Good Strategy Bad Strategy Examples
Good Strategy/Bad Strategy Examples applies Rumelt's kernel test to real collapses like Kodak and Mxit, so the difference between a real diagnosis and wishful thinking stops being abstract.
Follow a real company through the same six steps, watching where its diagnosis held and where it quietly slipped into wishful thinking.
Reach for this when…
- The kernel test feels abstract until someone can point to a company that got it wrong.
- A team needs a case they didn't live through, so ego doesn't block the lesson.
- You want to show what "mistaking goals for strategy" actually looks like in practice.
How to run it
- Pick a real case, for example Kodak or Mxit.
- Identify the diagnosis the company actually made, or failed to make.
- Identify the guiding policy that followed, or didn't.
- Check whether the actions taken were coherent with each other.
- Name which bad-strategy hallmark the case shows most clearly.
- Apply the same test to the participant's own organisation.
A worked example
Situation. Sofia Aguirre runs a craft brewery in Cordoba, Argentina, expanding from its taproom into national retail, and her board kept citing "we need to grow" as if that were a strategy.
Applied. She walked the board through Mxit - the South African chat app with 10 million users before WhatsApp arrived, which never named the real threat until too late - then asked: what's our Mxit?
Result. The board named its own blind spot: assuming tap-room loyalty would transfer to supermarket shelves with zero brand recognition. The national push was redesigned around building shelf recognition first.
The catch
Case studies teach pattern recognition, not application - a team can nod along to Kodak's mistake and still make the same one, because their own diagnosis is harder to see than someone else's in hindsight. Picking cases isn't neutral either: the wrong one (too extreme, too distant) lets people dismiss the lesson as "that couldn't happen to us".
If the case makes people say 'we're different', you've picked the wrong case or stopped one question too early.
Origin: Richard Rumelt (theory); Dale Williams (curated examples)