Methodology

Trust in this space is earned by showing your work. Here is exactly what the Birkin Score is, where every number comes from, and — just as important — what it is not.

What the number actually is

The Birkin Score is a position indicator: it shows where your documented pre-spend ratio (your cumulative boutique spend ÷ the bag's retail price) sits within the community-reported distribution of ratios at which buyers reported being offered that bag. It is “your position vs community-reported outcomes,” not your chance, probability, or odds of an offer. No public source proves a probability, so we do not claim one.

Where every number comes from (the receipts)

• The reported ratio distribution & the ~1.4× baseline — from Orange Book, a public crowd-sourced archive of 931 Hermès purchase reports (retrieved 13 Jul 2026). We store a dated snapshot.
• Per-city medians — Orange Book's own per-city figures, but we publish a city's number only when it has at least 8 reports (e.g. Tokyo 1.5× / 40 reports, Paris 1.0× / 20, London 1.2× / 10, New York 1.2× / 9). Thinner cities (and any we have no data for, like Dubai) fall back to the community baseline and the result says so plainly. We never invent a city number.
• The concept & the 1:1 rule-of-thumb — our own transcripts of ~59 buyer videos, plus specialist write-ups (PurseBlog, Madison Avenue Couture). These confirm the range (1:1 typical, 2:1+ competitive) but contain no per-city calibration.
• Bag retail prices — tracked public retail, the denominator of the ratio.
• Your anonymous check — city, bag, banded spend (no identity, no exact figure). This is what will let each city's number get truer over time.

What is deliberately NOT in the number

• Category mix, months, visits, sales-associate relationship, and region. The community believes these help — but no public data proves any of them changes the outcome, and Hermès publishes no such rule. So we show them as labelled context, and they do not move your score.
• We removed the old v1 weights (50/25/15/10) and its per-city medians after checking them against real data — several were materially off (e.g. New York was 1.6× in v1 vs 1.2× reported). Publishing invented precision is exactly what we won't do.

Honest limitations

• Success-reporting bias. People post after they get the bag, so reported ratios skew optimistic. Orange Book says this itself; we repeat it on every result. The fix is collecting non-offer outcomes too — which is why we're building outcome reporting.
• Small samples. Only a handful of cities have enough reports to publish; we show the count and fall back honestly everywhere else.
• Boutiques are people. No model captures one associate's judgment. This measures patterns, never promises.

Before we would ever say “chance” or “probability”

We would first need at least 100 outcome-complete reports (including 30+ “no offer” cases), an interpretable model with a stated uncertainty range, and statistical plus legal review. Until then, this stays a descriptive position, not a prediction.

Prices & the Orange Peel Index

Every price on the site is an estimate with a named source, stored per spec in our data (retrieved 13 Jul 2026):

• Retail — documented US boutique list prices after the January 2026 increase (Sotheby's, PurseBop, PurseBlog, BagUSeek). This is also the denominator of the Birkin Score.
• Resale — pristine secondary-market asking/sold prices for a standard colourway, from 2026 marketplace listings (Fashionphile, 1stDibs, Sotheby's, The Luxury Closet, Ginza Xiaoma).
• The Orange Peel Index — the secondhand premium over retail as tracked by Bernstein's Secondhand Pricing Tracker (average across all Birkin/Kelly auction sales), reported by Fortune and CNBC in December 2025: it fell from 2.2× at the 2022 peak to ~1.4× in late 2025. Those two years are Bernstein's readings; the years between are interpolated, and we say so.

One honest wrinkle we show rather than hide: the index (~1.4×) is an all-size auction average, while the per-spec premiums you see (a Birkin 25 at ~2.1×, a Mini Kelly near 2.9×) are pristine small sizes on consignment — so they sit above the index. Different measurements, both real. None of these are appraisals or financial advice, and a true many-listing median only arrives with Supply View.

The Orange Peel premium-index methodology

The Orange Peel Indexis our single headline reading of the Hermès game's secondhand premium — how much a Birkin or Kelly resells for as a multiple of its retail price. Because a premium is the one number the whole market reacts to, we hold it to the same receipts-first standard as everything else. Here is exactly what it measures and, just as important, what it does not.

• What it measures. A market-wide ratio: resale price ÷ retail price. A reading of 1.4× means the average bag resold for about 140% of its list price. It is a trend indicator for the category, not a valuation of any one bag.
• The basis (Bernstein). The headline trend uses Bernstein's Secondhand Pricing Tracker — an average across Birkin and Kelly auction sales of all sizes — as reported by Fortune (17 Dec 2025) and CNBC (16 Dec 2025). It fell from ~2.2× at the 2022 peak to ~1.4× in late 2025.
• Readings vs. interpolations (disclosed, never hidden). 2022 and 2025 are Bernstein's actual reported readings. The years in between are drawn as a straight line between those two anchors — an interpolation, not separate measurements — and we mark the interpolated years as hollow points and say so in text. We will not present a line segment as a reading it isn't.
• Auction vs. marketplace (two different lenses). The Bernstein index is an auction average across all sizes, so it runs lower than the per-spec premiums we show elsewhere, which are marketplace asking/sold figures for pristine, in-demand small sizes on consignment. A Birkin 25 near 2.1× or a Mini Kelly near 2.9× is not a contradiction of a 1.4× index — it is a different, narrower measurement, and we label which is which rather than blend them into one false number.
• Why the distinction is the trust asset. Anyone can post a premium multiple; the credibility is in disclosing the sample, the method, and the gaps. Naming the auction-vs-marketplace difference and flagging every interpolated point is how a price index earns citation instead of skepticism.

A true many-listing marketplace median — a live figure built from the full set of current listings rather than an anchor-and-interpolate line — arrives with Supply View. Until then the index stays a cited trend with its readings and interpolations shown, never an appraisal and never financial advice.

Cite this data

Our methodology, the Orange Peel Index, and every price and score figure are our own compiled, dated work. You're welcome to cite them — with a link back to the page you drew the figure from. Please don't republish our datasets wholesale or scrape them without permission, and keep each number's label (“estimated,” “community-reported,” “cited median”) when you quote it. Cite as: “Orange Peel — Methodology,” orangepeel.to/methodology.

Our commitments

• We will never present the score as a guarantee or a probability.
• We will never invent a data point — every number traces to a cited source or an anonymous submission, with its sample size shown.
• We publish updates here whenever the model changes.