When I teach key account managers on Cranfield’s KAM Best Practice programme, I start with a simple two-by-two grid. It has two lines, four boxes and one warning that I repeat all through the course. Don’t be very good at implementing a bad strategy.
Four ways to run a business
The grid has worked for decades, and in the age of AI agents it’s more relevant than ever. What’s new is that AI agents can make any business far better at implementing whatever it has chosen to do, right or wrong. That changes which box you’d want to be in, and how fast you get there.
Where does the grid come from?
I learned it from Malcolm McDonald, emeritus professor at Cranfield School of Management and author of Marketing Plans, which has sold over half a million copies. He has spent a career on marketing strategy, and this grid is one of the simplest tools he uses.
It rests on a distinction often credited to Peter Drucker, from The Effective Executive. Efficiency is doing things right. Effectiveness is doing the right things. Tactics are about the first. Strategy is about the second. Put one on each axis and every business lands in one of four boxes.
What do the four boxes mean?
Thrive is the right strategy, done well. The business has chosen the right customers and the right offer, and it delivers them efficiently.
Survive is the right strategy, done badly. It’s clumsy and wasteful, but it’s heading the right way. Customers forgive a lot when the offer is right.
Die slowly is the wrong strategy, done badly. The business is heading the wrong way, but its inefficiency slows it down. There’s time to notice and turn.
Die quickly is the wrong strategy, done well. This is the box that surprises people. A business that is very good at the wrong thing gets to the wrong place faster, and spends more getting there.
Kodak is the example I use. It was superb at making and selling film. One of its own engineers, Steve Sasson, built an early digital camera in 1975, and Kodak’s leaders doubted anyone would want to look at their pictures on a television. The film business stayed brilliantly efficient for years after the market had started to move, and Kodak filed for bankruptcy in 2012.
What does AI change?
Speed. AI agents can now research, draft, analyse and follow up at a scale no team could manage by hand. On my slide I put it as “AI agents can 100x both of these”. Both columns get faster: the business doing the right thing and the business doing the wrong thing.
That has an uncomfortable consequence. AI moves a business to the right on the grid, and then multiplies whatever direction it was already heading. Moving up the grid is still a decision people have to make.
AI speeds up the direction you're already heading
The thriving business more than doubles. The one in the “die quickly” box loses most of its value in the same five years. AI made both of them more efficient. Only one of them was pointed the right way.
I made a similar point about The Goal. Speeding up a step that isn’t the bottleneck adds nothing to the money coming in. Speeding up a business that’s heading the wrong way does worse. It takes away the time it had to notice.
If everyone gets the same tools, what’s left?
Michael Porter answered this thirty years ago in What Is Strategy?, his 1996 Harvard Business Review article. He separated operational effectiveness, which means performing similar activities better than rivals, from strategy, which means performing different activities, or the same ones in different ways.
His warning was that operational improvements get copied. When every company adopts the same best practices, everyone gets better and nobody gets ahead. He called competition on operational effectiveness alone “mutually destructive”, and wrote that “the essence of strategy is choosing what not to do”.
AI is the fastest spread of best practice there has ever been. Every competitor can use the same models, often within weeks of release.
Everyone moves right. Nobody moves up.
Once everyone is efficient, efficiency stops being an advantage. The only thing that still separates competitors is the vertical axis: what each one has chosen to do.
Why does domain knowledge matter more now?
My slide said that domain knowledge has never been so important, and this is why. AI is very good at doing things right. It’s much less able to tell you what the right thing is for your customers, in your market, at this moment. That still takes people who know the customers, the market and the history.
In key account management, the change is easy to see. An AI tool can now draft a plan for every account in an afternoon. The hard question is which accounts deserve a plan at all, and what the business should stop doing for the rest. That’s a strategy question, and I’d want my most experienced people answering it.
It’s also the first step in the delegation loop I wrote about this week. Choosing the right job comes before briefing it, letting it run and checking what comes back. A perfectly delegated wrong job is still a wrong job.
What would I ask an LSP owner?
Most people who read this blog run language service providers, and I think the grid is a useful way to look at the industry right now.
AI has made translation dramatically more efficient. An LSP can now produce far more words, far faster, for less. But if the strategy is to sell those words by the word as the price per word keeps falling, then that LSP is getting very good at a business the market is moving away from. On the grid, that’s a fast slide to the right along the bottom row. I wrote about the pressure on the industry in Nobody asked the language industry.
So these are the questions I’d ask. What would your customers pay for that they can’t get from a machine? Which services would you stop offering if you were starting today? And is your AI investment making you better at your strategy, or only faster at your old one?
Where does that leave us?
The grid is still the most useful one I know, and AI has raised the stakes. Being efficient used to be an advantage. Now it’s something every competitor can buy. What can’t be bought is the decision about where to point all that speed.
I’d rather be in “survive” with the right strategy than in “die quickly” with the best tools on the market. With the right strategy, AI will fix the efficiency. With the wrong one, it will only get you there faster.
Sources
- McDonald, M. (2014) How to develop a winning strategy to boost sales and profits. Presentation at Cranfield School of Management, 4 February.
- York St John University Professor Malcolm McDonald, staff profile
- Porter, M. E. (1996) What Is Strategy? Harvard Business Review, November to December. Summary at MAAW
- Drucker, P. F. (1967) The Effective Executive
- PetaPixel (2017) What Kodak said about digital photography in 1975
- Wikipedia Kodak and Steven Sasson