Rating and ranking methodology

1. Three kinds of rating — not to be confused

Every offer may display a rating out of 5. Behind that “/5” sit three quite distinct things — and we never mix them up:

• Épargna editorial rating. When a member of the editorial team has analysed an offer in depth, they set a rating out of 5: this is Épargna’s own assessment, based on verified public facts and on the trend in public reviews. It is not a user review, and it is never turned into aggregate review “stars”.
• Editorial index (automatic). As long as no editorial rating has been set, we compute an editorial index from objective, public criteria. It is flagged as such (a “*” marker / “editorial index”) so that it can never be mistaken for a user rating.
• Member reviews. The genuine star reviews left by our members on the site: they alone show an average and a review count, and they alone produce structured stars for search engines.

A point of honesty: as long as a rating is only an editorial index, we emit no aggregate-review structured data (AggregateRating) in the page’s code. And an Épargna editorial rating, although set by the editorial team, is no more turned into aggregate review “stars”: only our members’ reviews are. We never turn an editorial judgement into customer reviews in search engines.

2. What the rating covers: service quality and fees

Our index combines two dimensions, and only two: service quality and fees. The result is an index from 0 to 100, converted into a rating out of 5 (index ÷ 20). Neither the welcome bonus nor referral rewards enter into it: they are information, not ranking factors.

In concrete terms, here is what we look at to establish service quality and the level of fees. We always describe a level and a nature (for instance “low” management fees), never a bare figure taken out of context.

3. The welcome bonus: information, not a criterion

A welcome bonus (cash, discount, free months) can be pleasant, but it says nothing about the quality of a service or its cost over time. That is why it does not enter our index or the order of the ranking.

• The bonus is still displayed as information on every offer, so that you have it in mind.
• You can, if you wish, sort by bonus yourself: that is then your explicit choice, not our default ranking.
• Any bonus amount shown is indicative and not contractual; only the partner’s terms at the moment you sign up are binding.

4. How offers are ordered

By default, the offers in a comparison are ordered by descending quality + fees index (the Épargna editorial rating where one exists, otherwise the editorial index). The welcome bonus plays no part in that order. You stay in control:

• sort by your own priority: quality + fees index, or — if you decide so — bonus amount;
• filter by sub-category and search for a specific offer;
• read the detailed strengths and limitations of each offer on its own page.

5. Our revenue plays no part in the calculation

This is a non-negotiable principle: the module that computes the index has no access to affiliate links or to paid referral codes. It works only on public editorial fields covering quality and cost (level and nature of fees, regulation, reliability, product range, experience). The size of a commission therefore cannot, by construction, improve a position or inflate a rating.

Our business model is set out in detail on the affiliate transparency page.

6. How careful we are with figures

The fees, bonuses, rates and terms of financial firms change constantly, and some vary with your profile. So as never to display a figure with false certainty, we apply a simple rule: every figure published is indicative and not contractual.

• We update our data by automated collection followed by human checking before publication.
• We do not publish a figure we are not able to source; when in doubt, we prefer a cautious range or a qualitative wording over a falsely precise amount.
• Only the terms displayed by the partner are binding at the moment you sign up: that is the figure to confirm yourself on the official site before committing.

7. Where the data comes from

Our pages combine two complementary sources:

• Automated collection (scraping). We periodically record the public information on partners’ sites (fees, offers, terms) to detect changes and feed the pages.
• Human curation. The editorial team selects the offers, writes the strengths and limitations, settles ambiguous cases, sets the editorial ratings and confirms the data against its official source before treating it as checked.

We do not claim to cover the entire market: our scope is the set of offers we have selected and studied. The absence of an offer says nothing about its quality.

Put plainly, our comparisons show a selection — the short list of the best recognised platforms on the market — rather than an exhaustive directory. That is a choice of transparency and relevance: we would rather compare a few serious, regulated and proven firms honestly than line up every offer in existence. A platform can therefore be active and perfectly respectable without appearing in our comparisons: its absence says nothing about its quality, only about the scope of our selection.

Check for yourself, without taking our word for it

A methodology is only worth something if it can be checked. These registers are public, free and authoritative in France: an authorisation, a registration or a public warning can be verified there in seconds. We reproduce no figures from them — we send you to the source.

8. Report an error

Despite our care, a piece of data can be out of date or inaccurate. If you spot an error (a fee that has changed, an expired bonus, a condition wrongly transcribed), tell us: we correct it and re-check the offer against its official source.

• Through our contact page.
• Where possible, give the URL of the page concerned and the exact figure you saw at the partner.