Fixed-Percentage R&D Budget Rule
Set R&D as a fixed percent of revenue so no one has to fight for it every year.
- Difficulty
- Easy
- Time to result
- ~months to results
- Steps
- 5
- Confidence
- 80%
Rather than negotiating the R&D budget annually — a process prone to short-termism, especially during downturns — Morris Chang set TSMC's R&D spending at a fixed 8% of revenue, regardless of the business cycle. The rule removed the yearly tug-of-war between the CEO and R&D leadership, gave R&D confidence to plan multi-year technology bets, and directly enabled the aggressive push into the 28nm 'sweet spot' that became a turning point for the company. Chang had tried to raise TI's R&D percentage as an executive there and was repeatedly denied; at TSMC, as CEO and chairman, he simply set the number himself. The rule works because it converts a recurring negotiation into a standing commitment, insulating long-cycle technology investment from short-term financial pressure.
Origin
Chang tried unsuccessfully to raise TI's R&D percentage as an executive there in the 1970s; when he became TSMC CEO he set his own fixed rule to avoid repeating that annual fight.
Core principles
- 01Recurring budget negotiations create uncertainty that undermines long-range planning
- 02A fixed percentage-of-revenue commitment removes the yearly argument
- 03Protecting R&D through downturns preserves long-term technology leadership
- 04Simplicity (a single number) beats complex annual justification processes
How to run it
- 1
Review historical spend range
Look at the last several years of R&D as a % of revenue to find a realistic baseline (TSMC was running 6-7%).
- 2
Set a single fixed target
Pick a target percentage slightly above the baseline to signal ambition (TSMC chose 8%).
Pro tip Round to a simple, memorable number so it's clearly a policy, not a negotiation.
- 3
Apply the rule through good times and bad
Hold the percentage steady regardless of recessions so R&D isn't cut in downturns.
- 4
Free R&D leadership from annual budget defense
Let R&D leadership focus on technology bets instead of budget justification each year.
- 5
Let the freed-up confidence generate bigger bets
Watch for R&D to propose more ambitious multi-year targets (like TSMC's 28nm and 7nm 'sweet spot' calls) once budget certainty exists.
In the wild
After Chang fixed R&D at 8% of revenue in 2009, the R&D team, freed from annual budget fights, began pushing the idea that 28nm would be a 'sweet spot' node — eventually convincing Chang to triple capital spending.
→ TSMC's leading-edge dominance traces directly back to the confidence created by the fixed R&D commitment.
Common mistakes
Letting R&D be re-litigated every year
Annual negotiation between CEO and R&D director (TSMC's prior practice, and TI's practice where Chang's requests were repeatedly denied) creates instability and short-term thinking.
Cutting R&D during downturns
Treating R&D like a discretionary expense that flexes down in a recession undermines the multi-year technology bets needed to stay ahead.
Is it for you?
Best for
Technology and R&D-intensive companies wanting to protect long-cycle innovation investment from short-term pressure.
Not ideal for
Early-stage companies with volatile or unpredictable revenue where a fixed % may be unaffordable.
From the transcript
“I want to make him at ease you know I mean doesn't have to argue doesn't have to to request every year so I almost…”
“8% regardless of uh whether there's a recession or not and just 8% off revenue”
From the episode
TSMC Founder Morris Chang
Morris Chang