01
Market analysis
Your audience, what people search, what is working on social and how much the competition advertises, crossed with prices, rates, weather and the calendar.
Orian AI · market intelligence
Your market's week, forecast and checked.
Orian reads public data and social media to show your audience, your competition and what is working in your niche. It forecasts the week and tests every piece of content before it goes live. You get the report every Monday.
-
01 / 05
The data arrives
The Orian Matrix collects what is public: statistics, prices, central bank series, searches, social video and posts, and the active ads in each niche. Open global sources such as the World Bank give it the big picture.
-
02 / 05
The world takes shape
Country by country, every data point lands where it belongs. Where it simulates audiences, the Matrix uses statistical people built from census data, never a file on a real person.
-
03 / 05
Your market in layers
Who is online, how much they spend, how many can buy and how they connect, on the same map.
- Online audience 73.6% of the world uses the internet (2025) · World Bank
- Consumer market US$ 6,801 household spending per person, world (2024) · World Bank
- Working-age population 5.36 billion people aged 15 to 64 (2025) · World Bank
- Mobile 111.5 subscriptions per 100 people, world (2025) · World Bank
-
04 / 05
Content gets read
Post, image, script, video or ad: before it goes live, the content is read on 5 measures, from attention to action, and tested with the niche's simulated audience. Here, a personal finance Reel.
A map of 5 calculated measures. Not a brain scan, not eye tracking.
-
05 / 05
It becomes your weekly report
Every Monday, in your inbox, a report you can use the same day. No login, no dashboard. Inside:
- The week's forecast what is likely to rise and fall in your niche
- Topics the themes and formats that are working
- Audience reaction who is in your target and how they tend to react · simulation
- Posts ready to record with day, format and time
- Scoreboard what we got right and wrong last week
-
Orian AI
Every Monday, your market read before you publish.
Orian AI · market intelligence
Your market's week, forecast and checked.
Orian reads public data and social media to show your audience and your competition, forecasts the week and tests every piece of content before it goes live. You get the report every Monday.
01 The data arrives. Statistics, prices, searches, social and ads, plus open global sources such as the World Bank. 
02 The world takes shape. Country by country, every data point in its place. 
03 Your market in layers. Online audience, spending, working-age population and mobile, from the World Bank. 
04 Content gets read. 5 calculated measures and the simulated audience reaction. Not a brain scan. 
05 It becomes your weekly report. Every Monday, with last week's scoreboard.
What Orian does
Market intelligence that turns into content decisions.
02
The week's forecast
What is likely to rise and fall in your niche over the coming days, with the odds on each side and the reason why. Every forecast is checked against the real result.
03
Audience reaction
A simulated population built from census data shows how many people are in your target, their purchase intent and how each group tends to react.
04
Content testing
Posts, carousels, images, scripts, videos and ads go through 5 measures and come back with what to keep, cut or rewrite.
How it works
Three steps, every day. And an honest scoreboard at the end.
-
01Fact
Public data, every day
The Matrix collects open sources: statistics offices, central banks, prices, weather alerts, Google Trends, YouTube, Instagram and the Meta Ad Library, always through public search. Every number keeps its source and date.
-
02Forecast and simulation
Simulation
The data feeds a simulated population built from census data. That is where the week gets forecast and content gets tested before it exists.
-
03Scoreboard
Accuracy scoreboard
Every forecast is checked against the real result. Misses stay on the scoreboard and correct the next calculation.
Every block in the report says whether it is a fact, a forecast or a simulation. You always know what was measured and what was calculated.
What you get
Every Monday, a report you can use the same day.
No login and no dashboard to learn. You get the reading done and the week's strategy.
- 10-second summary
- Audience size and reaction
- Topics, formats and hooks that work on social
- Competition: how much and how the niche advertises, where ad data is public
- The week's forecast, with the odds on each side
- Posts ready to record, with day, format and time
- Last week's scoreboard
Weekly report
Niche: home renovation · worldwide
Sample week
10-second summary FactForecast
In English-language renovation content, long videos earn the most views per hour while Shorts carry the volume. Worldwide search interest usually dips over the next weeks: in the past five years it rose in only one of them. A week to build a library, not to chase reach.
577/hviews per hour, long videos (28 min)
496median views of a Short
8%of titles ask a question
168videos read in one pass
Seasonal forecast: worldwide search interest Forecast
| Week | Direction | Past 5 years |
|---|---|---|
| 1 | ↓ Usually down | up in 1 of 5 |
| 2 | ↓ Usually down | up in 1 of 5 |
| 3 | ↑ Leans up | up in 3 of 5 |
| 4 | ↓ Usually down | up in 1 of 5 |
Market Fact
Household spending per person goes from US$ 195 to US$ 46,041 across the 153 countries with data, with US$ 4,026 in the median country: the same video meets very different budgets.
Posts ready to record
- Mon“The full tour: one room, start to finish.”Long video, 28 min
- Tue“Before and after in 30 seconds.”Short
- Fri“What a kitchen remodel really costs.”Mid video, 12 min
Scoreboard
The following Monday, every line is checked against the real result: hit close or miss.
Sources: YouTube public search in English (one reading), Google Trends worldwide (5 years, weekly) and the World Bank.
Content testing
Before it goes live, your content goes through the Matrix.
