Tests, tools & research
The sex toy price index: the method, published before the numbers
This site covers eight English, French, German and Italian versions of the same catalog, which puts a genuine cross-market price comparison within reach. It also means getting it wrong would be easy, so the method comes first.
The observation behind this project is easy to make and hard to quantify. The same toy, from the same manufacturer, carries prices in Britain, Canada, Australia, France, Germany and Italy that do not line up once you convert them. Some of the gap is tax. Some is distribution. Some is what a brand thinks a market will bear. Nobody publishes which is which, and the result is that readers in the smaller markets routinely overpay without any way to know it.
A multi-market site is the only kind of publisher that can even attempt this, because it needs the same product list, the same observation dates and the same currency treatment applied across every market at once. A single-country site can only ever report its own prices.
The five things that make a cross-market price comparison go wrong
- Tax quoted inconsistently. Consumer prices in the European markets include value added tax; North American listings exclude sales tax, which then varies by province and by state. Comparing the two as published compares a tax-inclusive number with a tax-exclusive one, and the gap it produces is not a price difference at all.
- The exchange rate on which day. A figure converted at today’s rate and compared with one collected three weeks ago is measuring the currency market, not the toy market. Any index worth reading states its conversion date and uses the same one for every observation in a round.
- Recommended price versus the price you pay. Manufacturer list prices are tidy, comparable and almost never what anybody pays. Street prices are what matter and they move weekly, which is harder to collect and the only thing worth collecting.
- Parallel imports and gray stock. A cheap listing in one market is often stock imported outside the manufacturer’s distribution, which usually means no local warranty. That is a genuinely lower price for a genuinely different product, and an index that silently includes it reports a saving that does not exist.
- Shipping, duties and the cost of returning something. A price that excludes delivery, or that becomes a customs charge on arrival, is not the price. For a reader the only number that means anything is the total at the door, including what it would cost to send it back.
The method we will follow
A fixed basket, observed on fixed dates, from the maker’s own store in each market wherever one exists, with the total delivered cost recorded rather than the headline price.
The basket. A small set of products that are sold in every market under the same model name, drawn from the catalog already covered on this site. Fixed in advance and changed only between editions, with any change stated. A basket that is reselected each round to make a point is the oldest trick in index journalism.
The source. The manufacturer’s own regional store first, because it is the one listing whose provenance is certain. Where no maker-run store serves a market, a named national retailer, identified in the table rather than averaged into anonymity.
The observation. All markets on the same day, within the same few hours, with the currency conversion taken at that day’s published mid-market rate and the rate recorded alongside the figures.
The number recorded. Delivered total for one unit to a residential address in the country’s largest city, tax included where it is included in the quoted price and stated separately where it is not. Not the list price, not the sale price of the week presented as normal.
The publication. Every observation visible, not just the averages, so that a reader who thinks we picked a bad retailer can see which one we picked and say so.
What has to be in place before the first edition
Fix the basket and publish it
The product list has to be settled and visible before any price is collected, so that nobody can accuse us of choosing the basket after seeing the numbers. Including us.
Build collection that records the whole checkout, not the listing
The delivered total only appears at the end of a checkout, with a destination address entered. Capturing it reliably in six or seven markets is the actual engineering work, and it is the reason this page exists before the index does.
Run one silent round and throw it away
A first round is used to find the places where the method breaks: a market where the maker has no store, a product renamed locally, a checkout that will not quote delivery without an account. Those get fixed before anything is published.
Publish with the raw table attached
Headline findings are fine. They have to sit on top of every individual observation with its source and date, downloadable, so the conclusion can be checked rather than believed.
Questions people ask
- Why not just publish what you have now?
- Because what we have now is the same thing anybody has: an impression that prices differ, formed by looking at a few listings. Publishing that as an index would turn an impression into a statistic, which is exactly the move this site refuses to make.
- Which country is cheapest?
- We do not know, and anybody telling you confidently has not accounted for tax treatment, delivery or the exchange rate date. That is the question the index is being built to answer, and it will answer it with the observations attached.
- Is the difference just VAT?
- Tax explains part of it and cannot explain all of it, because gaps persist between markets with similar tax rates. Distribution structure and what a brand charges a given market appear to account for the rest, which is the interesting part and the part nobody publishes.
- Can I buy from a cheaper country instead?
- Often, and sometimes it genuinely saves money. Weigh it against delivery, possible import duty, and the fact that a product bought outside your region’s distribution may carry no local warranty. The index will record delivered totals precisely so this comparison is possible.
- Will you take a brand’s word for its own prices?
- No. Every figure will come from a live checkout we completed up to the point of payment, not from a price list a manufacturer sent us. A brand’s stated price and the price at your door are frequently different numbers.
- When is the first edition?
- When the collection described above works in every market, and not before. We would rather this page say "not yet" for another year than publish a number we cannot defend.