BuyBlock — decide with data, not dopamine

Is that Black Friday deal real?

Every discount is a claim about two numbers, and only one of them is real. Here is how to check the other one — and how to tell, in about a minute, whether the offer in front of you is worth taking.

BuyBlock guide · Updated August 2026

The reference-price trick

A price on its own is almost impossible to judge. Is 149 a lot for a pair of headphones? Not unless you already know what headphones cost. So the retailer helps: was 299, now 149. You are no longer judging 149. You are judging the gap — and the gap looks enormous.

That is anchoring, one of the sturdiest findings in behavioural science: people judge a number against whatever number they saw first, even when told the first one is arbitrary. In a shop it is not arbitrary. It is chosen. The "was" price is the part of the offer that does the persuading, and the part nobody checks.

So the cheapest way to manufacture a discount is not to cut the price you pay. It is to raise the price you compare against: nudge the shelf price up through late October, hold it there until it looks normal, then "cut" it on the Friday. Effect on your bank balance, nothing. Effect on your perception, half price.

We know this is a real practice rather than shopper paranoia, because regulators wrote rules to stop it. Since 2022, EU law has required that a trader announcing a price reduction advertise it against the lowest price that trader charged in the previous 30 days — not yesterday's price, and not the manufacturer's suggested figure. The EU's top court reinforced the point in 2024: the percentage has to be calculated from that 30-day low, so a shop cannot raise its price for a week and then discount from the inflated number.

The UK has no fixed 30-day formula, but the principle is identical. Guidance approved under UK consumer protection law tells traders a "was" price should be one the product was genuinely available at, for a meaningful period, shortly before the promotion. The Competition and Markets Authority — which can now enforce consumer law directly rather than through the courts — treats a reference price the goods were never really sold at as misleading.

Rules like these do not get drafted about imaginary problems, and enforcement inevitably runs behind the volume of offers published in a single November week. So the shopper's conclusion is unchanged: treat the "was" number as marketing copy until you have verified it yourself. Consumer groups run the same exercise every year — track a basket of Black Friday products across the months either side — and a substantial share turn out to have been available at the same price or cheaper at some other point. The proportion moves. The finding does not.

How to check a real price history

Only one number tells you whether a discount is real, and it is not the RRP, the list price or the retailer's own "was". It is the lowest price this exact product has actually sold for recently. Ninety days is the right window: long enough to contain a previous promotional cycle, short enough that you are still looking at the current model.

Three checks, in order:

1. The history at the retailer you are on. For Amazon listings, third-party price trackers plot the last year or more, and the shape of the graph settles it in five seconds: a genuine cut looks like a new low, a manufactured one looks like a step up followed by a step back down. On EU storefronts the prior-price line that must accompany a promotion is doing this work for you — read it before the big red percentage.
2. The price at three or four competitors, right now. One retailer's history tells you about one retailer's habits. Four retailers' current prices tell you what the product is worth this week. If one sits far below the rest, that is information rather than luck: check for a marketplace seller with no manufacturer warranty, a grey import, a different regional model, or refurbished stock presented as new.
3. Where the product sits in its lifecycle. A three-year-old model with a successor already announced is not on sale, it is being cleared, and it will keep getting cheaper. A model launched six weeks ago is very unlikely to be genuinely discounted, whatever the badge says.

A real discount is below the trailing 90-day low, on the current model, in new condition, and roughly matched by at least one competitor. A fake one is a large saving measured against a number nobody has paid in a year, from exactly one seller, on a model you cannot find anywhere else.

The five signs of a fake deal

1. The anchor is a list price, not a price

Watch the wording. "RRP", "list price", "UVP", "compare at", "value" — all ways of quoting a number without claiming anyone ever paid it. Suggested prices are frequently fiction in mature categories, set high so that permanent discounting has something to discount from. If the page will not say our price was, assume it never was.

2. The countdown is doing the arguing

Doorbusters, "only 12 left", timers that quietly reset when you reload in a private window. Scarcity is a fact when it is real and a pressure device when it is not, and the test is free: open the page in a fresh window and see whether the clock and the stock counter tell the same story. A good price does not need a stopwatch to close the sale.

