Roger Martin on going on the offensive with creative strategy
I recently came across an interesting article by Roger Martin, “Going on the Offensive with Creative Strategy,” which questions a foundational principle of modern business. His argument is a direct challenge to the way many of us, myself included, have been taught to approach new ideas.
The core argument
Martin posits that the business world is dominated by “analytics”—executives who believe that the future can be accurately predicted by rigorously analysing data from the past. On the other side are the “creatives”—the innovators and designers trying to build something new.
The problem arises when creatives are forced to justify their new ideas using the analytics’ framework. They are asked to prove future success with historical data, which is an impossible task for something that has never existed before. As Martin notes, this is a game creatives are destined to lose.
Creatives have a strong tendency to attempt to defend their ideas while implicitly or explicitly accepting the dominant premise.
The flaw in the analytical premise
Martin argues that this reliance on historical data is fundamentally flawed, a point made centuries ago by thinkers like Aristotle and Charles Sanders Peirce. Aristotle pointed out that data analysis is only useful if we are certain the future will be identical to the past—a rare occurrence in business. Peirce went further, stating that no truly new idea has ever been proven analytically in advance.
When analytics ask creatives to prove a new idea in advance analytically, they are asking the creatives to do something that has never been done in the history of the world — good luck with that!
This creates an environment where forecasting becomes a comforting illusion rather than a useful tool for genuine innovation.
It is the great pacifier of modern business — revenue forecasting makes analytics feel good, even if it provides no more value than popping a pacifier in a baby’s mouth.
A critical reflection
Martin’s argument made me reflect on my own work. I see a large reliance on trying to predict business outcomes based on past data, which is exactly what he criticises. It is important, however, to distinguish this from other types of prediction. For instance, in predictive maintenance, we are not forecasting unpredictable human behaviour in a market. Instead, we are predicting machine failure based on sensor data and physical principles. We can say, “We are 95% confident this machine will fail within the next 30 days.” This is a calculated probability within a more constrained system, allowing for informed operational decisions. It is fundamentally different from trying to forecast revenue for a product that does not exist yet.
Martin’s core point about business strategy remains valid and important. Truly great business ideas are rarely born from maintaining the status quo. Sticking to the proven path may yield marginal gains with some confidence, but it will not lead to breakthrough innovations. Look at a tool like the Cursor IDE; it was not created by marginally improving existing code editors but by fundamentally rethinking the developer workflow with AI at its core.
Why an offensive stance is better
To escape this trap, Martin urges creatives to stop defending their ideas on the analytics’ terms and instead go on the offensive. This means questioning the assumptions behind the status quo.
For Innovation in Modern Business, the only Defense is Offense.
Instead of trying to prove your new idea will work, challenge the assumption that the current model will continue to work indefinitely. This shifts the conversation from defending the unknown to comparing two possible futures: one based on a new idea and one based on the increasingly fragile assumption that nothing will change.
This “offensive” mindset is more powerful in general because it is proactive. It is about taking control of the narrative, setting the terms of the discussion, and actively shaping the future rather than passively reacting to the past. It is a shift from defence to creation.
