something wrong — a spec, a link, a piece of arithmetic — tell us at info@chainringclub.com and we will fix it and log the correction here with the date. Being visibly correctable is worth more than being confidently wrong.

Corrections log: nothing yet. The site launched in July 2026. When there is something here, it will stay here.

Who writes this

One person: Stephen V.He is not a mechanic, an engineer, a coach or a certified bike fitter, and you will never see a credential after his name on this site, because he doesn’t have one. He rides, he wrenches on his own bikes, and he reads spec sheets properly. There is no editorial team. Where a page draws on his actual experience — the how-to guides, mostly — it says so, and it is about him, never about a product we don’t own.