The Consumer Snack Index: September 2026

Pumpkin hit 2.5x August’s rate. The Bureau is monitoring the situation.

From the Bureau of Snack Statistics, a very unofficial division of Hoot.

September brought a changing of the guard. Burgers stepped back. Pumpkin moved in. Soup made a late entrance. Ice cream declined to acknowledge the transition.

Welcome to the Consumer Snack Index, our monthly look at what Hoot users actually log. The everyday lunches, the weekend drinks, the foods that come back every fall, and the occasional entry that deserves its own investigation.

This is not a report on how perfectly anyone ate. It is a chance to notice things. Some useful. Some surprising. Some involving a single candy corn.

Here is what the Bureau found.

The Big Number: 9 million calories

Hoot users logged approximately 9 million calories in September, along with 496 kilograms of protein. The protein alone weighs roughly as much as a concert grand piano. A substantial amount of lunch.

September was also Hoot’s biggest logging month yet. Total meal logs rose 50% from August, and the average number of meals logged per person who logged food rose 35%. The single-day record was broken six times, with September 30 finishing as the biggest day in Hoot history.

More logging gives us more to look at. But it also creates an important distinction: a food can appear more often simply because there are more logs overall.

So, for the food trends below, we compare each category’s share of all logs, not just its total count. Pumpkin does not get extra credit for Hoot having a bigger month.

Seasonally Adjusted: the Great Grill-to-Gourd Handoff

Pumpkin appeared at 2.5 times August’s rate and 10 times July’s rate. The Bureau has moved it from “seasonal interest” to “ongoing situation.”

It had company. Pie, chili, cereal, and pizza all gained ground, while several summer regulars lost some of theirs.

Candy and the combined beer, wine, and cocktail category each rose 41%. Sweet potato and squash climbed 34%, while stew and pot roast rose 25%. The Bureau is calling this the Great Grill-to-Gourd Handoff.

Change in each food category’s share of logs, September versus August.

Not everyone participated.

Ice cream’s share rose 16% and continued climbing in the second half of the month. Iced coffee and cold brew finished September 14% above August, although their late-month numbers tell a different story.

Ice cream, apparently, does not report to the calendar.

The Turn: soup season is officially called

Looking only at September’s monthly number, soup barely moved: up 1% from August. That sounds uneventful. It was not.

Comparing September 16–30 with September 1–15, soup’s share of logs jumped 43%. Chili rose 33%, cinnamon 21%, and sweet potato and squash 16%.

Then September 29 arrived. Soup, chili, and stew logs reached more than double their early-September rate, making it the biggest soup day of the month.

The Bureau is officially calling soup season. Effective September 29.

There was a smaller plot twist in the pumpkin department. Pumpkin spice drinks fell 33% in the second half of September, while the broader pumpkin category kept climbing. Iced coffee and cold brew also fell 20% between the first and second halves.

The useful part is not that everyone should switch to soup. It is that a monthly average can hide a change already happening underneath it.

September did not eat the same way from start to finish.

The Game Day Report

Some food traditions showed up clearly in the logs.

Friday was pizza night. Friday’s pizza logging rate was more than double Monday’s, with Sunday a close second. Wings, nachos, and dip appeared at an 81% higher rate on Sundays than Saturdays.

The Bureau recognizes a game-day pattern. It cannot confirm what was on the television.

Weekend drinks followed their own schedule, with Saturday’s rate running at twice the midweek rate. Then Monday arrived: drink logging fell 69% from Sunday, while salad logging rose 23%. The Bureau calls this the Monday Reset.

To be clear, Monday’s salad is not a punishment for Sunday’s pizza. They are both part of the picture.

And yes, Taco Tuesday held up. Across July through September, taco logs ran 45% above the rest of the week. Folklore confirmed, with a larger sample than a single month.

Holidays, real and made up

National Cheeseburger Day, September 18, saw burgers logged at their usual rate. National Coffee Day, September 29, saw coffee logged at its usual rate, too.

Hoot users appear comfortable having coffee and cheeseburgers without a commemorative occasion.

Soup, meanwhile, took September 29 for itself.

The Night Shift Report

Noon was the busiest logging hour. But the more interesting pattern was how the foods being logged differed throughout the day.

Logs at 6 a.m. averaged a Hoot Nutrition Score of 73.4, compared with 63.2 at 11 p.m. There was also an afternoon dip: average scores fell from 70.7 at 2 p.m. to 68.4 at 3 p.m., before recovering by dinner. Late-night logs carried more sugar, too: 11 p.m. logs contained 42% more sugar than 7 a.m. logs.

