AI and Augmentation
Roger Martin's question about AI is not whether to adopt it, but how. His distinction is simple and load-bearing: substitution replaces people to cut cost, augmentation extends their judgement to build advantage. Most companies do the first and call it strategy.
Martin pictures a ladder. On the bottom rung, AI is faster clerical labour. On the top rung, it is a thinking partner that sharpens human judgement. The strategic act is to climb it on purpose, not to bank the savings on the first rung and stop.
Substitution-first is a trap for one plain reason: every competitor can buy the same cost cut at the same time. That is Operational Excellence Is Not Strategy wearing an AI badge. Under Winner-Takes-Most Dynamics, parity tools hand nobody an edge. Advantage lives where rivals cannot or will not follow, the Can't-Won't Test: configurations they cannot easily assemble, proprietary data that compounds, and capabilities that people and models build together.
For a leader, then, this is a cascade question, not a tooling one. Where does AI change Where to Play, opening problems you could not serve before? Where does it change How to Win, through intelligence specific to your assets that compounds over time? And which Enabling Management Systems make augmented work the daily norm rather than a stalled pilot?
The seam with our OKR system is the position we hold everywhere else: AI is the accelerant, not the point. A company that only harvests substitution savings has built the AI version of an all-operational OKR portfolio, better numbers, no theory of advantage.
One line to keep: AI is the accelerant, not the point; substitution-first adoption is an all-operational OKR portfolio in disguise.
Our synthesis of Roger Martin’s published work, sources credited. Read the originals: they’re excellent.
Why is substitution-first AI adoption a trap? +
Because cost reduction is available to every competitor at once, the AI version of operational excellence that is not strategy. Under winner-takes-most dynamics, parity tools produce no advantage. Advantage comes from configurations, proprietary data and redesigned workflows rivals can't or won't copy.
How does this connect to OKRs? +
Same position we hold throughout: AI is the accelerant, not the point. A portfolio that only harvests substitution savings is an all-operational OKR set in disguise, better numbers, no theory of advantage. Climb the augmentation ladder deliberately instead.
Reading about method is not the same as running it. We install this system and build the capability that stays.
Turn strategy into executed strategy →