US electronics discounting is likely to get shallower this holiday season, not deeper. The specific call: when Adobe publishes its 2026 holiday wrap in early January 2027, the average peak discount on electronics is likely to print below the 30.9% recorded in 2025, breaking two consecutive years of deepening promotion, even as electronics dollar sales keep growing. That is a narrow, checkable claim with a fixed publication date attached to it.
The reasoning does not rest on the memory shortage itself, which has been visible in component markets for a year. It rests on three pricing decisions taken in roughly the last five weeks, which together suggest the shock has stopped behaving like a uniform cost pass-through and started behaving like a strategic choice. When the same input cost produces opposite pricing decisions at competing manufacturers, the promotional calendar that follows tends to fragment rather than deepen.
In short
- The prediction: average peak electronics discount depth in the 2026 US online holiday season is likely to come in below 2025’s 30.9%, the first compression in three years, while electronics online dollar sales still rise year on year.
- Signal 1: three console makers repriced within five weeks and moved in three different directions, with Microsoft raising Xbox by $100–150 effective August 1, Sony cutting the PS5 back by $100–150 effective August 21, and Nintendo lifting the Switch 2 to $499.99 effective September 1.
- Signal 2: Best Buy’s second quarter, reported August 27, showed computing average selling prices up in the mid-teens while computing units fell in the high single digits, with management naming memory as a computing-specific cost issue.
- Signal 3: retail memory pricing re-accelerated in August rather than cooling, with the widely tracked 3DCenter DDR5 retail index moving to 486% of its July 2025 baseline from 445% a month earlier, a 9.2% monthly step inside the window when Q4 inventory is committed.
- The timeframe and the main risk: checkable in Adobe’s holiday release in early January 2027 and again in Best Buy’s Q4 print in late February or early March 2027; the clearest way to be wrong is that television and appliance promotion, which is not memory-bound, holds the blended average near 31%.
Why this matters now
Discount depth is the single most useful summary statistic for a holiday season, because it captures what retailers actually decided rather than what they said. Dollar totals conflate price and volume, and traffic figures conflate intent and conversion. Peak discount depth, measured consistently across years, tells you how much margin the category was willing to surrender to move goods.
Adobe’s 2025 numbers set a clean baseline. Consumers spent a record $257.8bn online between November 1 and December 31, up 6.8% year on year, with electronics contributing $59.8bn on 8.2% growth. Discounts on electronics peaked at an average of 30.9%, up from 30.1% the prior year. That is two consecutive years of the category buying its growth with deeper cuts.
The question for 2026 is whether that trend continues into a year when the dominant input cost has roughly quintupled at retail. This publication previously argued that the memory shock would reach consumers mainly as specification compression rather than headline price rises, with less RAM and less storage behind the same badge. The evidence since then suggests that call was directionally right for computing and incomplete for hardware where configuration cannot flex.
Consoles are the clearest example of that limit. A console is sold as a fixed specification against a published price, so a manufacturer facing a memory shock has only two levers: move the price, or absorb the cost. In the last five weeks all three major console makers pulled one of those levers, and they did not pull the same one.
Signal 1: three console makers repriced in five weeks, in three directions
The console segment repriced three times between August 1 and September 1, 2026. Microsoft moved first, with Xbox Series X and Series S increases taking effect on August 1: roughly $100 on 512GB models and $150 on 1TB models, with the 2TB configuration discontinued. The company attributed the move directly to components, stating that console storage and memory prices have increased by more than 2.5x and that it expects a further doubling by the autumn of 2027.
Sony went the other way. Having raised all three PS5 configurations on April 2, taking the standard console to $649.99, the Digital Edition to $599.99 and the Pro to $899.99, it reversed that increase effective August 21, returning the line to $549.99, $499.99 and $749.99. Reporting on the reversal ties it to a sharp fall in PS5 hardware sales through the spring, following the earlier repricing. Sony faces the same memory market as Microsoft and chose share over margin heading into the holiday.
