Finding Value in Your Own Data: Pricing, Customers, and Assortment

In episode two, the hosts move from external data sales to decisions inside a company. Nikolay Golov selects three accessible cases from public vendor and consulting materials: pricing, customer profitability, and assortment. Andrey Tsybin and Alexander Polomodov ask which inputs are needed, whose work changes, and how to distinguish an appealing case description from a reproducible outcome.

At Motor City Industrial, branches had set prices independently; combining sales and pricing data supported a more consistent approach. Nikolay recommends a bounded first project, while Alexander emphasizes rules that keep working afterward. Much of the complexity can lie in reconciling sources, ensuring complete records, and getting people to follow a new process. The published case does not disclose every cost, so the estimates discussed are not a ready-made budget.

The logistics case connects customer income with support and transaction-handling costs. Excel may suffice for the arithmetic, but honest records and willingness to act are essential. The hosts discuss limits: network effects, customers using several products, and assumptions behind future income. Andrey also asks whether an observed improvement was actually caused by the intervention.

For assortment decisions, storage, delivery, and joint purchases complicate the calculation: a low-margin item may support the whole basket. The discussion turns to observing customers and an example from hotel search at trivago. The hosts distinguish correlation, explanations of behavior, and causal verification. They finish by advocating an understandable first step using available data, with analytical complexity proportionate to the decision and the cost of error.