About The Froth
We're cafe lovers who can't walk past a specials board, so we track which new drinks and dishes are catching on across 8 Australian cities.
The Froth in numbers
- 8Australian cities
- 30,172cafes we watch
- 38cafe trends tracked
- Dailyfigures rebuilt each morning
Where the figures come from
Six kinds of public information, and a person checks every name before anything counts.
- Posts
Cafe and customer posts
What cafes and their customers post publicly about what they're pouring, read in Sydney, Melbourne, Brisbane and Gold Coast. A customer's post that names the cafe counts the same as the cafe's own. We read cafe posts from Jul 2025.
- Reviews
Customer reviews
Customer reviews that name a drink or dish, in every city we cover. They go back years, further than the posts we read.
- Menus
Cafe menus
Cafes' current menus. A drink or dish on a menu counts that cafe for the last 90 days, and menus give the prices.
- Searches
Searches
Monthly searches in each city and across Australia, with every name a trend goes by added up. They show demand, not cafes.
- Creators
Food creators
Posts by food creators in Sydney, Melbourne, Brisbane and Gold Coast. Counted separately and never as a cafe selling it, because a creator can post about a drink or dish before cafes do.
- Press
Food press
Food articles. They add places to a trend's “Where to get one” list and map, but never to the share of cafes.
- A person checks every name
Checked by a person
Rising words in posts and searches flag a possible new drink or dish. A person decides which ones become trends, and checks every name a trend goes by before a post using it counts.
- Figures for each city
Figures
For each city: the share of its cafes with a trend, the change among the same cafes over 90 days, and searches with their base months named. Figures are never added up across cities.
- A stage in each city
Stages
Each city gives every trend one of five stages, from fixed rules checked every morning.
How we count, in full
- A cafe counts from its first post or review
- A cafe counts for a trend from the first public post or customer review we found about it. A drink on a cafe's current menu also counts that cafe for the last 90 days. The first post or review is when we first saw it, not the day the cafe started selling it.
- Share of cafes, not a count
- The share is the cafes we see well (an Instagram post, or a current menu we have read) that have the trend in a post, a customer review or their menu, divided by every cafe we see well in the same period. Cafes we only know from Google reviews are left out of both, because reviews rarely name a drink. A chain counts once, however many shops it has. The number of cafes we follow keeps growing, and a raw count would rise for that reason alone.
- The same cafes, the last 90 days against the 90 before
- A change counts only cafes we hold posts or reviews for from at least 180 days back that have recent ones, so newly added cafes never make a trend look like it is growing. A percentage needs at least 10 of those cafes with it before; under that we give the two counts. The windows end on the latest day with cafe posts or reviews (now 13 Nov 2026), not today, so late posts never show up as a drop.
- Searches, with their base months named
- Search volume is the latest month of searches, with all of a trend's names added up. Search change compares the latest three months with the same months a year earlier when we have them, otherwise the first three months we collected. The base months are always named next to the figure.
- Every name, counted together
- Cafes sell the same drink under different names. Each trend has a checked list of names, and a post using any of them counts for the trend.
- Unfinished periods are never a drop
- A month or quarter that isn't over yet is marked as partial on charts, never shown as a fall.
The five stagesHow the stage is chosen
The rules are checked in this order: fading, then emerging, rising, peaking and mainstream. The first rule that fits wins.
A cafe has a trend when it is in the cafe's own posts, customer posts, customer reviews or its current menu. In Perth, Adelaide, Canberra and Hobart we read reviews and menus only.
How the stage is chosen
The rules are checked in this order: fading, then emerging, rising, peaking and mainstream. The first rule that fits wins.
A cafe has a trend when it is in the cafe's own posts, customer posts, customer reviews or its current menu. In Perth, Adelaide, Canberra and Hobart we read reviews and menus only.
Each city gives every trend one stage, from fixed rules checked every morning.
Emerging
Emerging: the rule
Fewer than 5 of a city's cafes have it, or the first cafe post or review we have is under 120 days old, or creators are posting it before any cafe has it. It also covers anything on a small share of cafes that is holding steady.
A few cafes have it, or creators have started posting it.
PistachioIn all 8 citiesRising
Rising: the rule
At least 10 cafes had it in the last 90 days, and either the number of the same cafes with it grew by a quarter or more on the 90 days before (counting only cafes we have read through both periods, and at least 10 of them), or it is new among them, or the city's searches rose by a quarter or more while those cafes held steady.
More cafes each quarter, or searches climbing.
Mont Blanc coffeeIn 7 of 8 citiesPeaking
Peaking: the rule
At least 17.5% of the cafes we see well had it in the last 90 days but searches are falling or well below their 12-month high, or fewer of the same cafes have it than in the 90 days before (from at least 10) while searches hold.
