Growth Strategy · Foresight · Signal Reading
What H3 Growth Requires That H1 and H2 Tools Cannot Give You
McKinsey's Three Horizons framework is the standard way to balance today's business against tomorrow's. The part it doesn't tell you is how to find Horizon 3 opportunities in the first place — because by definition, they don't show up in the data your H1 and H2 tools were built to read.
The Framework This Builds On
McKinsey partners Mehrdad Baghai, Stephen Coley, and David White introduced the Three Horizons model in their 1999 book The Alchemy of Growth. The framework splits a company's growth agenda into three parallel streams: Horizon 1 is the core business generating most of today's profit and requires defending and extending it. Horizon 2 is emerging businesses with product-market fit in sight, needing scaling and selective investment. Horizon 3 is early-stage options — ideas and technologies still being validated, five to ten years from mattering.
The framework's own guidance recommends roughly a 70-20-10 split of resources and attention across the three horizons. It remains one of the most widely taught growth frameworks in business strategy, still referenced directly on McKinsey's own site as an "enduring idea."
What the Three Horizons model doesn't provide is a method for actually finding H3 candidates. H1 strategy is well served by competitive intelligence and financial analysis — the data exists because the market already does. H2 can still lean on emerging market data and visible competitor moves. H3, by construction, can't use either of those, because the opportunity hasn't shown up in conventional business data yet. That's what makes it H3.
The tools built for H1 and H2 are tuned to detect what's already happening. H3, by definition, is what hasn't happened yet.
A Real Field Built Around This Exact Problem
Technology forecasting — using patent filings and academic publication data to predict where a technology is heading before it's commercially visible — is an established academic discipline, not a Vaxa invention. Researchers including Daim and colleagues have published peer-reviewed work in the journal Technological Forecasting and Social Change specifically on using bibliometric and patent analysis to forecast emerging technologies, building on decades of patent-citation research going back to the 1980s and 1990s.
The practical logic is straightforward: patents are typically filed 18 to 36 months before a related product reaches market, which makes patent filing activity one of the earliest available signals of where a technical field is actually heading — earlier than product announcements, earlier than most competitive intelligence, and earlier than analyst coverage.
Early-stage venture capital is the second concrete, data-backed signal — not because VCs are always right, but because pre-seed and seed-stage capital represents the earliest point where real money is being staked on a specific technical bet, ahead of any product or revenue data existing at all.
PitchBook's Emerging Tech Indicator, which tracks pre-seed, seed, and early-stage deals specifically in emerging digital technologies, recorded $33.1 billion deployed across 874 such deals over the most recent year tracked — a record dollar figure even as deal count stayed below the five-year average, reflecting fewer, larger, higher-conviction early bets rather than broad speculative spread.
The pattern worth watching isn't the aggregate number — it's concentration. When early-stage capital clusters unusually heavily into a narrow technical category (the same PitchBook data shows this happening recently in AI-native biotech and AI-powered cybersecurity specifically), that concentration is a signal in itself: it means a specific group of specialist early-stage investors, whose job is exactly this kind of judgment, have independently converged on the same bet.
Why This Signal Gets Missed
Analyst coverage and competitive intelligence both describe what's already visible — an opportunity someone else has already found and moved on. Patent and publication activity describes what's being built before it's a product at all, which is exactly the lead time an H3 search actually needs.
Proof This Works at Scale: Procter & Gamble is the clearest large-scale proof that this kind of structured outside-signal scanning isn't hypothetical. In 2000, incoming CEO A.G. Lafley concluded P&G's "invent it ourselves" model couldn't sustain growth — the company had roughly 9,300 researchers, but nearly 2 million comparably capable scientists and engineers were working outside it. The resulting Connect + Develop program, launched as a permanent operating structure rather than a one-time initiative, is credited with over 35% of P&G's marketed products by 2006, a share that later grew past 50% — documented directly in Huston and Sakkab's account in Harvard Business Review, March 2006.
Cross-Referencing, Not Relying on Either Alone
Neither signal is reliable by itself. Patent activity can reflect defensive filing as easily as genuine innovation. Early VC activity can reflect a fad as easily as a real inflection point. The real value is in the intersection: a technical field showing both rising patent/publication activity and rising early-stage capital concentration, at the same time, is a materially stronger signal than either one alone — because it means both the people building the technology and the people willing to bet money on it independently arrived at the same place.
01
Patent & Publication Activity
Patents are typically filed 18 to 36 months before a related product reaches market — one of the earliest available signals of where a technical field is heading, earlier than product announcements or analyst coverage.
02
Early-Stage Venture Capital
Pre-seed and seed-stage capital represents the earliest point where real money is staked on a specific technical bet, ahead of any product or revenue data existing at all.
03
Read the Intersection, Not Either Alone
A technical field showing both rising patent and publication activity and rising early-stage capital concentration at the same time is a materially stronger signal than either one alone — it means the people building the technology and the people willing to bet money on it independently arrived at the same place.
04
Make It a Recurring Read, Not an Annual One
Patent activity and early-capital concentration both shift meaningfully within 12-to-18-month windows. An H3 read done once a year during an annual offsite is already stale for most of the months it's supposed to inform.
The Actual Discipline This Requires
The hard part isn't accessing the data — patent databases and VC deal trackers are commercially available. The hard part is treating H3 signal-reading as a disciplined, recurring practice rather than something done once during an annual strategy offsite. A technology's patent activity and early-capital concentration both shift meaningfully within 12-to-18-month windows, which means an H3 read done once a year is already stale for most of the year it's supposed to inform.
This is exactly the gap Vaxa's Future Foresight practice exists to close — signals converted into decisions before the market sees them, run as a standing discipline rather than an annual exercise. The patent and publication bibliometrics described above is the kind of work that sits under Technology Horizons, Vaxa's deep analytical research capability on emerging technologies. Cross-referencing that signal against domain-expert judgment — the step that turns a data pattern into a confident call — is what Insights Panels, Vaxa's proprietary primary research model built on direct access to domain experts, is built to do.
Data You Can Buy, a Call You Can't
Patent filings, academic publication trends, and VC deal data are all commercially available to anyone willing to pull them — none of it is proprietary. What's harder to buy is knowing how much weight to give each signal when they only partially agree, which happens far more often than the clean case where every source lines up. Strategic Innovation — Vaxa's proprietary methodology, built and refined across real engagement work — is what turns a partial, cross-referenced read into a timed decision, rather than a general sense that something is happening in a space.
The Actual Question
Which of your organization's growth bets are genuinely H3 — five to ten years out, still being validated — and are they being read through patent and publication signals cross-referenced against real capital flow, or through the same competitive-intelligence tools built for H1 and H2 that H3, by definition, can't show up in?
Vaxa's Growth Strategy in Action.
Related Material
Baghai, M., Coley, S. & White, D., The Alchemy of Growth: Practical Insights for Building the Enduring Enterprise, Perseus Publishing, 1999 — origin of the Three Horizons of Growth framework. · McKinsey & Company, "Enduring Ideas: The Three Horizons of Growth," mckinsey.com.
Daim, T. et al., "Forecasting Emerging Technologies: Use of Bibliometrics and Patent Analysis," Technological Forecasting and Social Change, 2006. · PatSnap, "Competitive Patent Monitoring for Tech Forecasting" — on typical patent-to-product lead times.
PitchBook, Emerging Tech Indicator (ETI) — early-stage venture capital deal tracking in emerging digital technologies. · Huston, L. & Sakkab, N., "Connect and Develop: Inside Procter & Gamble's New Model for Innovation," Harvard Business Review, March 2006.