How to Tell a Real Tech Deal From a Fake One

Update, 8 October 2026. The portable speaker shortlist was re-read on 8 October and gained a flagship-tier pick, the JBL Xtreme 5, which averages the same 9.7 as the budget JBL Flip 7. The count in check four was six of eight shortlists in which the budget pick matched or beat the flagship; with portable speakers added it is seven of nine. The two gaps that favour the flagship are unchanged.

Update, 21 September 2026. The robot vacuum and TV shortlists were re-read on 21 September. The count in check four still holds at six of eight, but the robot vacuum gap it cites fell from 1.3 points to 0.8, because the Roborock Saros 20’s average moved from 9.0 to 8.5. We had called that the one real gap; at 0.8 it is the largest in the comparison but less than half a star on a single review, and the paragraph now says so. A link below that still gave an older count, four of six, now matches the page it points to.

We do not publish deal lists, and we are not going to. A list of live prices is wrong within hours, it cannot be checked, and it is where a site’s credibility goes to die. What we publish instead is the number a deal has to beat — a list price, with the date we read it and the source we read it from, on every product in every shortlist here.

That is deliberately less exciting than a countdown timer. It is also the only part of a sale that is checkable.

Check one: what was the price before the sale?

A discount is a percentage of a number the seller chose, and sellers move that number.

The cleanest example available this year is Microsoft’s. Xbox console prices rose by $100 on 512GB models and $150 on 1TB models on 1 August — before the season, by Microsoft’s own announcement. Any November percentage is calculated against the raised figure, so the same headline discount returns less money than it did in July, and a saving measured against the price you remember from last year is not a saving at all.

This is why a dated list price is worth more than a percentage. Ours are on the product cards, and they say when we looked.

Check two: is it the current generation, or the one being cleared?

Sales are how last year’s stock leaves the building, and that is not automatically a bad thing — but you should know which one you are buying.

Two live examples from our own pages. Samsung’s Galaxy Watch 9 (affiliate link to Amazon, opens in a new tab) arrived in August and scored lower than the Watch 8 (affiliate link to Amazon, opens in a new tab) at both outlets that rated both, for $30 more, on the same design and sensors — so the previous model is both cheaper and better reviewed. Samsung’s soundbar line splits the other way: the roundups disagree about whether the HW-Q990F (affiliate link to Amazon, opens in a new tab) or the newer HW-Q990H (affiliate link to Amazon, opens in a new tab) is the pick, which is exactly the confusion a sale exploits.

A deep discount on a product three weeks before its replacement ships is a clearance, not a bargain — unless the replacement is worse, which happens more often than the marketing suggests.

Check three: does the thing you must buy afterwards swamp the discount?

The purchase price is frequently the smaller number.

• A console at 20% off still needs online play, which runs from $19.99 a year on one platform to about $120 on another, and storage that is proprietary on two of the three.
• A tablet at $399 needs a keyboard and stylus that are sold separately and can add hundreds.
• A streaming subscription is where the ad-free tier costs $132 a year on one service and $90 on another for the identical removal.

Work out the two-year cost before the discount impresses you. On subscriptions and consumables it will frequently exceed the saving.

Check four: is the cheap one already as good?

This is the check that saves the most money, and it is the one nobody markets to you.

Across our shortlists, in seven of the nine categories carrying both a budget and a flagship pick, the budget pick matches or beats the flagship on averaged review score — laptops, mini PCs and tablets outright, soundbars, wireless earbuds, smartwatches and portable speakers level. In the two where the expensive product wins, robot vacuums and ereaders, it wins by 0.8 points and 0.3 respectively. Neither is as large as the single point that half a star is worth on one review, though the vacuum gap is the one reviewers explain: all three outlets that scored the budget machine mark down its mopping.

The implication for a sale is direct: a 20% discount on the flagship is often a worse purchase than the full price of the budget model, because you are paying a premium to reach a score the cheaper thing already had. The full working is in the cheap one wins three times out of four.

What the season actually looks like

Adobe’s data puts the ceiling on electronics discounting at 30.9% off list at its 2025 peak. That is the good end, not the typical one, and it arrives on specific days rather than across the month.

The more useful finding in the same data is behavioural: discounting moves people upmarket rather than saving them money. Shoppers do not spend less; they spend the same on a better model. That is worth knowing about yourself before you open a sale page, because it is the mechanism the whole season is built on rather than an accident of it. Our piece on why headphones are the biggest spike covers where that pressure lands hardest.

How to use this site during a sale

1. Open the shortlist for the category and note the list price and the date we checked it.
2. Subtract the advertised price from that, not from the crossed-out number on the retailer’s page.
3. Check the generation — our picks name the model, and say when a newer one exists and scores worse.
4. Add the two-year running cost where there is one: subscriptions, storage, consumables.
5. Compare the budget pick’s score to the flagship’s before paying the premium.

If a deal survives all five, it is probably real. We will not email you about it, and there is no timer on this page.