Widespread, but demand is easing off its high.
Cold brewIn 2 of 8 citiesMainstream
Mainstream: the rule
At least 28% of the cafes we see well had it in the last 90 days, and nothing says it is rising, peaking or fading.
On a big share of cafes, holding steady.
MatchaIn 4 of 8 citiesFading
Fading: the rule
No cafe had it in the last 90 days, or the number of the same cafes with it fell by a quarter or more on the 90 days before (from at least 10 cafes) while searches did not rise, or searches are under half their 12-month high and cafes are not picking it up.
Cafes or searches dropping away.
Dubai chocolateIn all 8 cities
What the numbers can't tell you
We only see what is shared publicly, so every count is a floor.
Every figure is an estimate
The figures are estimates from public information and may be incomplete, out of date or wrong. Prices, menus and opening hours change, so check with the cafe before you go. Our Terms of use explain what you can rely on.
Quiet cafes are missed
A cafe that sells a drink but never posts about it, and whose customers and menu never mention it, won't be counted. Shares are measured against cafes with an Instagram post or a current menu in each window, so quiet cafes don't skew them, but what they sell is missed.
First seen is not where it started
The earliest post or review we have is when we first saw it, not launch day. A trend may have started earlier or somewhere else, so we never name a cafe or suburb as where a trend began.
Small numbers move a lot
In Sydney about 1,829 cafes had an Instagram post or a current menu we read in the last 90 days, so one cafe is roughly 0.1% of share; in smaller cities one cafe counts for more. A small trend can change stage on one or two cafes.
Searches are seasonal
When a search change is measured against the first three months we collected rather than the same months a year earlier, a summer drink can rise for seasonal reasons. The base months are always named so you can judge.
Drinks and food, counted apart
We report drinks, drink formats (such as cold foam) and flavours used in drinks, and food: dishes, bakery items and sweets. Each kind has its own matching rules, so a drink named after a food (a Dubai chocolate latte) is never counted as the food. Every list can show drinks or food on their own.
Captions stay private
We show counts, dates, cafe and suburb names and links to public posts, never the text of anyone's post.
How often it updates
- OvernightNew posts, reviews, menus and searches come in
- Every morningFigures and stages are rebuilt, then the site refreshes
- 13 Nov 2026The latest day with data, where every window ends
How to cite us
Name The Froth, link to the page you quote and give its data date.
“Source: The Froth, thefroth.au/trends/pistachio, data to 13 Nov 2026.”
See every trend we trackFacts to quote
- The Froth watches 30,172 cafes across 8 Australian cities (Sydney, Melbourne, Brisbane, Perth, Adelaide, Gold Coast, Canberra and Hobart), counting the bakeries, dessert and tea shops that sell the same drinks and food, as of 13 Nov 2026.
- It tracks 38 cafe trends, and gives each one its own stage in each city.
- Its figures come from cafe and customer posts (in Sydney, Melbourne, Brisbane and Gold Coast), customer reviews, cafe menus, food press and search demand. Reviews go back years; cafe posts are read from Jul 2025.
- Food creators' posts are counted separately, never as a cafe selling a drink or dish.
- The figures are rebuilt every morning, Sydney time, with new data added overnight.
Questions about the method
Where does The Froth get its data?
From public information: posts from cafes and their customers (in Sydney, Melbourne, Brisbane and Gold Coast), food creators' posts, customer reviews, cafe menus, food press and search demand in each city and across Australia. It watches 30,172 cafes across 8 cities. Reviews go back years; cafe posts are read from Jul 2025. Food creators' posts are counted separately.
Does the earliest post or review mean that cafe started the trend?
No. The earliest post or review we have is only the first one we collected. We only see what we collect, so a trend may have started earlier, at another cafe or in another city. The Froth never names a cafe or suburb as where a trend began.
Why does The Froth use a share of cafes instead of a count?
Because the number of cafes we follow grows over time. A share (cafes with the trend divided by the cafes we see well, with an Instagram post or a current menu, in the same period) shows real spread; a raw count would rise just because we added more cafes.
Why do the 90-day windows not end today?
They end on the latest day with cafe posts or reviews. Both arrive in batches, so ending on today would make the last few days look empty and show a fake drop.
Can I quote The Froth's figures?
Yes. Each trend page has a press card with a ready-made quote and the figures. Please name The Froth as the source, link to the trend page and give the data date (currently 13 Nov 2026).
Does The Froth publish people's posts or captions?
No. Captions and raw posts stay private. The site shows counts, dates, cafe and suburb names, and links to the public post where we saw a trend.