Nintendo then moved up. The Switch 2 went from $449.99 to $499.99 in the US effective September 1, an increase of about 11%, with company commentary in May pointing to memory component costs alongside exchange rates and oil prices. Three manufacturers, one input shock, three different answers, all inside five weeks.
| Maker | Action | Effective | US price move | Stated driver |
|---|---|---|---|---|
| Microsoft (Xbox Series X and S) | Increase | August 1, 2026 | +$100 on 512GB, +$150 on 1TB; 1TB Series X to $799.99; 2TB discontinued | Storage and memory costs up more than 2.5x, further doubling expected by autumn 2027 |
| Sony (PS5 line) | Reversal of an April increase | August 21, 2026 | Back to $549.99 standard, $499.99 Digital, $749.99 Pro | Not framed as a cost move; follows a sharp spring decline in hardware sales |
| Nintendo (Switch 2) | Increase | September 1, 2026 | $449.99 to $499.99, about 11% | Memory component costs, exchange rates, oil prices |
The analytically useful detail is the timing rather than the direction. All three moves landed before the holiday promotional calendar rather than after it, which is unusual. Manufacturers normally hold list prices steady into Q4 and let retailers do the discounting, because a list move during the season reads as panic.
Repricing in August and early September does something specific: it resets the reference price from which Black Friday discounts are calculated. A 25% cut from a base that just rose 11% is not the same offer as a 25% cut from last year’s base, and Adobe’s methodology measures depth against the prevailing list price, not against the prior year.
Signal 2: Best Buy has already shown the price and volume split
Best Buy’s second quarter, covering the period to August 1 and reported on August 27, is the closest thing the market has to a clean read on how the memory shock is landing at the till. Headline results were strong: enterprise revenue of $9,779m, comparable sales up 4.1% enterprise-wide and 4.5% domestically, domestic online revenue of $3.0bn on 5.1% comparable growth, and a domestic gross profit rate of 24.0% against 23.4% a year earlier. The company raised full-year guidance to comparable sales growth of 1.9–3.0% and adjusted EPS of $6.70–6.90, per the company’s own results release.
The interesting number is not in the release. On the earnings call, management disclosed that computing average selling prices rose in the mid-teens while computing unit sales declined in the high single digits. That is the memory shock, quantified at the point of sale, in a single sentence.
Management also described the mitigation explicitly, pointing to trade-ins, financing and promotional tactics as the tools being used to help customers manage elevated price points. Note the ordering. Trade-in and financing are named before promotion, which is a meaningful signal from a retailer describing how it intends to defend a category whose prices have moved beyond what units can absorb.
Two further details from the call sharpen the read. Management said it did not see material evidence of broad customer demand pull-forward following a major product price announcement, which weakens the argument that consumers are simply buying early and will disappear in Q4. And it flagged that the memory-cost issue has primarily affected computing rather than other categories, which is precisely the fragmentation this analysis expects to show up in the blended discount figure.
| Metric (Best Buy Q2 FY27, reported August 27, 2026) | Value | What it implies |
|---|---|---|
| Enterprise revenue | $9,779m, comparable sales +4.1% | Category demand is not broken |
| Computing average selling prices | Up mid-teens year on year | Cost is reaching the shelf, not being absorbed |
| Computing units | Down high single digits | Volume elasticity is already binding |
| Domestic gross profit rate | 24.0% vs 23.4% | Margin expanded, aided by $34m of tariff refunds |
| Q3 comparable sales guidance | 1.0–3.0% | Deceleration assumed into the holiday run-up |
| Stated mitigation | Trade-in, financing, promotion | Promotion is the third lever, not the first |
One caution on reading this quarter. The gross margin expansion was helped by $34m in tariff refunds, a one-off that flatters the comparison and gives the company headroom it may not have next year. Retailers sitting on unexpected refund income can afford to promote more aggressively than their cost curve implies, and that cuts against the prediction rather than for it.
Signal 3: component costs re-accelerated in August, inside the holiday buy window
The third signal is the one most likely to be missed, because the prevailing narrative through mid-year was that the memory surge had begun to cool. TrendForce’s July 3 guidance put third-quarter conventional DRAM contract prices up 13–18% and NAND up 10–15%, a sharp deceleration from roughly 60% in the prior quarter, and attributed the slowdown to buyers reaching an affordability limit rather than to supply returning.
