Back-to-school Facebook ads

Shoppers start earlier than most campaigns do: NRF surveyed 7,677 consumers on 1 to 8 July 2026 and found 62% had already started. Once the season ends, Meta's Ad Library keeps showing only ads that are still running, so the rival ads you can study disappear unless you saved them.

Reading rival ads well means knowing which are really about school, which angles repeat, and what your own customers have said that points the same way.

In a count of 543 ads, only 46 used back-to-school wording. The sections below show what those 46 did, and a weekly sheet turns rival angles and customer signals into a ranked test list for next season.

The label misleads because a pulled set contains mostly ads that are not about school: only 46 of 543 ads in the count (8.5%) use back-to-school wording.

The Ad Library is Meta's public database of ads currently running. A back-to-school set pulled from it needs filtering before you read it.

The count covered 543 US ads with start dates from 13 July to 8 September 2026, from 324 brands, collected on 23 October 2026. No brand has more than two ads in the set, so ad counts per brand say nothing about spend.

To reproduce the 46, search the set for "back to school", "BTS" (back to school) or "new school year" in copy or headline. That leaves 46 ads from 31 brands. The rest include degree or course ads that matched on the word "school".

Back-to-school ads are different from school or university recruitment ads, which Facebook Ads for Education covers, and from pure sale events such as Black Friday.

Rivals mostly led with a discount, a style collection or a routine and time-saving story. Each of the 46 ads was assigned one primary angle, and the coding is a judgment call.

Only 11 of the 46 carry a numeric discount such as a percentage or dollar amount. Image ads are 11 of the 46 (24%), against 30 of the 543 (5.5%) overall. Video is 19 of the 46 (41%), against 351 of 543 (65%).

None of the 46 carries a ship-by or arrive-before-school message, and none leads with a short list of what to buy first. Those are open slots, not proven wins.

Shoppers buy early and again late, so rival ads you see in September describe the second push. NRF found that 46% of shoppers who had not yet completed half their purchases were waiting for better deals, and 47% planned to buy only essentials first (NRF, July 2026). Another NRF release found 52% of K-12 (kindergarten through grade 12) shoppers finish one to two weeks before school starts.

Deloitte surveyed 1,207 K-12 parents on 22 to 29 May 2026 and found 48% of planned spend lands by the end of July and 31% in early August. It also found that 31% are "hyper value-seekers", meaning they use four or more cost-saving behaviors.

In the count, 26 of the 46 back-to-school-worded ads (57%) started on or after 24 August. That counts only ads still live at collection: 539 of the 543 had no end date. Testing an early-season launch is better than assuming the early window is empty.

Score each angle on four inputs, then test the top three. This is a scoring model, not an official method.

1. Filter. Pass: the ad's copy or headline names a school moment. Fail: drop it.
2. Code and count. Give each ad one primary angle and count brands, not ads.
3. Score. Score = 2 × customer signal + rival score + untested − effort.
   • Customer signal: 0 = no mentions, 1 = mentions from 1 to 2 people, 2 = mentions from 3 or more people, 3 = 3 or more people across 2 or more sources such as reviews and comments.
   • Rival score: 0 brands scores 0, 1 to 2 scores 1, 3 to 5 scores 2, 6 or more scores 3.
   • Untested: 0 = run in the last 12 months, 1 = run once, 2 = never run.
   • Effort: 0 = reuse existing assets, 1 = new edit, 2 = new shoot.
4. Brief. For the top three, write one hypothesis, one variable, one success metric and a stop rule, using an ad creative brief.

An angle no rival uses may be open, or it may not sell. That is why customer signal counts double. Pull those signals with review mining. Ads are evidence of what rivals run, not of what works: the Meta Ad Library shows no spend or results for ordinary commercial ads.

• Myth: A set of ads labelled "back to school" is all about school. Fact: Only 46 of the 543 ads counted use back-to-school wording; the rest include degree or course ads that matched on the word "school".
• Myth: A rival ad that keeps running must be earning money. Fact: It shows only that the advertiser kept it live, since the Ad Library shows no spend or results for ordinary commercial ads.
• Myth: An angle no rival uses is a free opening. Fact: It may be open, or it may not sell, which is why customer signal counts double in the score.

Deepsolv's creative intelligence platform combines competitor activity, customer signals, ad-performance data and past ad learnings into ranked weekly test plans and execution-ready creative briefs, which is what the scoring model above does by hand. Deepsolv also publishes the free-to-browse Back to School page behind the 543-ad count, and its own ads are not in that set. See how Deepsolv pricing works.