Paid Media
Q4 Planning

Holiday Demand Is a 13-Week Curve. Is Your Media Plan Chasing a Three-Week Spike?

Rosie Osmun
·
November 21, 2026
·
10
Minutes
Holiday Demand Runs 13 Weeks. A broad lime shopper demand curve from October through January with a narrow black ad cost spike at BFCM labeled ad costs +84%
Table of Contents
IntroductionKey InsightsConclusion

Ask most brands when the holiday season happens and they will point to five days: Thanksgiving through Cyber Monday. You may even be hearing the new trendy term, Cyber 5.

But when we actually follow the money, the picture changes. Shoppers spread their holiday spending from early October into January. Ad auction pressure, by contrast, spikes for about three weeks and then falls away.

Most media plans follow the cost curve instead of the demand curve. Brands arrive late, judge the peak on flattering platform metrics and pull back in December, right as demand holds and competition thins.

We pulled the most reliable third-party data we could find to map both curves: platform earnings filings, measurement firms, consultancy surveys and peer-reviewed research. We left out agency-produced benchmarks. Here is what the data shows, and how a sophisticated media plan actually builds with it.

Most consumer holiday dollars land outside Cyber Week

Adobe tracks US online retail spending across retail sites. In the 2025 season (Nov. 1 to Dec. 31), shoppers spent $257.8 billion online. Cyber Week accounted for $44.2 billion of that, about 17%.

Adobe's own daily figures let us split the season into three blocks:

  • Nov. 1 to 26: about $93.2 billion, or 36%
  • Cyber Week: $44.2 billion, or 17%
  • Dec. 2 to 31: about $120.4 billion, or 47%

The 30 days after Cyber Monday were the largest block of online spending all season. Cyber Week days ran hotter, at roughly 2.3 times an average day, but the other 56 days still carried five of every six dollars.

The 2026 forecast keeps the same shape. Adobe expects $275.1 billion online for November and December, with Cyber Week at $47.5 billion, or 17.3%. Adobe also forecasts $95.8 billion in October. Count October and Cyber Week drops to roughly 13% of fourth-quarter online spending.

Where 2025 US online holiday dollars landed: Nov. 1 to 26 $93.2 billion (36%), Cyber Week $44.2 billion (17%), Dec. 2 to 31 $120.4 billion (47%) of $257.8 billion. 2026 forecast: Cyber Week $47.5 billion, 17.3% of $275.1 billion. Source: Adobe

Shopper surveys point the same way. Deloitte's 2026 Holiday Retail Survey asked 4,000 US consumers when they plan to spend. Households planned:

  • $123 before the end of September
  • $151 in early October
  • $178 in late October
  • $240 in early November
  • $359 in late November
  • $330 in early December
  • $204 in late December
  • $70 in January

Late November is the single biggest period, but it holds only 22% of the total spending budget. Early December is close behind. And 28% of holiday budgets are expected to be spent by the end of October, up from 24% last year and the highest share in six years. Intent to shop October sales events has climbed from 24% of shoppers in 2023 to 42% this year.

December isn't a tail, either. Salesforce found US online sales grew 9% in the last two weeks of December 2025, against 4% for the season overall. Pickup orders peaked on Dec. 22 at 35% of online orders, and Adobe saw curbside pickup hit 39% on Dec. 23.

After Christmas, the flow starts to reverse. About one in seven of the season's returns landed between Dec. 26 and 31. Gift cards topped holiday wish lists at 47% in NRF's survey, and most get redeemed in January and February. That window is the season's fifth quarter, which most plans treat as dead time.

Ad spend often focuses on the narrow, most expensive window

The cost side looks very different for most brands, however. Lebesgue, an analytics platform for ecommerce brands, tracked the median brand's daily Meta CPM against its September 2025 level:

  • Within about 10% of September until Nov. 10
  • Up 27% by Nov. 20
  • Up 53% on Thanksgiving
  • Up 84% on Black Friday
  • Up 71% on Cyber Monday
  • Back to 20% above September by Dec. 3

The 2025 spike also started later and ran smaller than the year before. Black Friday week was 59% above September in 2025, against 80% in 2024. Costs first ran 10% above September on Nov. 10 in 2025, versus Oct. 14 in 2024.

