Decision forecasting
We tell you what people will do before you spend, and we keep score
Informativ forecasts how each group of people will respond before you launch, then grades every forecast against the results.
The problem
What weak insight, planning and execution cost each year
Whether you are winning customers, investors or voters.
- Insight ~$22B research that decisions never use Origin
- Planning ~$11B plans that are never graded Flight Plan
- Execution $26.8B ad spend wasted Co-pilot
How these figures were calculated
Insight and planning are Informativ estimates. Insight: $56B global market research spend, 2024 (ESOMAR) × 40% of key decisions made without consumer insights (Market Logic, 2024). Planning: $111.3B US marketing consulting, 2026 (IBISWorld) × 10%, the minimum share of spend 78% of decision-makers believe is wasted through poor measurement (Haus, 2026). Execution: ANA Programmatic Transparency Benchmark, Q2 2025, global, annualized.
What is different
What Informativ does differently
Informativ forecasts how people will respond before you spend, then checks each forecast against what actually happened.
A forecast before you spend
We forecast results before launch.
Person by person, or group by group
We show how each person or group is likely to respond.
The reason behind each forecast
Each forecast says why people will respond.
Graded against the result
Every forecast is locked before launch and scored against what happened.
The technology
How the system works
- 15 years of research
- Proprietary models
- No third-party AI
- No personal data
Accountability
Every forecast is sealed, then graded
- Sealbefore launch
- Runin your channels
- Gradeagainst results
Tests on historical data. A live test against a control group is next. See our journey.
Working with us
Three steps to a graded forecast
- 1Back-testWe forecast a past result blind. You grade us.
- 2ForecastSealed before your next launch.
- 3Run it liveIn your ad platforms or your own channels, with Co-pilot guidance if you want it.
Get in touch
See a forecast on your own data
Send us a past decision. We will forecast it blind, with no integration or media spend.
Products
Three products, four ways to buy
Organizations buy by project or by subscription. Technology platforms license our forecasts for their own clients. Every forecast is sealed and graded.
For organizations
Brands, agencies and other organizations
- 1 · Understand Origin Who buys, why, and what holds the rest back. Addresses ~$22B
- 2 · Plan Flight Plan A sealed forecast and a plan ready to run. Addresses ~$11B
- 3 · Run Co-pilot While your campaign runs, you send us results daily or weekly. Ava sends back the changes to make, and your team makes them. Addresses $26.8B
Co-pilot works from a Flight Plan, so the two are bought together.
Two ways to buy
- What you buy
- One Origin study, one Flight Plan, or Co-pilot for one campaign.
- How it’s billed
- A fixed fee per study or campaign. Co-pilot is billed monthly while the campaign runs.
- What you buy
- Any one product on an annual plan, or all three as Informativ Complete.
- How it’s billed
- An annual fee, at a lower cost per study or campaign than buying by project.
For technology companies
Enterprise license and infrastructure
Two ways to bring graded forecasts to the customers on your platform.
- What you buy
- Informativ’s forecasts and guidance in your product. One agreement covers every client on the platform, with no rebuild needed.
- How it’s billed
- By usage, above an annual minimum.
- What you buy
- Ava running inside your decision path, under your name.
- How it’s billed
- An integration fee, then usage above an annual minimum.
- Ad-buying and media platforms
- Customer engagement platforms
- Data and identity platforms
- Research and insights providers
- Recruitment and investor platforms
Which option fits
Where to start
| If you… | Start with |
|---|---|
| want to understand who your buyers are and why they buy | Origin, as a project |
| are about to launch a campaign and want to know what will work | Flight Plan, as a project |
| want guidance while a campaign is running | Flight Plan with Co-pilot |
| run campaigns or outreach throughout the year | A subscription, or Informativ Complete |
| run an ad, engagement, data, research or recruitment platform | Enterprise license or infrastructure |
Not sure? Most clients start with a back-test: we forecast a past decision blind and you grade us. Get in touch.
Beyond marketing
Wherever a decision depends on how people respond
- Account opening
- Patient enrollment
- Investor sign-up
- Candidate recruitment
- Resident communication
- Business attraction
Our journey
How we have built and tested Informativ
Milestones as they complete, and results as they are graded. Each stage starts when the one before it passes.
-
Working prototype
Running on real client material.
-
First complete version of the engine
Version 1 of Ava, completed August 15.
-
Measurement standard
Rules for sealing and scoring, fixed before any grading.
-
Sealed tests on historical data
Sealed before each outcome was revealed.
88%winning ad picked in 2,057 preregistered tests91.7%accuracy in an unfamiliar industry, no retraining98.3%same verdict on a blind re-run0invented findings when source material was thin -
Historical campaign re-run
Ava’s plan applied to a past campaign’s full record. Details on request.
-
Full system build and simulation
Every part connected and run in a simulated ad marketplace. No media spend.
-
Live campaign
Confirms bidding, tracking and reporting on a real platform.
-
Live validation against a control group
About two months, graded against a randomized control group.
-
First design partners
Back-tests on partners’ own records, graded by them.
Want to talk about a validation on your own data? Get in touch.