August did not follow that script at retail. The 3DCenter DDR5 retail price index, published at the end of August, moved to 486% of its July 2025 baseline from 445% a month earlier, a 9.2% month-on-month increase. A 2x32GB DDR5-6400 kit that sold for roughly EUR 181 in July 2025 was around EUR 1,093 in August 2026, an increase of about 503%.
US channel tracking tells the same story in dollars. A mainstream 32GB DDR5-6000 kit was running around $392 in August, against roughly $110–140 as recently as the third quarter of 2025. These are retail component prices rather than contract prices, so they overstate what a large OEM pays, but they are the prices facing the system builders and small assemblers who supply a meaningful slice of holiday PC inventory.
Timing is what makes this a signal rather than a data point. August and September are when Q4 holiday inventory is committed and promotional plans are locked with vendors. A cost curve that re-accelerates during the buying window, rather than during the selling window, constrains what a merchant can commit to before the season opens.
| Signal | Date | Type of evidence | Independent of the others? | What it implies for discount depth |
|---|---|---|---|---|
| Console repricing in three directions | August 1 to September 1, 2026 | Manufacturer list price actions | Yes: OEM pricing decisions | Reference prices reset upward for two of three makers before the season |
| Best Buy price and volume split | August 27, 2026 | Retail point-of-sale disclosure | Yes: retailer demand data | Elasticity already binding; promotion ranked third among levers |
| August memory index re-acceleration | End of August 2026 | Component market price tracking | Yes: upstream input market | Cost pressure rising during the inventory commitment window |
What the pattern suggests
Read together, the three signals point to a holiday season in which electronics promotion fragments rather than deepens uniformly. The category splits into memory-bound goods, where the cost curve is genuinely constraining, and everything else, where normal competitive dynamics apply. A blended average discount that has risen for two years is likely to stall or fall when a large slice of the basket stops participating in the discount cycle.
The console divergence is the strongest evidence for fragmentation, because it shows the shock is no longer deterministic. If memory cost mechanically dictated pricing, all three makers would have moved in the same direction. Sony’s reversal demonstrates that a manufacturer can still choose to eat the cost when the volume response is bad enough, which is exactly what a fragmented promotional landscape looks like from the supply side.
Best Buy’s figures then show what happens on the demand side when the price does move. Mid-teens ASP growth against high single digit unit declines is roughly consistent with an elasticity around negative one, meaning the category is trading dollars for units close to one for one. A merchant looking at that ratio has limited incentive to discount deeply: the extra units bought with a price cut do not obviously pay for the margin given up.
That is the core mechanism behind the prediction. Deep discounting is rational when a modest price cut buys a large volume response, and much less rational when the response is roughly proportional and the underlying cost is still rising. The August component re-acceleration removes the escape route of discounting from cheap inventory bought earlier in the year.
Wider context: promotion is moving off the price tag and onto the payment
If headline percentage-off compresses, the promotional pressure does not disappear. It relocates. Best Buy’s own framing puts trade-in and financing ahead of promotion, and that ordering is likely to characterise the category more broadly this season.
The infrastructure for that shift is already in place. Adobe recorded more than $1bn of buy now, pay later volume on Cyber Monday alone in 2025, on a record $14.25bn single-day total, with mobile accounting for 56.4% of transactions across the season. A category facing mid-teens ASP inflation is a natural home for instalment products, because instalments reduce the salience of the sticker without reducing the price.
Trade-in works the same way, and better for the retailer. A trade-in credit is a targeted discount granted only to customers who bring inventory the retailer can resell, so it lowers the effective price for the buyer without lowering the advertised price for everyone. Expect trade-in offers on phones, tablets and consoles to carry more of the promotional weight than in prior seasons.
Adjacent levers are moving in the same direction. Free-shipping thresholds have been drifting upward across the sector, and this publication has argued that those thresholds are likely to rise again before Black Friday, which is another way of tightening the effective offer without touching the item price. Retail media is doing similar work on the margin line, and the case that retail media funds holiday margin rather than the shelf price fits neatly alongside a season where the shelf price is the one thing merchants are reluctant to move.
