Commando. Military. Cop. battle strategies.

 In a battle, to conquer enemy territory, you first send in the commandos to distract the watch tower. Land your ship on the beach front and the military barges in. The military pushes aggressively and captures enemy land. Then the cops come in to restore peace and enforce rule of law.

A successful startup (or even a feature) goes through those 3 phases. See Paul Graham’s blog on this.

Commando

• < 5% odds of success. A 1% shot at blowing up Adolf Hitler is an opportunity of a lifetime.

• Team mates die all the time. No big deal. (Employee attrition)

• Civilian casualties are OK. (Customers). If a stadium needs to be blown up to take down hitler, too frigging bad. Greater good.

• Step changes in territory. 0 to 1.

• Largely un-coordinated. Speed is everything. Disrespect for rules.

Military

• 20-70% odds of success. Can’t indulge in suicide missions and risk too many lives. But need to take chances and risks.

• Death is quite common, but still 10% or less.

• Civilian casualties are avoided but not uncommon. (Customer bad experience)

• High compound growth rates in territory. Or revenue.

• High rates of co-ordination. Large teams. But still small autonomous subteams, loosely coordinating.

Cop

• 99% odds of success. If the robber has a gun, call in a 100 backups.

• No chances of death.

• No civilian casualties.

• Stable territory. Or revenue.

• Extreme coordination and process.

Any good decision making should consider the cost of failure and pick the right strategy for the job. And products need to be rolled out to increasing number of customers as maturity rate increases and the products solidify. You should know where in the product life-cycle the feature you are building stands.

Cops and commandos find it hard to get along. Click here to see how it goes.