After the peak, prices kept falling. Supermetrics, which analyzed $675 million in Meta spend from more than 5,000 retail advertisers, found Christmas-week cost per click ran 11% to 17% below September. Superads' median retail Meta CPM fell from $21.68 in November to $13.49 in December, and January 2026 was the cheapest month it tracked across the year.

Demand runs long, costs spike short: Deloitte planned holiday spend per household by period from before September's end ($123) to January ($70), peaking at $359 in late November, against a Meta auction cost index with September 2025 at 100: 113 on Nov. 10, 127 on Nov. 20, 184 on Black Friday, back to 120 by Dec. 3, and Christmas-week CPC at 83 to 89. Sources: Deloitte, Lebesgue, Supermetrics, Superads

Advertiser budgets follow the cost spike more often than the demand curve, which is something we typically see in audits and correct in our media planning.

In Supermetrics' data, Black Friday week alone took 10% to 13% of Meta budgets spent from September through December, at 1.8 to 2.8 times the spend of a normal September week. Then budgets dropped. Cyber Monday week kept only 56% to 79% of Black Friday week's spend, a cut of 21% to 44%, and spend kept tapering from there.

That pullback lands right on top of the biggest block of the season's spending.

How hard the peak hits depends on category and channel. Lebesgue measured Black Friday week CPM increases of 116% for home and decor, 76% for electronics and sporting goods, 66% for fashion and 42% for beauty. Billy Grace, an attribution firm, found the same week ran 72% above baseline on Snapchat, 51% on YouTube, 45% on Meta and TikTok, 26% on Pinterest, 25% on Bing and 22% on Google.

This year starts from a higher floor. Meta's own earnings show the average price per ad impression in the US and Canada up 14% year over year in the first quarter of 2026, 20% in the second and 23% in the third, while ad impressions there grew just 8% in the third quarter.

Black Friday is part of Q4, not the whole quarter

This is where the usual argument goes wrong. A lot of commentary treats the BFCM peak as a trap. The data doesn't support avoiding it, but it does support planning longer-range.

In Lebesgue's data, Black Friday week CPMs rose 59% over September, but conversion rates rose 48%. Cost per purchase fell 20%, and Meta-reported ROAS rose 19%. Billy Grace found the same pattern: cost per order down about 18% and ROAS up about 26%. For the median brand, on platform metrics, Black Friday week was the most efficient week of the season.

So the peak is a reasonable buy. What a typical plan misses sits around it, for four reasons.

Averages hide the last dollar. A falling median CPA says nothing about the final dollars added at the peak. In one large study of TV advertising across 288 consumer brands, the last dollar returned less than it cost for more than 80% of them. Nielsen finds about a quarter of channel investments sit past the point of best return.

Peak-week conversions are the most likely to happen anyway. Platform-reported purchases include people who were already buying, especially when audience exclusions are murky. In a Facebook campaign studied by Northwestern and Meta researchers, a simple exposed-versus-unexposed comparison showed 416% lift, while the randomized test showed about 77%. The same intent spike that makes Black Friday CPA look great also makes it the week where the share of truly incremental conversions could be lowest.

ROAS ignores the discount. Adobe expects average discounts of up to 14% in early November, 21% before Thanksgiving, about 30% in Cyber Week and 10% to 15% through December. Take an illustrative product with a 50% gross margin at full price. At 30% off, its margin falls to about 29%. At 10% to 15% off, it stays between 41% and 44%. A cheaper CPA Cyber Week order can still earn less profit per ad dollar than a December order. While ROAS might look great, margins may not.

The biggest gap opens after the peak. Brands cut Meta spend by 21% to 44% after Cyber Monday week, yet Dec. 2 to 31 carried 47% of online holiday dollars, late December grew fastest and costs fell back toward or below September levels.

The money a typical plan leaves behind sits in what it under-buys (the October awareness build, December demand capture and January reach) and in what it can't see (incrementality and margin).

