US Food & Drinks Ads, September 2026: Curiosity Gap Slipped to 28.4% and Nothing Large Took Its Place. So Where Did the Share Go?
Scope: Food & Drinks ads in the US, as the eCommerceBenchmark Ad Library tracks them. 12 brands, 571 ads active in September 2026, compared with August 2026.
Most ad reports hand you a ranking. Then they walk off.
This one tells you what to do with it.
For each of September's top three hooks, top three angles and top three concepts you get the same five things: what it is, why it matters to you, what moved since August, which real brands ran it, and one thing to steal. The run starts at curiosity gap, 28.4% of the month's ads, and ends at Side-by-Side Comparison, 2.5%. Then the small rows that actually grew, and one ad to brief this week.
- A brief decides three things, not one. Hook, angle and concept. Ask for a stronger hook and stop there, and the person making the ad chooses the other two for you.
- The default in each group is the one every shopper has already met. It is the row where your ad is most likely to look like the ad above it in the feed.
- Each default has a shelf life. In the sample of long-running ads, 100% of urgency / FOMO's ads ran sixty days or more, against 78.1% of Cinematic B-roll's. That is what tells you where your creative time is worth spending.
But first, three words that most briefs mix up.
Hook, angle, concept: three different jobs
Mixing these three up is the most common reason an ad brief goes wrong. A hook is defined as the first move an ad makes to stop a scroll: its first line or first seconds. An angle is defined as the reason the ad gives a shopper to buy. A concept is defined as the shape of the creative: how it is filmed, laid out or staged. A brief that asks for "a stronger hook" and stops there has decided one of the three and left two to chance, and the person who shoots it fills in the gaps with whatever they made last time.
| Word | In one line | Plain example |
|---|---|---|
| Hook | The first move an ad makes to stop a scroll: its first line or first seconds. | "I kept buying the wrong bag of coffee." |
| Angle | The reason the ad gives a shopper to buy. | "One scoop, no grinder, no mess in the morning." |
| Concept | The shape of the creative: how it is filmed, laid out or staged. | A hand lifting one jar on a plain kitchen counter, text on screen. |
The plain examples are made up to show the idea.
How to read a rise or a fall
Every item below shows its share of September's ads, how that compared with August, and how many brands ran it: curiosity gap, say, at 28.4% against 30% in August, on ten of the twelve brands here.
- Rising. A bigger share of ads uses it, the way solution-aware went from 1.1% to 3.9%. Brands may be copying each other, or testing a gap. Either way, it is getting more crowded.
- Falling. A smaller share uses it, the way curiosity gap went from 30% to 28.4%. That doesn't mean it failed, and nothing here records a brand stopping an ad.
- Flat. Its share held. Treat it as a stable baseline.
- Small. A move under half a point is small, like relatability's 17.5% against 17.4% in August. Check the brand count, because bold claim's rise to 2.3% sits on one brand.
One more rule. A bigger share doesn't mean a row works better. Product Demo Close-Up staged 32.6% of September's ads, and no sales, spend or profit figure sits behind that share. Each item also shows how many of its ads in a sample of long-running ads ran sixty days or more, 97.3% in that shape's case, and the sample is ranked by time on air, so the numbers read high. Read them as a hint about staying power, nothing more.
Ready? Start with how ads open.
The top 3 hooks: how ads open
Together, these three open about half of September's US Food & Drinks ads. Below each one you get how crowded it is, how long its ads lasted in that longevity sample, and how many of the twelve brands in this cut ran it.
Share of September 2026 ads by hook. Source: eCommerceBenchmark Ad Library, 2026.
Share of each hook's ads that ran sixty days or more, in a sample of long-running ads. Reads high: a staying-power hint, not a result. Source: eCommerceBenchmark Ad Library, 2026.
1. Curiosity Gap: it keeps the payoff for the next line
What it is. An opening that tells you something is there without telling you what it is, so the only way to find out is to keep reading. seattlecoffeegear.com starts on "The Diletta Bello" and leaves the promise to its headline. No benefit, no price, no verb. Just a name you either know or want to look up.
