Stay ahead of the curve
Get exclusive insights delivered directly to your inbox. Join thousands of tech leaders.
TL;DR
- A technology roadmap sets which technologies a company will develop or adopt, and when. Timing is where most roadmaps fail.
- Early and late are not equally bad. Early can be slowed down; late usually has to be bought.
- Western Digital could have entered flash storage for $65M in 2009. It paid $19B for SanDisk in 2015. The rate had flagged flash in 2001.
- Replace dates with triggers tied to a measured improvement rate.
Timing is the question R&D leaders actually ask
R&D leaders rarely struggle to name the likely winner. They struggle with when to move, and what it will cost to move then. Commit too early and you pay for an immature technology. Commit too late and you spend a year developing a product for a market that has already moved on.
A technology roadmap that only lists dates cannot answer that.
Early and late are not equally bad
| Moving early | Moving late | |
|---|---|---|
| Main risk | Paying for an immature option | The window has closed |
| Can you recover? | Usually: slow down, scale back, keep a small team | Rarely: the capability takes years to build or must be bought |
| What it costs | Some wasted effort | Market position, acquisition premiums, people |
| What protects you | Staged funding and stop rules | A standing watch on the rate, with triggers |
Late usually means buying your way in, at a price set by everyone who already saw what you missed.
What waiting cost Western Digital
Backtest · Data storage · 2001–2026
The late entry cost $19 billion
The rate flagged flash in 2001. Western Digital's way in cost $65M in 2009 and $19B in 2015.
Improvement rate per year, computed from patents public at each date · GetFocus backtest: HDD vs SSD
The GetFocus storage backtest shows the curve. In 2001 solid-state storage was improving at 49.6% a year against hard drives' 29.2%. The gap widened every year after.
- 2005: Apple discontinued the hard-drive iPod Mini at peak sales for the flash-based Nano.
- 2009: Western Digital bought SiliconSystems, a small enterprise-flash maker, for a reported $65M. Eight years after the signal, the entry price was still small.
- 2010: SSD filings overtook HDD filings, nine years after the rate.
- 2015: Western Digital agreed to buy SanDisk for about $19 billion.
The technology choice was never in doubt. The late entry cost $19 billion: full price for a future the rate had flagged fourteen years earlier.
Why change still surprises prepared teams
Exponential improvement looks flat, then sudden. A technology compounding at 50% a year doubles in under two years; one at 30% takes more than two and a half. For a decade the gap is small in absolute terms, then it is everything. Teams that watch performance see the change when it is already obvious. Teams that watch the rate see it when the curves first separate.
Replace dates with triggers
Instead of "adopt in 2029", a resilient roadmap says "move when X happens." Triggers that work:
- A challenger's rate overtakes the incumbent's on the job you care about. Sodium-ion crossed LFP in 2019 (71.0% vs 70.7%) and has pulled away since (see the radar).
- The rate gap passes a threshold you set, for example 1.5× the incumbent.
- New entrants appear on a fast-improving route.
- A competitor commits capacity, not just research.
- Your next platform window approaches. Time the evidence for the decision point, not the calendar.
GetFocus recomputes every route's Technology Improvement Rate monthly and fires an alert the moment a trigger you set is crossed.
How to change course without losing credibility
Leaders worry that reversing a recommendation after twelve months will cost them credibility with the business. The fix is to publish the trigger with the recommendation. Then a change of course is the plan working, not a U-turn.
FAQ
What is a technology roadmap?
A plan that links business goals to the technologies needed to reach them, with an order and timing for developing, buying or adopting each one.
How do you create a technology roadmap?
Start from the business goals and the functions you need. List every option for each function. Measure how fast each option is improving. Then set staged investments and explicit triggers for when to move, scale or stop.
How do you decide when to invest in a new technology?
Invest when the challenger's improvement rate is clearly above the incumbent's and the gap will close within your planning horizon, and before your next platform decision. When building the capability takes years, late is the bigger risk.
How often should a technology roadmap be updated?
Most companies revisit it yearly. The triggers should be watched continuously, so a change between reviews is caught in time.