There is a regulatory dimension worth watching too. Targeted, individualised offers are the natural substitute for broad discounts, and several US states have new rules taking effect that constrain exactly that practice. The surveillance pricing bans going live on October 1 land immediately before the season and may limit how finely retailers can personalise the substitute offers.
Implications for retailers, brands and marketplaces
For electronics retailers, the planning implication is that gross margin rate is likely to be defensible this season in a way it has not been for several years, while comparable unit volumes are not. Merchants who set incentive plans against unit throughput rather than margin dollars are likely to find the targets mismatched to the environment. The categories to watch are computing and gaming hardware, where the memory content is highest.
For device makers, Sony’s reversal is the most instructive datapoint in the set. It establishes that a large manufacturer facing a genuine cost shock will still retreat on price when the volume response is severe enough, which caps how far competitors can push. Anyone modelling further list increases into 2027 should treat that ceiling as real rather than theoretical.
For marketplaces and third-party sellers, shallower first-party discounting is usually good news, because it widens the price umbrella under which independent sellers operate. The offsetting risk is inventory cost: sellers who bought memory-heavy stock at August prices carry a higher basis than a large retailer buying on annual contracts, so the umbrella may not be as generous as it looks.
For investors, the checkable consequence is a divergence between dollar growth and unit growth across the electronics complex in Q4 reporting. Companies reporting solid revenue growth on declining units are behaving as this analysis expects. Companies reporting unit growth with compressed margin have chosen the other path and are worth separating out.
For consumers, the practical reading is that waiting for a deep cut on memory-heavy hardware is a weaker strategy this year than usual, while televisions, appliances and accessories are likely to promote normally. That asymmetry is unusual and is not yet widely understood.
Caveats: what could go wrong
The most likely way this prediction fails is compositional. Adobe’s electronics bucket is broad, spanning televisions, computing, phones, audio and accessories, and only part of it is memory-constrained. Television panel pricing is driven by a different supply chain entirely, and an aggressive TV promotional war would pull the blended average back toward 31% regardless of what happens in computing. Best Buy’s own quarter showed home theatre delivering its strongest growth since fiscal 2022, with domestic TV sales up more than 10%, which is what a healthy promotional TV market looks like.
The second risk is that retailers have more promotional firepower than the cost curve suggests. Tariff refunds have been substantial across the sector, with Best Buy booking $34m in the quarter and Walmart citing a $2.9bn refund which it deployed into more than 11,000 rollbacks in Q2, up from 7,200 in Q1. That is a large pool of unexpected margin dollars available to buy traffic, and Walmart’s behaviour so far shows a clear willingness to spend it on price.
The third risk is that the memory surge genuinely cools before the season. TrendForce’s own framing has the increases decelerating, and if fourth-quarter contract prices flatten while retailers sell through inventory bought at earlier prices, the constraint loosens at exactly the wrong moment for this forecast. The August retail index re-acceleration argues against that, but retail spot pricing and OEM contract pricing can diverge for months.
The fourth risk runs the opposite way from the thesis and deserves attention. Falling units can provoke deeper discounting rather than shallower, if merchants decide that defending unit share matters more than margin rate. Sony’s August reversal is precisely that behaviour, and if it proves contagious the season could feature deeper cuts on hardware than any cost analysis would predict.
Finally, there are exogenous factors that could swamp the signal entirely. A major software or content catalyst can reset gaming demand independently of price, and Best Buy management specifically flagged the fourth-quarter release of Grand Theft Auto VI as a support for the gaming category. Input costs elsewhere are moving too, and the argument that fuel and energy are likely to outrank tariffs in November retail guidance is a reminder that memory is one line in a longer cost stack.
| Scenario | What happens | Adobe electronics peak discount, holiday 2026 | Leading indicator to watch |
|---|---|---|---|
| Base case | Memory-bound categories promote lightly, others normally | Below 30.9%, likely in the high 28% to 30% range | Black Friday computing offers advertised as trade-in or financing rather than percentage-off |
| Composition offset | TV and appliance promotion carries the blend | Roughly flat at 30.5% to 31.5% | Early November TV doorbusters at or below 2025 price points |
| Share defence | Merchants chase units, following Sony’s reversal | Above 31% | A second manufacturer reversing a 2026 list increase before November |
How to score this prediction
A forecast is only useful if it can be marked. There are three checkpoints, each with a fixed date and a public source, and the first two matter most.