That does not make 11 p.m. a forbidden time to eat. These numbers describe the food logged at different hours. They do not show that the hour caused a lower score, and they do not measure anyone’s afternoon energy level.

The useful question is more personal: What do I tend to reach for at that time of day?

A recurring afternoon snack or late-night meal can be something to understand, not another rule to follow.

The Fiber Deficit

Here is one pattern that barely changed.

On days with at least five food logs, average logged fiber was 19.7 grams, almost identical to August’s 19.8 grams. About half of those days reached 100 grams of protein, while only about one in four reached the report’s 25-gram fiber benchmark.

Those are different nutritional benchmarks, not equivalent goals. And five logs do not guarantee that someone recorded everything they ate. Still, within this more heavily logged group of days, fiber remained a consistent gap.

That is the kind of thing tracking can help bring into view. Not just how much you ate, but something you might otherwise overlook.

Hoot says: Curious about your own fiber? Look at one snack you already eat. Beans, berries, or oats are places to start. No complete meal-plan overhaul required.

Salt Watch

Sodium moved in the other direction.

Among days with at least five logs, 53% exceeded 2,300 milligrams of sodium, up from 47% in August. Average logged sodium on those days was 2,722 milligrams.

That does not tell us which meal, ingredient, or habit explains the change. It gives the Bureau something to watch, particularly as soup and stew move back into the rotation.

For an individual user, the useful part is seeing where sodium shows up across their own meals, rather than guessing from a food’s name or reputation.

Photo Evidence

About 35% of September’s logs started with a photo. But some foods were considerably more likely to get a close-up than others.

Sushi led at 58%, followed by pasta at 57% and salad at 56%. Drinks were photographed 28% of the time, donuts 24%, and pumpkin-related foods 19%.

The Bureau recognizes that not every snack requires a photo shoot.

The interesting part is that these foods still made it into the record, with or without a picture. That is more useful than a collection of meals chosen only because they look good on camera.

Leading Indicators: October is already here

A few smaller signals suggest where next month’s report might go.

Apple cider appeared at nearly three times August’s rate, although the sample remains small. Maple rose 43% in the second half of September, and hot chocolate was up 19%. Halloween-themed foods had already started appearing.

These are early signals, not headline conclusions.

Then there was the candy corn.

One candy corn was logged in September. Estimated calories: seven.

The Bureau is treating this as a cameo, not a trend. Individual candy corn reports are not a requirement.

Next month, we will be watching whether pumpkin keeps climbing, how much Halloween shows up in the logs, and whether soup settles into its newly declared season.

Most Wanted

Through all the seasonal movement, one familiar food held its position.

Banana was the most widely logged food for the second month running. Chocolate, meanwhile, appeared in one of every 27 logs.

Among popular foods, hard-boiled eggs had the highest average Nutrition Score at 98.9. Chocolate chip cookies had the lowest at 34.9. Both got logged.

That last part matters.

Tracking is not about building a record of the foods you think you should have eaten. It is about getting a useful picture of the foods you did eat.

Including the cookie.

Log it. Learn from it. Move on.

There is no perfect September hiding in these numbers. There is pizza on Friday, more soup toward the end of the month, a persistent fiber gap, and ice cream continuing its business as usual.

Some patterns are worth acting on. Some are simply interesting. You do not need to turn every observation into an instruction.

That is what we mean when we say food tracking should make you smarter, not stricter. A little more understanding. A useful next step when you need one. Then back to your day.

Log it. Learn from it. Move on.

How the Bureau counts

This report uses anonymized, aggregated Hoot meal logs from September 1–30, 2026, with earlier periods used for the comparisons described above. Hoot users are people who track their food, not a representative sample of the U.S. population. The findings describe what was logged, not necessarily everything that was eaten.

Food trends are measured as a share of all logs. Categories use keyword matches on food names, so they are not perfectly precise: for example, the candy category can include some branded protein bars. Days with five or more logs are used for the fiber and sodium comparisons, but should not be read as verified complete days of intake.

Time-of-day analysis uses local time in North American and UK time zones. Day-of-week and holiday findings use local dates for North American users. Game-day findings involve small daily category samples. These patterns are associations, not evidence of causes or weight-loss outcomes.

The Bureau of Snack Statistics is not a real government agency. The data is real.

For reporters and researchers: Statistics may be used with credit to the Hoot Bureau of Snack Statistics.