Sometime last autumn a friend messaged to ask whether a particular laptop was worth buying because it was "40% off." I looked it up. The reference price had only appeared on the retailer's site for six days in September before dropping to the current "sale" price — which was, incidentally, where it had been priced for most of the previous year.

This is the kind of thing that made DealRadar exist. The discount percentage tells you almost nothing useful. What matters is whether the price you're paying right now is actually low relative to what the item normally costs.

So we did something methodical. We pulled 40 "sale" or "aanbieding" listings from five major Dutch electronics retailers over a six-week period and ran each against 90-day price history data. The results were, to put it diplomatically, mixed.

What we looked at

We focused on mid-range electronics — laptops between €500 and €1,200, tablets, wireless headphones, and small home appliances that cross into electronics (robot vacuums, air fryers with smart features). Deliberately broad, because we wanted to see if patterns held across categories or were specific to one.

For each listing, we recorded the displayed "original" price, the "sale" price, and the percentage discount. Then we pulled the 90-day price history for the same product across all available sources and calculated the 90-day median price.

The question we were asking was simple: is the sale price below the 90-day median? If yes, the deal is real. If the sale price is at or above the median, the percentage is fiction.

The actual numbers

Of 40 listings, 22 had a sale price at or above the 90-day median. That's more than half. Of the 18 that were genuinely below median, 11 of them came from the same two retailers. The other three retailers had, between them, five real deals out of 29 listings.

A sale label is not evidence of a low price. It's evidence that someone wants you to think the price is low.

The particularly frustrating pattern we saw was what we're calling the "October reference price" — a product listed at an inflated price for a short window, then dropped to a lower price that is maintained for months, then periodically relabelled as a discount from the original inflated reference. This is legal under current EU pricing regulations as long as the reference price was actually charged at some point in the past 30 days (the Omnibus Directive sets this threshold), but it's misleading in practice.

Which retailers did well

We're not naming retailers here partly because this is a single study over a limited time window and partly because patterns shift. What we can say is that the better performers tended to have more transparent price history on their own product pages — they showed historical pricing voluntarily, which both reflects honesty and allows customers to self-verify.

The retailers that performed poorly were more likely to show only the reference price and the current sale price, with no history visible on the page. Which isn't surprising.

What this means for your buying decisions

A few practical notes from running this analysis:

  • Percentage discounts above 40% should make you more skeptical, not less. Larger discounts from inflated reference prices are a consistent pattern.
  • Price history tools take about two minutes to use. Browser extensions that overlay price history on product pages are useful but check what data sources they use — some EU retailers aren't well covered.
  • The 90-day window matters more than the 30-day window for most categories. Holiday promotions artificially inflate the 30-day median.
  • Laptop prices are particularly noisy around back-to-school periods (August–September). The same laptop can vary by 15–20% just from retailer promotion cycles within a single week.

We'll continue tracking this. The six-week sample is enough to see directional patterns, but a longer study would let us say more about seasonality and whether retailer behaviour changes around major sale events.

If you want to see the methodology behind the Deal Score in more detail, the Radar Method page covers how we weight the four factors.