The future is a claim. We quantify it.
Every forecast an institution publishes is a claim about the future with a date attached. We have parsed 4.14 million of them from 630,000 sources and scored where they converge. Your board asks why you believe a market moves in 2030. You answer with evidence.
Nextatlas reads people. Foresight reads institutions.
300,000 innovators across more than 40 verticals. The unit is the entity: matcha, lion's mane, quiet luxury. It tells you what people are already doing, and it holds up about three years out.
630,000 institutional sources. The unit is the claim: a dated, directional statement about what changes. This is the layer that works from three to ten years out, before there is any behavior to observe.
Both layers use the same method and differ only in what they count. Run them together and the coverage runs from next quarter to next decade.
Anatomy of a claim.
A claim says how something changes, and by when. We pull every one out of its source and break it into the same fields. A forecast buried in a 90-page energy report and a single line in a patent filing end up in the same table.
How a claim earns a probability.
A single forecast tells you what one analyst believes. What we score is agreement. When institutional reports, patent filings, and innovator behavior point the same direction inside the same window, that convergence is measurable, and it is the basis of every probability on this page.
Reports, news, academic papers, government publications. Where institutions put the future in writing. The most explicit of the three sources, and the slowest.
120 million filings, read at publication. A patent is the first public trace of a private bet, filed years before a product exists, and it shows you where R&D money is already committed.
We track 300,000 early adopters continuously. They are the control on the other two sources. If a projected shift has nobody living it yet, you find out here.
Possible, plausible, probable.
Each band does a different job. You commit budget against what is probable, hold options against what is plausible, and keep watch on what is merely possible so it cannot arrive unannounced.
Every bet gets a date.
The model groups claims and patents into opportunity areas and scores each one for readiness. Every area carries a horizon. That is what lets a team sequence its bets and defend the order it chose.
Illustrative. Two inflection points define the commercial window.
Market ready. The problem is named and patents are already clustering on it.
In development. Positioning narrows from generic to condition-specific and categories begin to compete.
Speculative. The category splits, and each branch is a separate bet you can start positioning against now.
It fits the way your team already works.
Our foresight scientists assemble the source pool and the claim taxonomy for your sector, then build the scenario set with your team, starting from the questions you need answered this year.
New claims arrive continuously and the probabilities move with them. The version you show the board in March has already changed by the time you present it in June.
A REST API and an MCP connector expose the claims, the scenarios, and the probabilities to your own tools and agents. Your analysts query it from where they already work.
Add Foresight to the Nextatlas trend layer on one subscription. Behavioral signal and institutional projection under one contract, through the same API.
Let's quantify your next decade.
Before your board asks how you know.