Posts, images, scripts, videos and ads. Content comes back in minutes with five calculated measures, the niche audience's simulated reaction and a suggested fix.
- Posts and carousels
- Images
- Captions and scripts
- Videos
- Ads
For video and images, the Matrix shows where the eye tends to go in the first 3 seconds and a heat map of the image. It is a calculation on the content, not eye tracking of a real person. When the post goes live, the real result enters the scoreboard and calibrates the next reading.
“How much of your income goes to debt?”
Attention
A question as the hook. In this niche, 13% of titles ask a question and 28% start from “how”.
Keep · test a “how” version
Comprehension
Jargon (“debt-to-income ratio”) in the first 2 seconds.
Rewrite · “how much of your pay is already spoken for”
Emotion
Fearful tone from start to finish, with no way out offered.
Adjust · close with one doable step
Trust
Cites central bank data. No promise of returns and no product recommendation.
Keep
Action
Two calls at the end: save and send to a friend.
Cut · keep a single call
Example in personal finance, with real hook shares from an English-language YouTube reading. The Matrix talks about the audience and the content, never about where to invest.
Real numbers
The scale of what we read. Every number with a source.
General and global numbers, the same for any niche.
- of the world's population uses the internet (2025)
- 73.6%World Bank, IT.NET.USER.ZS
- people aged 15 to 64 (2025)
- 5.36 bnWorld Bank, SP.POP.1564.TO
- household spending per person, world (2024, constant 2015 US$)
- US$ 6,801World Bank, NE.CON.PRVT.PC.KD
- mobile subscriptions per 100 people, world (2025)
- 111.5World Bank, IT.CEL.SETS.P2
- countries with internet data in our map layer
- 169World Bank, latest year available
- English-language videos read in one pass, across two niches
- 382YouTube, public search
- years of weekly search history behind each seasonal forecast
- 5Google Trends, worldwide
- measures on every piece of content we test
- 5attention, comprehension, emotion, trust, action
Where we are
Deep where the data is deep.
The Matrix reads a market at full depth only where local public data is rich enough to check its own forecasts against. Everywhere else it starts from open global sources and goes deeper one market at a time. Tell us your market and we will tell you, plainly, how deep we can read it today.
What Orian has found for clients
The Matrix crosses searches, prices, weather, the calendar, social media, ads and each client's simulated audience. These are real cases from the first companies Orian works with, described by type of business.
Freight marketplace
What we read
When fuel prices jump 8% or more, truck drivers search for loads more, 1.25 times on average, for 1 to 3 weeks (it happened in 3 of 4 increases). Companies hiring freight do not search more.
What changed in the strategy
In the proposed recalibration, fuel prices weigh only on drivers, and for up to 3 weeks. For companies hiring freight, fuel is not a hook.
Matrix event study: weekly fuel prices and Google Trends.
High-end homebuilder
What we read
In the simulated audience, the number one objection is trust in the builder (59%), ahead of price (46%). Across profiles in the niche, owners speaking on camera get about 6 times the engagement. And a post whose text does not match its photo stops making sense.
What changed in the strategy
A series with the person in charge of the build answering questions on camera went into the plan, and posts were remade so the text matches the photo.
Matrix audience simulation; public reading of profiles in the niche.
Moving company
What we read
Of 22 conversations an ad brought in, the simulation estimated that about 7 would be vague requests or cheap small jobs, not real moves. In testing, a carousel with real photos reached 3.81 conversations per thousand, against 2.44 for an illustrated one.
What changed in the strategy
The ad is now judged by conversations that become moves, checked one by one. In the schedule, the real-photo carousel moved ahead of the illustrated piece.
Matrix audience lab; Meta ads.
Fashion app
What we read
The Reel with a story, a “no buy month” challenge, reached 1.8 times the profile's average Reel. The ones built on a standalone quote stayed near or below average.
What changed in the strategy
The story became a planned series: the challenge countdown continues day by day, in the same format that worked.
The profile's Instagram, Reel plays against its own baseline.
Rental real estate
What we read
After a major flood, searches for apartments to rent reached 1.8 times normal in week two and stayed above normal for 7 weeks. Searches for moving barely changed, with a small peak only in week five.
What changed in the strategy
After a weather event, expect weeks of rental demand, not a spike in moves. The weight of weather in the forecast is under review: more on rentals, less and later on moves.
Google Trends; a rental price index.
Open scoreboard
The Matrix labels what is data, what is reference and what is still a hypothesis. Every forecast ships with a date to be checked, and where it misses, the scoreboard shows it.
- 55% data
- 12% reference
- 33% hypothesis
Matrix Fidelity Index.
Privacy
Aggregated data only. No files on people.
Public, aggregated sources
Statistics offices, central banks, searches, public posts and public ads read together, never profile by profile.
Simulated people are not real people
The people in the Matrix are statistics built from census data. They have no name, no ID number and no social media account.
No lists, no tracking
Orian does not buy contact lists or match personal data. Competition is read in aggregate, without exposing brands. This site uses no cookies and no trackers.
Data protection first
Our work follows the data protection laws that apply to it. What a client sends us serves only that client's report.
Contact
Let's read your market.
Tell us your niche and your region. The first conversation is to see whether the Matrix already reads what you need.
- orianai@orianai.com