- Early January 2027: Adobe’s holiday season wrap, covering November 1 to December 31, 2026. The prediction is correct if average peak electronics discount depth prints below 30.9% while electronics dollar sales grow year on year. It is wrong if depth exceeds 30.9%, and partially wrong if depth falls but dollar sales also fall, since that would indicate demand failure rather than promotional discipline.
- Late February or early March 2027: Best Buy’s fourth-quarter results. The supporting claim is that computing average selling prices rise year on year again while computing units decline year on year again. Two consecutive quarters of that split would confirm the elasticity read.
- Through the season: the qualitative tell is the structure of advertised offers. If flagship computing and console promotions lead with trade-in credit, instalment terms or bundled content rather than a percentage off the list price, the substitution thesis is holding regardless of where the blended average lands.
Worth stating plainly: the second checkpoint could confirm while the first fails, and that combination would still be informative. It would mean the mechanism is real in computing but too small to move a blended category average, which is a narrower and more accurate version of this argument.
Frequently asked questions
What exactly is being predicted?
That the average peak discount on electronics during the 2026 US online holiday season, as measured and published by Adobe in early January 2027, is likely to come in below the 30.9% recorded for 2025, while electronics online dollar sales still grow year on year. The supporting claim is that Best Buy’s Q4 print will likely again show computing average selling prices up with computing units down.
Is this just the memory shortage story again?
No, and the distinction matters. The memory shortage has been well covered for a year and its first-order effect, higher component costs, is priced into most commentary. The claim here is second-order and concerns the promotional response: how retailers and manufacturers choose to distribute a cost shock across list price, discount depth, specification and payment terms.
Why do the console price moves matter more than the component data?
Because they show the shock has become a choice rather than a mechanism. Component data tells you costs rose, which is not in dispute. Three manufacturers responding to the same cost with three different pricing decisions inside five weeks tells you the pass-through is discretionary, and discretionary pass-through is what fragments a promotional calendar.
Could shallower discounts simply mean weaker demand rather than discipline?
That is a fair challenge and it is why the prediction has two parts. Shallower discounts alongside growing dollar sales indicate merchants holding price into decent demand. Shallower discounts alongside falling dollar sales would indicate something closer to a demand failure, which would not vindicate this argument.
What is the strongest argument against the prediction?
Composition. Adobe’s electronics category blends memory-bound goods with televisions and appliances that face entirely different supply chains. Best Buy reported domestic TV sales up more than 10% with the strongest home theatre growth since fiscal 2022, which is consistent with a competitive TV promotional market that could hold the blended average near 31% on its own.
Does this contradict the earlier forecast of specification cuts rather than price rises?
It refines it rather than contradicting it. Specification compression remains the likely path where configuration is flexible, which mainly means laptops and phones. Where the specification is fixed and published, as with consoles, the only available levers are list price and absorbed margin, and August showed manufacturers using both.
What should a retailer do differently because of this?
Plan margin rate and unit volume separately rather than assuming they move together, and set incentives against margin dollars in memory-heavy categories. Build the promotional calendar around trade-in and financing offers for computing and consoles while keeping conventional percentage-off promotion for categories that are not memory-bound.
Is there a read-through to non-US markets?
Partially. The component cost shock is global and the console list moves were broadly international, so the cost side travels. The promotional side does not travel as cleanly, because discount conventions, instalment credit availability and pricing regulation vary substantially by market, so the Adobe metric should be treated as a US-specific test.
When will it be clear whether this was right?
The first hard read arrives with Adobe’s holiday wrap in early January 2027. Directional evidence should be visible earlier, from the structure of advertised Black Friday offers in the second half of November 2026, and confirmation or refutation of the supporting claim follows with Best Buy’s fourth-quarter results in late February or early March 2027.