Why it matters to you. Ten brands in this cut ran it, the widest base of the three hooks here. In the longevity sample, 95.1% of its ads ran sixty days or more, a shade under urgency / FOMO's share of the same sample. So it is the safe, busy choice: the row your shopper has met most often, and the row where your next ad is most likely to sound like the one above it.
What moved. 28.4% in September, 30% in August. It is still the opening more of this cut's ads use than any other, and it gave ground while nothing large moved in behind it. Our reading: the share went out in small pieces rather than to one new default. The data records the share, not the reason.
Four ads, four ways to hold something back. seattlecoffeegear.com names the machine and stops, on "The Diletta Bello". drinkpoppi.com opens on "POV: this is your sign to treat yourself to a refreshing lemon lime poppi today 🍋", which hands the shopper the moment and lets them supply the reason. liquiddeath.com starts on "MT Everest is no longer impressive." and keeps its own product out of the first sentence entirely. spiceology.com puts a line about pot pie in front and saves the point for its headline. The one worth copying is seattlecoffeegear.com's, because each word does one job: name the thing, then stop, so line two has something left to say.
That's the crowded one. Now the one that starts in the kitchen.
2. Relatability: it starts in the shopper's own kitchen
What it is. An opening that describes the shopper's own habit back to them before the ad asks for anything. cometeer.com starts on "If you bought beans from all of these amazing roasters, it would take you so long to go through them all that the coffee would just go stale." The product is nowhere in that sentence. A cupboard is.
Why it matters to you. It sits second of the three hooks on share, but on a smaller base: seven brands ran it, against ten for curiosity gap. In the longevity sample, 95% of its ads ran sixty days or more, the lowest of these three and still high. That mix, a solid share on fewer brands, is where the room is: you can open in the shopper's kitchen without joining the queue the leading hook has.
What moved. 17.5% in September, 17.4% in August. That is a move under half a point, which is small, so treat it as a baseline that held and look at it again next month before you read anything into it.
One ad, one picture of the problem. cometeer.com is the only ad this cut's sample offers for the row, and its line spends every word on the shopper's cupboard: beans bought, beans unfinished, coffee gone stale. The order is the thing to copy. The habit comes first, the consequence second, and the brand arrives only after you have agreed with both.
That's the one with room. Now the one that runs on a clock.
3. Urgency / FOMO: it puts a clock on the decision
What it is. An opening that attaches a deadline or a shortage to the decision, so the shopper is choosing when as well as whether. It is the smallest of these three hooks by share, and the only one of the three whose whole job is the calendar.
Why it matters to you. Eight brands ran it, a wide base for a row of 28 ads. In the longevity sample, 100% of its ads ran sixty days or more, the highest of these three hooks. Read that as staying power in a sample ranked by time on air, not as a result. The practical read: this is the tool you reach for when you have a real date to name, not a mood to set.
What moved. 4.9% in September, 5.2% in August. A small move, and the row stays where it was in the order: third of the three, carried by more brands than its ad count suggests.
That's how ads open. But why would anyone buy?
The top 3 angles: why the shopper should buy
Together, these three argue nearly half of September's ads. Same five things for each: what it is, how crowded, what moved, who ran it, what to steal.
Share of September 2026 ads by angle. Source: eCommerceBenchmark Ad Library, 2026.
Share of each angle's ads that ran sixty days or more, in a sample of long-running ads. Reads high: a staying-power hint, not a result. Source: eCommerceBenchmark Ad Library, 2026.
4. Problem-aware: it names the annoyance before it names theproduct
What it is. An argument that names the annoyance first, so the shopper agrees there is a problem before the product shows up as the answer. javvycoffee.com starts on "33g protein in one caramel iced coffee, yes please", which puts the thing the shopper is short of in the first three words and leaves the brand to the label.
