The Trust Centre collects everything we publish about our own standards in one place, so you never have to hunt through footers to find out who is behind a recommendation. If you only have thirty seconds, read the eight commitments below — they are the rules the rest of the site is built on.
What does SirenPicks commit to?
Eight things, each with a mechanism behind it rather than a slogan: every score comes off an instrument on a published five-criterion protocol, body-safe material verification is a gate rather than an average, nothing on the site is for sale that changes a verdict, our ownership of the Slow Duo line is disclosed on every page and on every pick, every review carries a named accountable author, corrections are published rather than deleted, and analytics stays off until you allow it. Where we fall short of one of these, the fix is public.
- Verifiable, not aspirational. Each commitment below is paired with the mechanism enforcing it and the place you can check it yourself.
- Nothing can be bought — not a score, a ranking, a placement or a kind word. Brand and retail engagements are limited to bench time and data licensing.
- Our conflict is permanent and public. We own Slow Duo. It is labelled, tested identically, and stated in the footer of every page.
- When we are wrong, the correction is dated and visible — never a silent edit.
Our commitments, in writing
- Commitment 01
Every score comes off an instrument. Noise is a decibel reading at 30 cm, battery is a timed run, material is a verification against a spec sheet. Nothing in a review is estimated, and no number is invented to fill a gap.
- Commitment 02
The weights are public and unchanged. Noise 30%, material safety 25%, battery 20%, build and cleaning 15%, value 10%. A 7.4 means the same thing today as it did last year — that is the whole point of publishing them.
- Commitment 03
Material safety is a gate, not an average. If skin-contact material cannot be verified as platinum-cure silicone, the device does not appear on this site at any score.
- Commitment 04
Nothing here is for sale. No placements, no sponsored posts, no paid rankings, no display advertising, no link sales. Brands can pay for bench time and data licences; they cannot buy a verdict.
- Commitment 05
Our own conflict is stated on every page. We own the Slow Duo device line. It is disclosed in the footer of every page, labelled on every pick, and explained in the editorial policy.
- Commitment 06
Accounts are named, not anonymous. Every review carries a named author accountable for it. There is no generic house byline.
- Commitment 07
Corrections are published, not deleted. Errors are fixed in place and dated. A review that cannot be supported by its measurements is retracted, and the page says so.
- Commitment 08
Your reading is not the product. Analytics is optional and off until you allow it. No ad cookies, no cross-site tracking, no data sales — because in this category, privacy is not a marketing line.
Commitments, mechanisms and how to check them
The list above is the promise. This table is the enforcement — and the place you go to catch us out. If a mechanism here does not match what you find on the page, that is a bug worth reporting and we will fix it.
| Our commitment | How it is enforced | Where you verify it |
|---|---|---|
| Scores come from instruments, not opinions | A five-criterion protocol with published weights: noise 30%, material 25%, battery 20%, build and cleaning 15%, value 10%. | Match any review's scores against the weights on How We Test. |
| Body-safe materials or no review | Material verification is a gate, not a weighted average. Unverified devices are excluded entirely. | Look for the material row in any review's results table. |
| Nothing is for sale that changes a verdict | No sponsored slots, no advertorial, no paid rankings, no link sales exist on the site — there is nothing to buy. | Try to find a sponsored or partner-marked placement. There is none. |
| Ownership is disclosed, not hidden | We own Slow Duo; every own-brand pick is labelled and tested identically, and the footer of every page carries the statement. | Footer of any page, the roundup index, and the ownership policy. |
| Accountability is named | Every review, roundup and face-off carries its author. No generic house byline exists. | The byline on any article, and the About page. |
| Corrections are published, not deleted | A written corrections process with response targets, including retraction when a review cannot be supported. | The corrections policy and the dated fixes on affected pages. |
| Your reading is not the product | Analytics is opt-in; the Google tag is not requested at all before you allow it, and ad signals are permanently denied. | The privacy page and the consent banner — decline and watch that nothing loads. |
The conflict we do not hide
We own the Slow Duo device line, and Slow Duo products are recommended in some of our roundups. Any review site where the house brand keeps winning deserves suspicion, so here is exactly how we handle it:
- We disclose ownership in the footer of every page, on the categories index, in the editorial policy and in our machine-readable organisation data.
- Every Slow Duo pick in a roundup or face-off carries an explicit "ours" marker.
- Slow Duo devices run the identical protocol and the identical published weights as every competitor. No internal criteria, no exceptions.
- Where a competitor wins on a weighted criterion, the competitor is ranked first and the reason is written in the verdict.
- Ask us for the best non-Slow-Duo option for your constraint and we will name it — that is an editorial recommendation, not a commercial one.
If you think that conflict makes our recommendations unusable for you, that is a fair conclusion. Our reviews publish the underlying measurements anyway, so the numbers remain useful even if you discount the verdict.
Everything we publish about ourselves
How we test
The full protocol: equipment, conditions, the five tests, the published scoring weights and the exclusions.
Read the method →Editorial policy
No pay-to-rank, who is accountable for each claim, our AI and sourcing rules, corrections and complaints.
Read the policy →Disclosure
Every commercial relationship in one place: affiliate links, samples, ownership, and what we refuse to do.
Read the disclosures →Privacy
What we collect, every cookie, retention periods, and your rights under GDPR, UK GDPR and CCPA/CPRA.
Read the privacy policy →Who we are
The named testers accountable for the scores, how the desk makes money, and what we are not good at.
Meet the desk →Corrections & contact
How to report an error, what happens next, and the response times we hold ourselves to.
Report an error →How we respond when something is wrong
- Factual error in a review — corrected in place, dated, and acknowledged. Target: acknowledged within two working days, resolved within ten.
- Disputed score — you get the measured data and the weights, and the editor's ruling is published with the correction rather than left private.
- Unsupportable review — retracted, with the reason stated on the page.
- Missing or unclear disclosure — treated as a bug and fixed immediately, because a disclosure that readers cannot find is not a disclosure.
What this site is not
We are not a shop, so we cannot process an order, refund a purchase or handle a warranty — those sit with the retailer you bought from. We are not a clinic, and nothing here is medical advice; if you have a health concern, speak to a qualified clinician. And we are not a generic content farm: there are ten tested devices on this site, not a thousand pages of rewritten spec sheets, and that is deliberate.
Independent corroboration
A trust page that only quotes itself is worth very little. What you can check independently: whether the published weights match the review scores, whether the disclosure appears on the articles where our own products win, and whether a device that lost on the bench is still published with the reason. Every one of those is verifiable from the pages on this site, and we would rather you checked than took our word for it.
If you find a page where any of the above does not hold, tell us. A trust failure we do not know about is the only kind we cannot fix.