Cor­po­ra­te way of thin­king: We care­ful­ly pre­pa­re a plan. Then we imple­ment it preci­se­ly. The results will be as sta­ted in the plan. Fai­lu­re is due to bad execu­tion.

Star­tup way of thin­king: My plan is not real­ly a plan, it’s just a sum of my hypot­he­sis. For sure it will be wrong but I don’t know whe­re and how. I need to run lots of struc­tu­red expe­ri­ments to test my hypot­he­sis. When I have vali­da­ted my assump­tions, I may have a plan that is worth somet­hing. Fai­lu­re is due to not run­ning enough dif­fe­rent “tests” with cus­to­mers.

Many star­tups beha­ve like cor­po­ra­tes in this regard, assu­ming their plan will work – just throw in money and people and voi­la. No. Be very awa­re that until you have vali­da­ted your assump­tions you do not have a plan. Then you can not be imple­men­ting a plan (read: sca­ling), all you can do is run expe­ri­ments.

The result of an expe­ri­ment is to inc­rea­se your know­led­ge – does this work, Yes/No. The job of a star­tup is to run expe­ri­ments. A good star­tup learns a lot whi­le spen­ding very litt­le effort and money. A bad star­tup spends a lot and doesn’t learn much.Cost efficiency of your lear­ning should be a key objec­ti­ve whi­le in the expe­ri­men­tal pha­se (=all pha­ses befo­re sca­ling, by which time you should have a vali­da­ted plan you can just imple­ment). Cost efficiency of run­ning your ope­ra­tion should be your key objec­ti­ve once at the sca­ling pha­se.