Why it matters to you. This is the base almost every brand in the cut argues from: ten brands ran it, the widest base of the three angles. In the longevity sample, 97.3% of its ads ran sixty days or more, the highest of these three. Crowded and steady, then. If you pick it, your job is not the angle, it is which annoyance you name.
What moved. 32% in September, 33.5% in August. It gave up ground and still argued more of the month's ads than the next two angles here put together.
Four ads, four annoyances. cometeer.com names the cupboard: "If you bought beans from all of these amazing roasters, it would take you so long to go through them all that the coffee would just go stale." javvycoffee.com names a number a shopper can check, "33g protein in one caramel iced coffee, yes please". liquiddeath.com opens on "MT Everest is no longer impressive.", a complaint about everyone else's bragging rather than about thirst. spiceology.com puts the annoyance in a line about what is inside a pot pie and leaves the rest to its headline. Copy javvycoffee.com's, because the first three words are a figure, and a figure is the one part of a claim a shopper can test without clicking.
That's the base. Now the one that grew.
5. Identity: it tells the shopper who drinks this
What it is. An argument about the kind of person the product is for, rather than the job it does. drinkpoppi.com opens on "POV: this is your sign to treat yourself to a refreshing lemon lime poppi today 🍋" and sells a shopper a version of their own afternoon.
Why it matters to you. Six brands ran it, fewer than problem-aware's ten. In the longevity sample, 79.3% of its ads ran sixty days or more, the lowest of these three angles. So it is the busiest of the three after the leader, on a smaller brand base, and the one whose ads show the least staying power in that sample. Plan on rewriting it more often than you rewrite the problem.
What moved. 10.2% in September, 9% in August. It is one of the two leaders on this page that rose enough to show in the movers chart further down.
One ad, one named moment. drinkpoppi.com's line does the work with its last two words rather than its first: "treat yourself" is the argument, and "today" is the close. Worth copying because it never describes the drink. It describes the permission.
That's the one that grew. Now the one with the fewest brands in it.
6. Comparison: it sets the product beside the thing they buy now
What it is. An argument that starts from what the shopper already buys, then puts your product next to it. javvycoffee.com opens on "Why is everyone dropping Starbucks for this protein coffee?", which names the habit in the first line and keeps its own product for the answer.
Why it matters to you. Three brands ran it, the fewest of these three angles. In the longevity sample, 86.7% of its ads ran sixty days or more, between the other two. Thirty ads on three brands is a row a few advertisers are holding open, which is exactly the shape of a row with space in it.
What moved. 5.3% in September, 5.4% in August. A small move, so read it as a row that held while the leader above it eased.
One ad, one habit named. javvycoffee.com puts a competitor the shopper already knows in the opening question and makes its own product the reply, with the rank claim left where it belongs, inside the brand's own words. The structure is the lesson: the shopper's current choice first, yours second.
That's why a shopper should buy. Now what the ad actually looks like.
The top 3 concepts: how the ad looks
Together, these three stage about 4 in 10 of September's ads. A concept is the shape of the creative: how it is filmed, laid out or staged. Fewer ads carry a concept than carry a hook or an angle, so these shares look small. Compare concepts only with each other, never with a hook or an angle share.
Share of September 2026 ads by concept. Source: eCommerceBenchmark Ad Library, 2026.
Share of each concept's ads that ran sixty days or more, in a sample of long-running ads. Reads high: a staying-power hint, not a result. Source: eCommerceBenchmark Ad Library, 2026.
7. Product Demo Close-Up: it puts the product in frame and lets it do the talking
What it is. The shape where the product is the subject, close and in use, with the words carrying the argument around it. seattlecoffeegear.com runs its "The Diletta Bello" line in this shape, one object and one label.
Why it matters to you. Eight brands staged it, the widest base of these three concepts. In the longevity sample, 97.3% of its ads ran sixty days or more. This is the control: the shape most of this cut's ads already take, which makes it the thing a new shape has to beat.
What moved. 32.6% in September, 34.4% in August. It eased and still staged more of the month's ads than the other two concepts here combined.