3. The bundle hides the item

Headphones plus a case plus three months of a subscription you would never buy alone. Bundles defeat comparison by construction — the package exists only here, so there is nothing to compare it against. Price the one thing you actually want, on its own, at three retailers. If the bundle only wins once you value the extras at their list prices, see sign one.

4. Condition is left vague

"Open box", "renewed", "certified refurbished", "outlet", "graded". Some are excellent value with a real warranty behind them; the problem is condition buried in a spec table while the discount is anchored to the new-item price. Two questions settle it: who is warranting this, and for how long? If the answer is a marketplace seller and 30 days, the comparison price is not the new price.

5. The model number exists nowhere else

Retailer-exclusive derivative models are standard practice in televisions and large appliances: a variant one digit off the mainstream model, with a different panel or a thinner spec, sold through one chain only. The exclusivity is the mechanism, not a perk — it makes the product un-comparable, and un-comparable products cannot be shown to be expensive. Search the exact model number in quotes. If only one retailer returns it, you have lost the ability to price-check, and that was the point.

The 60-second check before you buy

Put together, the check is short. Open a second tab. Search the exact model number. Look at three or four sellers. Find the trailing low. Confirm the condition and the warranty. That is under a minute of work, and it kills most bad deals outright.

Then comes the part that actually decides the outcome: do not buy it in the same session.

The architecture of a sale event exists to compress the decision into the moment of maximum wanting — the countdown, the stock counter, the one-tap checkout, the saved card. None of it survives a delay. If the price is genuinely good it is still good tomorrow, and the offers that vanish overnight were mostly deadlines wearing the costume of a deal. The highest-yield move in November is not a better price tracker. It is a gap between wanting and paying.

To see the format first, we publish verdicts on popular products at Should you buy it? — real prices across retailers, review highlights, and a buy, wait or skip call on each. A few live examples: the PlayStation 5 Pro (reseller markup over MSRP), AirPods Pro 3 (below-retail listings worth checking), the Dyson Airwrap, the Steam Deck OLED (buy direct, skip the third-party markup) and the Stanley Quencher.

Frequently asked questions

Are Black Friday deals actually cheaper?

Some are, and the skill is telling which. Discounts tend to be real where a reference price is visible and comparable — hardware with a model number every retailer stocks, especially when a successor is on the way. They tend to be theatre where nobody knows what the item should cost: furniture, mattresses, jewellery and some premium small appliances have run near-permanent "sale" pricing for decades, precisely because the customer has no benchmark. Verify per product, not per event.

How do I check a price history?

On Amazon, a price-tracking site plots the listing's history back over a year, which is the fastest answer available. In the EU, a promotion must be shown against the seller's lowest price of the previous 30 days, so the disclosure is already on the page. Everywhere else, use breadth instead of depth: the current price at four competing retailers approximates the market price closely enough to catch an inflated anchor. And if you are going to wait a fortnight, screenshot the price with the date — that turns "I think it was cheaper in October" into evidence.

When do Black Friday prices really drop?

The reliable trigger is not a date, it is a lifecycle event. The deepest and most durable cuts land when the successor is announced or starts shipping, because that is when the retailer genuinely needs the old stock gone — which is why last year's laptops fall after a chip refresh, and why television prices are usually at their most honest as the new season's lineups arrive rather than in November. Black Friday is the moment retailers most want you to buy, which is not the moment the price is lowest.

Is quoting an inflated "was" price illegal?

Advertising a reduction is legal; misleading people about what the product previously cost is what the rules address. EU law fixes the comparison to the trader's lowest price of the preceding 30 days, and UK guidance requires a "was" price to be genuine and recent. Enforcement lags the volume, so verify the anchor before you value the discount.

---

Next: if the honest answer is that you overspend in November because you overspend generally, read No-Buy 2027: a practical guide.

Research your own impulse buys with BuyBlock.