Evidence before excitement
A trend list is easy to fill with guesses. This one has to show dated sources for every trend, and it keeps what was observed apart from what is estimated.
Find marketing trends with AI. Five signals, each with its sources.
DFIRST Trend discovery is an AI trend finder that helps you find marketing trends with evidence. Name a topic and a market, and it searches the live web for five trends, each with who it affects, dates, source links and a content experiment to try. It will not call a trend rising without comparative evidence.
- Five trends in every run
- Dates and source links for each
- A content experiment for each trend
What you get. Trends you can act on.
Two of the five trends from two runs, shortened. In the real result, every source is a link.
Example output
Trend 2 of 5: sparkling water as an evening drink
Who it affects: adults in Spanish cities who are cutting back on alcohol.
Dates: signals from spring and summer 2026, compared with coverage from 2024.
Sources: a 2026 food and drink trade report; a national newspaper feature on alcohol-free evenings.
Content experiment: three reels of a ZESTA lime can poured into a wine glass at sunset. Compare saves with the usual daytime posts.
Example output
Trend 4 of 5: run clubs as a social scene
Who it affects: new runners who join for company as much as for fitness.
Dates: coverage from 2025 and 2026.
Content experiment: a KINETRA post inviting followers to a Saturday group run, tested against a product-only post.
How to find trends with DFIRST.
Three steps. The first one starts on this page.
- 01
Name the topic and market
Describe what you want to scan in the box above: the category, the country and the period. Ask it to separate new signals from habits that are already common.
- 02
DFIRST searches the web
It gathers public evidence with links and dates, then picks five trends and checks each against what it found. Without comparative evidence, nothing is called rising.
- 03
Get five trends to act on
Each trend comes with who it affects, its dates, source links and one content experiment to test. The list is saved to your session and your Data Room.
Made for brands. Not a blank prompt box.
From signal to a post
Every trend ends with a content experiment, so the list turns into something the team can make and measure. Take one into a session and Firsty can draft it with you.
One question, or every day
This tool answers the question you ask, when you ask it. For trends watched for you, the Trends module follows posts, sounds, formats and creators on Instagram, TikTok and YouTube and shows what is gaining ground.
Who uses it.
Social and content teams
Ideas for next month's posts, based on what is really happening in the category.
Strategists
A quick scan of a market before a campaign, a launch or a new product.
Agencies
Trend pages for a pitch, with sources the client can open.
At a glance.
The facts about Trend discovery, as it works in DFIRST today.
- Input
- A topic and a market, in your own words
- You get
- Five trends, each with who it affects, dates, source links and a content experiment
- How it works
- A web search for evidence, then the trend list
- Model
- Gemini 3.8 Flash with web search
- Where it goes
- Your session and your Data Room
Questions people ask.
About Trend discovery and how it fits the rest of DFIRST.
How do I find marketing trends with AI?
Describe the topic and the market in the box at the top of this page and continue in DFIRST. It searches the web, then returns five trends, each with who it affects, dates, source links and a content experiment you can try.
Does the trend finder search the web?
Yes. The first step searches the live web and collects evidence with links and dates, and the trend list is built from that evidence. It reads public web pages, so it sees what is written about social platforms rather than every post on them.
How is this different from the Trends module?
This tool runs once, on your question, from public web sources. The Trends module keeps watching Instagram, TikTok and YouTube around your brand and shows what is rising over time. Use this for a quick scan and the module for the everyday view.
How does it decide something is a trend?
It needs evidence. Every trend must come with dated sources, and the tool is told not to call anything rising without comparative evidence, such as an earlier and a later source. Where the evidence is thin, it says so.
Can it predict what will go viral?
No. It reports what public sources show now, with dates, and suggests one experiment per trend so you can test it with your own audience. Your own results are the real answer to whether a trend works for your brand.
Next steps. Tools that pick up where this one ends.
Start it here. Finish it in DFIRST.
Your trend discovery draft waits on this device until your access starts. DFIRST opens to people on the waitlist first.
Updated