A typical plan vs. a sophisticated plan

A typical plan vs. a sophisticated plan. Budget timing: typical launches in mid-November, stacks budget into Black Friday week, then cuts 21 to 44% after Cyber Monday; sophisticated paces to demand, building in October near September costs and sustaining spend through December. Judging the peak: average platform CPA and ROAS vs. marginal, incremental and margin-adjusted return with holdouts set up before the peak. Channel mix: Meta and Google take about 80% of Cyber Week spend vs. shifting peak weight to channels that inflate less and testing under-funded growth channels off-peak. Measurement: platform ROAS with optional experiments vs. MMM calibrated with incrementality tests covering retail media and creators. Offers: one sitewide Black Friday offer vs. timing each category's offer to its deal window and buyer mindset. Late December to January: dead time vs. the season's cheapest reach against gift card redemption, gift-related purchases and January intent

The measurement row is where most brands are stuck. Nielsen's 2025 survey of about 1,400 marketers found 85% feel confident they measure full-funnel ROI, but only 32% actually measure traditional and digital spend together. In The CMO Survey, 70.6% of marketers respond to pressure to prove value by focusing on short-term results, and only 57.7% run experiments. IAB's State of Data 2026 found 75% of marketers say their measurement falls short.

Timing by category matters, too. Not just for discounts but for understanding what buyers are likely to respond to and when.

Adobe expects the deepest discounts on electronics and computers on Cyber Monday, on toys, TVs, apparel and furniture on Black Friday, and on sporting goods on Thanksgiving. Google's When to Buy Guide puts the best fragrance deals in the two weeks after Black Friday weekend, handbags in mid-December, and large luggage and motorized tech toys in late-December clearance. A single sitewide calendar misses most of those windows.

Beyond discount optimization, there’s also buyer mindset timing. People shopping for gifts in early October have a different mindset than those shopping on December 15 last minute. And, people are also still buying for themselves throughout this time period as well!

How much efficiency is really on the table

In audits, we often see a typical media plan leaving 20% to 40% of its potential on the table. This has been validated repeatedly after restructuring client media plans.

From independent research, the conservative end is about 15% to 20% from moving money within the same budget. McKinsey found 15% to 20% of marketing budgets could be reallocated or returned without losing ROI, across about 400 client engagements. Google's Meridian modeling projects 14% to 16% more sales at the same budget for its partners. Vendor media mix modeling case studies land near 17%.

The top end comes from building better measurement. Analytic Partners' 2026 study of 455 marketing decision-makers found the top 22% of organizations earn about $40 more return per $100 of media, roughly 40% above their peers. Nielsen found half of channel plans are underinvested, by a median of 52%, and that fixing it lifts ROI by a median of 50%, though that gain comes from spending more where it works rather than shuffling the same dollars.

For specific tactics the gap can be total. Branded search and retargeting can return close to zero incremental sales, while platform reports show them as top performers.

Put together with creative improvements, a range of roughly 15% to 40% is defensible: the low end from moving money, the high end from building the measurement to know where to move it + creative that works harder.

We have seen the cost of planning from last year's playbook firsthand. Heading into BFCM 2024, predictive modeling for MagBak showed that holding the prior year's spend levels could raise Meta customer acquisition cost by as much as 60%. The response was to diversify channels and validate new audiences + creatives well before the peak. Blended revenue was up 40% year over year by January. And, that same result was virtually repeated the following year.

What to do with the rest of this season

There is still time to act on most of this for 2026.

  1. Don't go dark on Dec. 2. Plan December as its own campaign, with budget that holds through the shipping cutoff peak (and/or in-store availability for retail brands).
  2. Judge the peak on margin, not just CPA. Build discount depth into your ROAS/CPA/MER targets so your plan adjusts for profit not just platform vanity metrics.
  3. Time offers to your category's window. If you sell fragrance, handbags, luggage, tech toys or highly gift-able items that ship fast, your best weeks may come after Cyber Monday.
  4. Plan the fifth quarter now. Late December through January brings the season's cheapest reach, plus gift card redemption, returns and fresh January intent.
  5. Start next year in October. That is when a growing share of budgets gets spent and costs still sit near their September baseline.

Holiday demand is a long curve. Plans built only for the spike leave money behind on both sides of it.

Want a second set of eyes on your holiday plan?

Neon Growth can map your budget to your category’s demand curve, from the October build through January.

Talk to a Partner

Sources

Rosie Osmun

COO

Agency Operations | Digital Creative Strategy | Paid Media Strategy

Subscribe to our newsletter

Get the latest growth marketing insights delivered to your inbox.

Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
Or make Neon a preferred source on Google

Improve marketing. Prove impact.

Bring your growth target and your biggest roadblocks. We'll show you where the next gains are.

Talk to a Partner →