What Happens When Your Competitors Catch Up?
Your AI initiative works. It creates measurable economic value. The company successfully captures part of that value. ROI is positive.
Time to celebrate?
Maybe.
But there is one more question: Can you keep the advantage?
This is the third stage of the Create → Capture → Defend framework. And it may become one of the hardest problems in enterprise AI.
Positive ROI is not the finish line
Consider a simple example.
An AI initiative generates $2 million in economic value this year. Next year it generates another $2 million. The internal dashboard looks healthy.
Value created: positive.
ROI: positive.
Trend: stable.
But something else has changed. Competitors have adopted similar technology. Their costs are falling faster. Their output is increasing faster. One has lowered prices. Another has improved service. A third is taking market share.
Your AI initiative is still producing value. Your strategic position is deteriorating.
That gives us an important distinction:
Absolute AI ROI ≠ competitive AI advantage
Internal performance tells you whether the initiative works for you. Competitive performance tells you whether it works well enough.
Only relative gains count
This is one of the strongest arguments in Harvard Business Review's AI and the Looming Competition for Margin.
The authors argue that companies cannot evaluate AI efficiency gains in isolation. If competitors achieve greater improvements, your own gain may do little more than keep you in the game.
That changes the question from: Are we improving? to: Are we improving faster than the competitive baseline?
That is a much more demanding standard. And it may become increasingly important if AI capabilities continue to spread.
What happens when everyone has access?
AI is unusual because many organizations can access similar foundation models, cloud infrastructure, developer tools, and software.
The HBR authors argue that this broad availability can make application-level advantages difficult to defend.
That does not mean every company will use AI equally well. Far from it.
But it does mean that: The model itself may not be the moat.
If everyone can buy increasingly capable intelligence, differentiation has to move somewhere else. Possibly toward:
- proprietary data
- better workflow design
- distribution
- customer relationships
- brand
- switching costs
- organizational learning
- superior economics
- faster execution
- better decisions
The technology may become more common. The organizational capability around it may become more valuable.
AI may compress the lifespan of advantage
This is still a hypothesis worth testing, not a settled fact. But it raises an important strategic question:
If successful AI applications can be copied more quickly than previous technological advantages, how much time will companies have to respond?
The HBR article argues that widespread technology diffusion can accelerate imitation and intensify competition.
If that dynamic becomes strong in AI, companies may face a continuous cycle:
The question is no longer merely whether the original AI investment worked.
It becomes: How quickly can the organization recognize that its advantage is fading and decide what to do next?
Defending does not mean freezing
“Defend” can sound like protecting an existing system forever. That is not what it means here.
Sometimes the right defense is to:
- scale faster
- redesign the workflow
- improve proprietary data
- lower cost
- switch models
- route tasks to cheaper systems
- change pricing
- reinvest in customer experience
- abandon an initiative and move capital elsewhere
The objective is not to defend every AI project. It is to defend the company's ability to create and capture value.
Sometimes that requires killing yesterday's successful idea.
The management system becomes part of the moat
That leads to what may be the most important implication of the framework.
As AI capability becomes easier to access, sustainable differentiation may depend increasingly on the organization's ability to: decide, adapt, allocate, redesign, measure, and execute faster than competitors.
Think about the feedback loop:
If competitive conditions are changing quickly, slow management cycles become expensive.
An annual portfolio review may discover deterioration long after the economics changed. Even a positive ROI metric can become misleading if it is viewed without trajectory or competitive context.
The question is not only: Is this initiative working?
It is: Is it still the best use of our capital now?
The AI race may be an allocation race
This is where Create → Capture → Defend changes how we think about AI strategy.
First: Create — Did AI produce real economic value?
Then: Capture — Did the company retain enough of that value?
Finally: Defend — Can it continue doing so as competitors respond?
The organizations that win may not simply be the ones that deploy the most AI. They may be the ones that can see more quickly:
- what is working
- what is deteriorating
- where advantage is moving
- and where money and management attention should go next
That brings us to the next question in this research program:
If AI speeds up competition, can your management system keep up?