Ecosystem Building Game Practice

Ecosystem building is the original assembly task of the McKinsey Solve assessment. In 2026 it is rarely reported: standard invitations open with Redrock (a data study) and then Sea Wolf, and some add Sustainable Futures Lab — see the Solve guide. It is still a good drill for the habit the newer tasks reward: filtering options against several constraints at once.

In the task you are given a location described by a few conditions, a set of species with their own requirements, and you have to build a food chain in which every species can actually live. In this trainer each round gives you a terrain, a pool of twelve species and five slots to fill, from the producer at the base upward.

Three rules have to hold at once. A species must tolerate the terrain — both its depth and its temperature range have to contain the location's values. It must eat the species directly below it in the chain, not merely something somewhere in it. And the species below has to supply at least as many calories as the one above requires. Every slot that satisfies all three counts, so a chain that holds for three links scores those three.

What actually helps: filter before you build — cross out every species the terrain rules out, and you are usually choosing between two or three candidates per slot rather than twelve. Work upward from a producer, because the calorie constraint only resolves once you know what is below. Check the predator's requirement against its prey's supply as you place it, not at the end, since one underfed link invalidates everything above it.