No More Elephant Carpaccio!

How to eat an ele­phant? Exact­ly: In small slices. With this sala­mi tac­tics you can mas­ter big tasks. Many rec­og­nize the iter­a­tive-incre­men­tal nature of agile meth­ods like Scrum and there­fore feel remind­ed of this ele­phant carpac­cio — a fal­la­cy which is based on the fun­da­men­tal con­fu­sion of com­pli­cat­ed­ness and complexity.

Sala­mi tac­tics has its jus­ti­fi­ca­tion for han­dling step-by-step tasks that are large and com­pli­cat­ed, but oth­er­wise well under­stood. Build­ing an air­plane or a car falls into this cat­e­go­ry. Such a prob­lem can be dis­man­tled and ana­lyzed. And the desired result can then be com­posed of the com­po­nents. This is cer­tain­ly the case today with many years of expe­ri­ence in air­craft and auto­mo­bile construction.

The art of sim­plic­i­ty is a puz­zle of complexity.

Dou­glas Horton

In the days of the Wright broth­ers or Got­tlieb Daim­ler, how­ev­er, these tasks were com­plete­ly dif­fer­ent. At that time those were com­plex prob­lems which were solved not by ratio­nal analy­sis, but through learn­ing by tri­al and error. Asked by a reporter how it feels to fail a thou­sand times, Thomas Edi­son replied: “I did­n’t fail 1,000 times. The light bulb was an inven­tion with 1,000 steps.”

Agili­ty pri­mar­i­ly trans­lates into flex­i­bil­i­ty. This adapt­abil­i­ty actu­al­ly results from a series of small steps. Ide­al­ly, each of those steps is a small exper­i­ment with the aim of learn­ing a lit­tle more about the chal­lenge and the pos­si­ble solu­tions. Like the sequence at the bot­tom of Hen­rik Kniberg’s draw­ing above. Each step results in a usable prod­uct that allows for expe­ri­ence to be gained. This is com­plete­ly dif­fer­ent from ele­phant carpac­cio, where it is only at the end that one notices whether the ele­phant is viable.

Suc­cess is not final, fail­ure is not fatal: it is the courage to con­tin­ue that counts.

Win­ston Chrurchill

The com­mon objec­tion that it is not effi­cient to build a skate­board, a scoot­er, a bicy­cle and a motor­cy­cle before the con­vert­ible is as jus­ti­fied as it is instruc­tive. In fact, this pro­ce­dure is not effi­cient if it is clear that the cus­tomer needs and wants the con­vert­ible and it is clear how con­vert­ibles are basi­cal­ly built. Then this still com­pli­cat­ed task can be ratio­nal­ly and effi­cient­ly solved by means of decomposition.

If all you have is a ham­mer, every­thing looks like a nail.

Abra­ham Maslow

How­ev­er, it is often not clear which prod­uct or ser­vice will be need­ed tomor­row and what a pos­si­ble busi­ness mod­el looks like. Or it is uncer­tain what the best tech­ni­cal solu­tion is — like for Edi­son, Daim­ler or the Wright broth­ers. Or all togeth­er. A great deal of unnec­es­sary suf­fer­ing aris­es from the fact that the entire arse­nal for effi­cient solu­tions to com­pli­cat­ed tasks is applied to a prob­lem that is actu­al­ly com­plex. So: please no more ele­phant carpac­cio when you real­ly need agility.