Four ads, one shape, four different jobs for the words. seattlecoffeegear.com gives the frame a name and nothing else. cometeer.com puts the stale-beans sentence over it, so the picture is the product and the text is the cupboard. drinkpoppi.com runs the permission line in the same shape. javvycoffee.com runs its question, "Why is everyone dropping Starbucks for this protein coffee?", over the close-up. Copy seattlecoffeegear.com's restraint: one object, one line, and the argument held back for the clip after it.
That's the control. Now the one that rose.
8. Cinematic B-roll: it sells the mood around the product
What it is. The shape where the setting does the work: the kitchen, the morning, the pour, with the product inside the scene rather than filling it.
Why it matters to you. Four brands staged it, fewer than the leader's eight. In the longevity sample, 78.1% of its ads ran sixty days or more, the lowest of these three concepts. Of the three shapes here, it is the one carried by fewest ads that still grew, which makes it the first thing to test against your close-up.
What moved. 5.6% in September, 4.7% in August. It is the second of the two leaders that rose enough to show in the movers chart below.
That's the shape that grew. Now the one that shows both options at once.
9. Side-by-Side Comparison: it shows both options in one frame
What it is. The shape that puts two things on screen together and lets the shopper do the choosing, with the words naming what the difference is.
Why it matters to you. Three brands staged it, the fewest of these three concepts, across 14 ads. In the longevity sample, 100% of its ads ran sixty days or more, the highest of these three. Read that as staying power in a sample ranked by time on air. A row this thin on brands is not a plan, but it is a clean second test.
What moved. 2.5% in September, 2.9% in August. It eased on a base of three brands, so a single advertiser's launch calendar could account for it.
So the three defaults all eased. Where did the share go?
What actually moved: the small stuff
Six of the nine leaders on this page eased, and two rose. Not one of the nine moved the way a market is supposed to move, into a new default. Instead, eighteen rows across the three rankings rose by a fifth of a point or more against August. Here they are.
Rows of the three rankings that rose by a fifth of a point or more, September 2026 against August 2026, in points. Source: eCommerceBenchmark Ad Library, 2026.
Of the nine leaders, none held exactly the share it had: six eased, two rose enough to show in the chart, and one rose too little to count. Most of the other risers are small rows that one to four brands write for, which is why none of them gets a heading of its own here. The reason to care isn't any single row. It's that several rose in the same month while every default gave ground. Let's look at one of them in real ads.
Direct question: it asks the thing the shopper is already wondering
What it is. An opening that puts the shopper's own half-formed question on screen, the way javvycoffee.com does with "Why is everyone dropping Starbucks for this protein coffee?"
Why people respond. You hear a question and you answer it before you decide whether to. Someone in a queue says "is that the one with the protein in it?" and your head is already going through what you bought last week. That is our reading of the attention, not a claim about sales.
A handful of brands carry this row: 21 ads from eight brands. javvycoffee.com's is the one ad the cut's sample offers for it, and the order inside the line is what makes it a question rather than a boast, because the brand a shopper already uses comes first and the product arrives at the end.
Several small rows rising in the same month usually hints at one ad nobody has built yet.
One ad to brief this week: the Morning Swap
Direct question, solution-aware, bold claim, selfie talking-head and montage all rose against August, and together they point at one ad. It opens on the comparison the shopper is already making, the middle shows the annoyance in their own kitchen, and the close hands them permission to buy today. This is our suggestion, built from the movers, not a finding: the library has no sales data, so it can't say the ad will work. That's what the test is for.
The format. A fifteen-second vertical video of four or five clips, one line of text on each, no voice-over needed: close to the montage shape that went from 0% to 1.1% of September's ads.
The hook. Model the first clip on javvycoffee.com's "Why is everyone dropping Starbucks for this protein coffee?": keep the shopper's current choice in the first line and nothing else on screen.
The middle. Model the second and third clips on cometeer.com's "If you bought beans from all of these amazing roasters, it would take you so long to go through them all that the coffee would just go stale.": keep the habit first and the consequence second, in the shopper's own cupboard.
