Rebuild-to-Sustain Advantage
Continuously experiment, kill failures, and scale proven advantages
- Difficulty
- Expert
- Time to result
- ~ongoing to results
- Steps
- 6
- Confidence
- 96%
Rebuild-to-sustain advantage starts from the premise that success attracts competition and makes yesterday's edge less valuable. Rather than defending the current operating model until it fails, leaders continually test new capabilities, products, data sources, and processes. Each initiative is bounded by explicit hypotheses and success or shutdown criteria. Failed initiatives are killed so their resources can be recycled; successful ones receive increasing capital, talent, and organizational support. Over time, the portfolio of experiments changes the firm's composition before external pressure makes transformation unavoidable. The mechanism is disciplined renewal rather than random novelty: persistent experimentation generates options, evidence determines which options survive, and scaling converts validated options into the next version of the organization.
Origin
Matthew Granade relayed Steve Cohen's explanation that sustained hedge-fund performance required rebuilding the firm repeatedly, on Masters in Business.
Core principles
- 01Current success attracts competitors and erodes its own advantage.
- 02Assume today's winning business will not remain unchanged.
- 03Run multiple bounded initiatives before the need becomes urgent.
- 04Kill failed experiments rather than defending sunk costs.
- 05Scale initiatives only after evidence of success.
- 06Treat organizational restlessness as a survival capability.
How to run it
- 1
Assume advantage will decay
Identify where competitors, technology, regulation, or customer behavior could weaken the current business.
Pro tip Examine the sources of today's strongest returns first.
Watch out Do not mistake recent performance for durable protection.
- 2
Create renewal hypotheses
Form specific hypotheses about new capabilities or approaches that could create the next advantage.
Pro tip Tie each hypothesis to an observed change in the environment.
Watch out Novelty without a mechanism is not strategy.
- 3
Run bounded experiments
Test several hypotheses at manageable scale with explicit owners, budgets, and evaluation periods.
Pro tip Design experiments so failure is affordable and informative.
Watch out Do not overload the core business with too many simultaneous initiatives.
- 4
Apply kill criteria
End experiments that fail predefined evidence thresholds and document what was learned.
Pro tip Set shutdown conditions before teams become emotionally invested.
Watch out Sunk-cost reasoning can turn experiments into permanent distractions.
- 5
Scale demonstrated winners
Increase resources and integrate initiatives whose evidence supports a durable advantage.
Pro tip Scale in stages while verifying that performance survives increased scope.
Watch out A small pilot may rely on conditions that disappear at scale.
- 6
Rebuild continuously
Repeat the cycle so the organization evolves before its existing advantage is exhausted.
Pro tip Maintain a visible portfolio of current, killed, and scaling initiatives.
Watch out Renewal becomes chaotic if experiments are not connected to strategic priorities.
In the wild
A hedge fund continually experiments with alternative data, systematic strategies, and hybrid human-machine portfolios. Weak initiatives are closed, while successful ones gain capital and staff until they become meaningful parts of the firm.
→ The organization develops new sources of advantage as older methods become crowded.
A successful workflow vendor tests AI-assisted features, usage-based pricing, and a new enterprise channel while its core product is still healthy. It kills the channel experiment, refines pricing, and scales the AI workflow after retention improves.
→ The business renews its product advantage before competitors force a defensive response.
Common mistakes
Waiting for a crisis
Transformation becomes more expensive and less flexible once the current business is already deteriorating.
Protecting every experiment
Refusing to kill failed initiatives drains resources and conceals which ideas genuinely work.
Scaling before proof
Large commitments made before validation turn manageable experiments into strategic liabilities.
Is it for you?
Best for
It is best for organizations in fast-moving, highly competitive industries where advantages decay quickly.
Not ideal for
It is not ideal for teams without enough capacity to protect core operations while running disciplined experiments.
From the transcript
“because I've rebuilt it four or five times UM.”
“if you think that whatever success you have today is going to be true tomorrow, you are really naive UM.”
“things will succeed or you know, and things will fail and will kill them, and things that will succeed will scale”
From the episode
Matthew Granade Discusses Quantitative Investment (Podcast)
Matthew Granade