Quantitative technology decision-making for deep tech · An MIT spinout

Deep tech decisions are timing decisions.

TechNext measures how fast defined technology domains are improving and forecasts when an emerging technology will cross the cost or capability threshold that changes your decision. For the investors backing it and the R&D and mission teams that will have to adopt it.

Start with the technology area you are exploring, or the decision you need to make.

THE WALL STREET JOURNAL.
“New Research Busts Popular Myths About Innovation”
FastCompany
“MIT built a Google search to spot the most important tech innovations of the future”
FINANCIAL TIMES
“Innovation still requires smart, even barmy, innovators”
The evidence base

Measured, published, and patented.

150,000+
Technology domains
Each with a measured annual improvement rate.
97.2%
Of the U.S. patent system mapped
In the published validation study (Research Policy, 2021). Global patent coverage in the system today.
12,099,572 B2
Issued U.S. patent
The method is covered by an issued patent, not a description of one.
MIT
Founded by MIT researchers
Spun out by the MIT professor and researcher who built the method.
In the field
The Northeast Microelectronics Coalition uses the platform to analyse and accelerate innovation in semiconductor manufacturing across the Northeastern United States.
Who we work with

Questions we help you answer.

Corporate R&D and CVC

The roadmap has more promising options than the budget can fund, and the case for each one rests on a claim about how fast it will improve.

How we work with you →
Tech investors

There is no revenue to test the thesis against, so the whole bet rests on whether the underlying technology keeps improving on schedule.

How we work with you →
Government & defense

Programs are committed years ahead of the technologies they depend on, and the capability picture can shift mid-program.

How we work with you →
The hidden speed limit

The slope changes first. The market notices later.

Every board, investment committee and R&D team has more emerging technologies to judge than judgment can cover. Analyst reports, press attention, founder stories and raw patent counts tell you how much noise a technology is making. None of them tells you how fast it is improving.

TechNext uses patented technology developed at MIT to precisely define technology domains and estimate the annual improvement rate of each one.

The difference

Most technology intelligence measures attention. We measure progress.

STATUS QUOWHAT WE ADD
Broad trend labelsPrecisely defined domains
Raw counts and expert opinionNetwork position + performance evidence
Single-point maturity claimsTrajectory and crossover windows
A static recommendationDecision triggers that are monitored
Track record

We say no as readily as yes.

Three public calls in opposite directions, each logged before the outcome. Dozens of other areas and calls available on demand.

YES, SOONER THAN CONSENSUS
xMEMS micro-speakers

Analysts had the cost stuck well above competitiveness. We recorded a forecast of a cost crossover ahead of that consensus, and logged the later observation against it.

NO, LATER THAN THE HYPE
Nuclear fusion

Developers were publishing net-energy targets. Our estimated rates for tokamaks and HTS magnets put self-sustaining fusion later than any of them. Announced dates have since slipped toward our window.

YES, YEARS EARLY
Solar PV vs. onshore wind

From a year-2000 model, the method projected the trajectory of solar becoming cheaper than onshore wind per MWh, matching empirical reality roughly two decades ahead of time (Benson & Magee, 2018).

After the answer, we keep watching

A forecast with an expiry date on it.

Every answer says what evidence would change it, and we watch for that evidence. When a trajectory moves, you hear from us.

How the measurement works →

Recent projects

Recent work, taken domain by domain.

Recent engagements. Each asks which of the competing approaches inside it are improving fast enough to matter.

Quantum computing
Competing qubit architectures against each other.
Energy storage
Chemistries and cell formats on their own curves.
Microelectronics
Built alongside our work with the Northeast Microelectronics Coalition.
See how we work, and what a commissioned analysis covers →

We add a rate to the process your team already runs.

Thirty minutes with the people who build the model.

Book a meeting
TechNext
We measure how fast technologies improve and forecast when a challenger crosses the threshold that changes your decision.
Provenance U.S. Patent 12,099,572 B2 Research Policy (2021) An MIT spinout
© 2026 TechNext, Inc. We forecast capability and cost thresholds. Adoption dates are somebody else's claim.