Technology Forecasting. Evidence Before the Decision.

Technology forecasting answers the question every R&D and corporate-development team eventually faces: not what a field looks like today, but where it is heading and how fast. We read that trajectory from hard signals — the S-curve position of a technology, the velocity of patent filings behind it, and the emerging sub-fields drawing new entrants — and turn them into a defensible three-to-five-year outlook you can plan and invest against.

What Technology Forecasting Actually Delivers

A technology forecasting engagement is not a trend piece or a guess dressed up with charts. It is a structured read of where a specific technology is on its development curve, how quickly it is moving, and which adjacent problems the next wave of R&D is about to attack. The output is a decision, not an essay: build, buy, licence, partner or wait — and by when.

Corporate strategy teams use technology forecasting to time market entry and avoid committing capital to a field that has already saturated. R&D leaders use it to decide which internal programmes deserve another funding round and which are chasing a plateau. Corporate-development and licensing teams use it to spot the assets worth acquiring while they are still undervalued.

What separates a useful forecast from an expensive opinion is the evidence base. We anchor every call in primary data — patent filing records, scientific-literature output, and standards and funding activity — so the outlook can be defended to a board, not just presented to one.

The S-Curve: Where a Technology Sits in Its Life Cycle

Every technology follows a broadly predictable path. In the emergence phase, filings and papers are sparse and experimental. In the growth phase, cumulative activity climbs steeply as the approach proves out and new entrants pile in. In maturity, the curve bends as the easy gains are exhausted, and in saturation it flattens as the field consolidates around a dominant design.

Plotted as cumulative patent activity over time, this trajectory traces the classic S-curve. Its value to technology forecasting is diagnostic: the same absolute filing count means opposite things depending on where the curve sits. Rapid growth on a young curve signals opportunity; the same growth flattening near the top signals a field about to consolidate and a poor moment to enter cold.

We fit the curve to real filing histories rather than eyeball it, then read the inflection point — the moment growth stops accelerating. That inflection is the single most actionable number a forecast produces, because it separates fields where a fast follower can still win from fields where only differentiated or defensive positions remain.

Filing Velocity: What Patent Momentum Reveals

S-curve position tells you the stage; filing velocity tells you the speed. By measuring how fast application volume is changing — not just its level — we separate fields that are genuinely accelerating from those coasting on past momentum.

The macro backdrop matters here. Global patent applications passed 3.5 million for the first time in 2023, up 2.7% and a fourth consecutive year of growth, while roughly 2 million patents were granted worldwide that year, a 10.1% jump and the fastest since 2012, according to WIPO’s World Intellectual Property Indicators. Against a rising tide, a technology whose filings are merely flat is quietly losing relative momentum.

So we always read velocity in relative terms: a field’s growth rate against the whole-portfolio baseline, and against its own three-year trailing average. Acceleration above baseline is the clearest early signal that a technology is entering its growth phase; deceleration below it, well before the absolute count peaks, is the earliest warning that a field is maturing.

Exploratory and Normative Forecasting Methods We Combine

No single method forecasts a technology well on its own, so we combine the two established families. Exploratory methods start from where a technology is and extrapolate forward; normative methods start from a future objective and work back to what has to be true for it to arrive.

The methods we draw on include:

• S-curve and trend extrapolation — fitting growth curves to filing and citation histories to project the next phase
• TFDEA (Technology Forecasting using Data Envelopment Analysis) — tracking the rate of performance improvement across successive product generations
• Bibliometric and patent citation analysis — mapping who cites whom to expose where a field is concentrating
• Delphi and expert elicitation — structured, multi-round input from domain specialists to calibrate the data
• Scenario planning and technology roadmapping — laying out plausible futures and the timing signals that distinguish them

Detecting Emerging Sub-Fields Before They Break Out

The highest-value forecasts identify a sub-field before its filing curve steepens — while the option to enter is still cheap. That means watching leading indicators rather than headline counts.