The close. Model the last clip on drinkpoppi.com's "POV: this is your sign to treat yourself to a refreshing lemon lime poppi today 🍋": keep the permission and the day in the same breath, where the shopper is already deciding.
The test. Run it against your best curiosity gap ad, the hook on 28.4% of September's ads, same audience and same budget split, for two weeks, judged on cost per purchase.
Two smaller tests can ride along, each as a whole ad against the ad it replaces, never folded into this brief and never a swap of the first line alone. One: an urgency / FOMO open with a real calendar date in place of "limited time", in the row eight brands already run, taken down the day the date passes. Two: an ad written for one named shopper, modelled on drinkpoppi.com's line, which names exactly who the moment is for, run against your broadest problem-aware ad, the angle on 32% of the month's ads.
The eCommerceBenchmark Ad Library sorts your ads into the same hooks, angles, concepts and offers used in this report and shows your mix next to the market, so you can see which shapes you have never shipped.
Audit my creatives, freeOne catch before you run off. Everything above is the whole of US Food & Drinks, as this library tracks it. Your corner of it might look nothing like it.
Why you need to know the hooks in your niche
A row that is crowded across the whole of US Food & Drinks can be empty in your corner of it. Your product line, your price point, your shopper: none of those is the cut. The row you think is yours alone can be the one all five of your competitors opened with last week, and the row this page calls thin can be the only one nobody near you is using.
- Shoppers tune out an opening they meet over and over. The fourth time the same first line scrolls past, it stops being a first line.
- When every ad sounds the same, price is all that's left to compare. You end up in a discount argument you didn't pick.
- A different opening is one line in the brief either way. Writing "hook: direct question" takes the same space as writing "hook: stronger".
- Different isn't the same as right. That's why you test one ad into the row nobody near you uses, and keep the curiosity gap control that opened 28.4% of September's ads funded, instead of switching the account over.
Your 30-minute check
- Collect. Pull twenty current ads from your five closest competitors in the free Meta Ad Library.
- Label. Write down each ad's hook, angle and concept, one line per ad.
- Count. The row that shows up most is your niche's crowded row; across US Food & Drinks, that's the curiosity gap.
- Find the gap. The row that shows up least, or never; check it against the movers above.
- Brief one test. One ad for the empty row, run against your control for two full weeks.
Want every number? Here are the full rankings.
The full rankings: hooks, angles and concepts
Use these as worksheets, not scoreboards. Write your own count beside each row, from your last month of ads or your competitors' current ones. The rows where your count is highest are the ones you're actually competing in, and across this cut those are curiosity gap at 28.4%, problem-aware at 32% and Product Demo Close-Up at 32.6%.
| Hook | September share | August share | Change |
|---|---|---|---|
| Curiosity gap | 28.4% | 30% | ▼ 1.6 |
| Relatability | 17.5% | 17.4% | ▲ 0.1 |
| Urgency / FOMO | 4.9% | 5.2% | ▼ 0.3 |
| Shocking statement | 4.4% | 4.3% | ▲ 0.1 |
| Direct question | 3.7% | 3.4% | ▲ 0.3 |
| Authority | 2.8% | 3.1% | ▼ 0.3 |
| Bold claim | 2.3% | 0.5% | ▲ 1.8 |
| Newness | 1.1% | 0.2% | ▲ 0.9 |
| Visual reveal | 0.7% | 0.2% | ▲ 0.5 |
| POV open | 0.5% | 0.2% | ▲ 0.3 |
| Confession | 0.4% | 0% | ▲ 0.4 |
| Price anchor | 0.2% | 0.2% | 0 |
| Angle | September share | August share | Change |
|---|---|---|---|
| Problem-aware | 32% | 33.5% | ▼ 1.5 |
| Identity | 10.2% | 9% | ▲ 1.2 |
