1.
Test before you scale. McDonald's risked 80 stores, not 8,000 — learned what they needed, then pivoted without damaging the core.
2.
New competitors don't always mean new categories. Sometimes a 'threat' is just a product gap you can fill with line extensions.
3.
Know when to kill experiments fast. Two years of data beat ten years of indecision — move on when you have your answer.