We track three signals. First, new-entrant velocity: when assignees who never filed in an area begin doing so in numbers, the field is drawing fresh capital. Second, citation bridging: patents that suddenly connect two previously separate technology clusters often mark the birth of a hybrid sub-field. Third, classification drift: a rising share of filings tagged with newer or finer patent classification codes signals that examiners themselves are recognising a distinct emerging area.

Artificial intelligence is the textbook case. EPO filings in core AI grew from 49 applications in 2015 to 2,243 in 2024 — a roughly twelve-fold rise, per the EPO Patent Index 2024. A team watching new-entrant and citation signals saw that curve steepen years before the absolute numbers made it obvious to everyone else.

What the 2024 Data Signals: AI, Energy and Biotech

A current read of the EPO Patent Index 2024 shows how differently mature and emerging fields behave. Overall European filings were essentially flat, down 0.1% after three years of strong growth — an aggregate that hides sharp divergence underneath.

Computer technology, which captures AI, machine learning and pattern recognition, became the single largest field for the first time at 16,815 applications, up 3.3%. Electrical machinery, apparatus and energy grew a striking 8.9% to 16,142 filings, reflecting the electrification and battery build-out. Biotechnology rose 5.4% to 8,479 filings, extending an uptrend that now spans at least eight years, while transport grew 3.5% to 10,026.

For technology forecasting, the lesson is that headline totals mislead. A flat aggregate masked double-digit acceleration in energy and steady, compounding growth in biotech — exactly the fields a forecast built on sub-field velocity, rather than the top-line number, would have flagged as the ones to watch.

From Signals to a Three-to-Five-Year Outlook Memo

Signals are only useful once they are reconciled into a single view. We synthesise the S-curve position, the velocity model and the emerging-sub-field scan into one outlook memo written for decision-makers, not analysts.

The memo states, in plain terms, which phase the technology is in, how much runway the growth phase has left, which two or three sub-fields are most likely to break out next, and what would have to change for that view to be wrong. Each call carries the specific timing signal behind it — the velocity threshold, the citation bridge, the new entrant — so the reader can watch the same indicator and know when the forecast is being confirmed or broken.

Because it is anchored in cited primary data and an explicit method, the memo survives scrutiny in a strategy review. That is the point: a technology forecasting output should make the timing decision easier to defend, not merely easier to illustrate.

How to Brief a Technology Forecasting Engagement

The biggest determinant of a useful forecast is a sharp brief. Name the specific technology and the decision it feeds — “should we license or build in solid-state battery separators, and by when” beats “tell us about batteries.” A decision-anchored question keeps the analysis pointed at an answer you can act on.

Tell us the time horizon that matters (most planning cycles run three to five years), the geographies you compete in, and any incumbents or assets already on your radar. The tighter the frame, the deeper we can go on the sub-fields that will actually move your decision, rather than spreading thin across a whole domain.

From there we run the S-curve fit, the velocity model and the emerging-field scan, and return the outlook memo with its underlying evidence. If the question spans sourcing or valuation, we fold in the adjacent work so the forecast connects directly to the next move.

What You Receive

• An S-curve maturity assessment placing your technology on its life-cycle curve
• A filing-velocity trend model showing whether momentum is accelerating or fading
• Emerging sub-field detection flagging the niches drawing new entrants first
• A three-to-five-year outlook memo with the timing signal behind each call
• An IP overlay showing who is building the position you would forecast against

Data Sources & References

This analysis draws on primary patent and market data:

• WIPO World Intellectual Property Indicators 2024 — Global patent filing and grant totals and growth rates used for the velocity baseline.
• EPO Patent Index 2024 — European filing counts by technology field, including AI, energy and biotechnology.
• Daim et al., Forecasting emerging technologies (Technological Forecasting & Social Change) — Peer-reviewed method combining bibliometrics and patent analysis with forecasting tools.
• Technology forecasting framework using patent analysis (MDPI Sustainability) — Applied patent-analysis roadmap methodology for technology forecasting.

Request a Technology Forecast

Tell us the technology and the decision it feeds, and we will map where it is heading.

Related PerspireIP work: Technology Scouting · Market Sizing & Opportunity Analysis · Technology Scouting in Biotechnology (case study).

Frequently Asked Questions