| Comparison | 5.3% | 5.4% | ▼ 0.1 |
| Objection crusher | 4.4% | 4.2% | ▲ 0.2 |
| Solution-aware | 3.9% | 1.1% | ▲ 2.8 |
| Mechanism | 3.5% | 3.6% | ▼ 0.1 |
| Social proof | 1.1% | 0.9% | ▲ 0.2 |
| Named enemy | 0.2% | 0.2% | 0 |
| Concept | September share | August share | Change |
|---|---|---|---|
| Product Demo Close-Up | 32.6% | 34.4% | ▼ 1.8 |
| Cinematic B-roll | 5.6% | 4.7% | ▲ 0.9 |
| Side-by-Side Comparison | 2.5% | 2.9% | ▼ 0.4 |
| Selfie talking-head | 2.5% | 1.8% | ▲ 0.7 |
| UGC Unboxing | 2.3% | 1.8% | ▲ 0.5 |
| Customer Testimonial Reel | 1.1% | 1.3% | ▼ 0.2 |
| Montage | 1.1% | 0% | ▲ 1.1 |
| Offer-first banner | 0.9% | 0.7% | ▲ 0.2 |
| Reaction | 0.9% | 0.2% | ▲ 0.7 |
| Grid swap | 0.5% | 0.5% | 0 |
| POV | 0.5% | 0.2% | ▲ 0.3 |
| Us vs. Them | 0.4% | 0% | ▲ 0.4 |
Your test card
Here is the Morning Swap from this article, set up the way a media buyer would run it, against the curiosity gap control that opened 28.4% of September's ads. Fill in the blanks with your own numbers, and write them down before you launch.
| Hypothesis | The Morning Swap, run beside my best curiosity gap ad on the same audience and the same budget split, will hold or lower my cost per purchase over two full weeks. |
| Change only this | The whole creative, as the Morning Swap brief writes it: the question opening, the two middle clips on the shopper's own kitchen, and the permission close. |
| Keep the same | Product, offer, audience, placement and budget split. |
| Funnel stage | Say first whether this ad is for prospecting (cold) or retargeting (warm). A deal in the first line is often right for warm audiences and wrong for cold ones, so test it where the ad will really run. |
| Setup | One control and one variant, same audience, same placement, even budget split. Do not edit either ad while the test runs. |
| How long and how much | Until each version has about 50 purchases, or two full weeks, whichever comes later. This is general testing practice (Meta's own learning-phase guideline is about 50 results per ad set per week), not a library finding. |
| Success metric | Cost per purchase, not clicks. |
| Stop rule | Pause a variant that has spent about 3 times your target cost per purchase with no purchase. If it is close, keep it running. |
| What counts as a win | The variant beats the control on the metric by an amount you would act on, and still does in a second week. |
September 2026 US Food & Drinks ads: what to brief next
- Curiosity gap opened 28.4% of September's US Food & Drinks ads, relatability 17.5% and urgency / FOMO 4.9% [eCommerceBenchmark Ad Library, 2026]: the lane every brief lands in by default.
- Problem-aware argued 32% of the month's ads, identity 10.2% and comparison 5.3%: together, nearly half of them.
- Product Demo Close-Up staged 32.6%, Cinematic B-roll 5.6% and Side-by-Side Comparison 2.5%, read against each other and never against a hook or angle share.
- Of its three hooks, urgency / FOMO holds longest in the longevity sample at 100% of its ads past sixty days; of its three concepts, Cinematic B-roll holds least at 78.1%.
- Eighteen rows rose by a fifth of a point or more, led by solution-aware from 1.1% to 3.9% and bold claim from 0.5% to 2.3%, each carried by three brands or fewer.
- Label twenty of your five closest competitors' current ads before your next brief: the row that never appears there is the one to test.
US Food & Drinks ad benchmarks: common questions
Keep one ad on curiosity gap funded as your control: it opened 28.4% of September's US Food & Drinks ads, so it is the row your shopper already knows. Then brief exactly one ad, the Morning Swap, and run it against that control on the same audience and the same budget split for two full weeks. Judge it on cost per purchase and nothing else. At the end of the fortnight you will have one decision you can defend, instead of a ranking you read and filed.