Skip to main content
Nick Ansel
Supply Chain SystemsField ResearchCompetitive StrategyGlobal ResearchEnterprise Transformation

The Hidden System of Work

A supply chain analyst counted boxes in her head for 10 hours a day. One extra zero sent five million t-shirts to sea.

Global field research revealed a shadow system of workarounds living outside enterprise software. That gap became a platform strategy and two product solutions that could catch a $10M mistake before it shipped.

Companye2open (via projekt202)
Period2017–2019
IndustrySupply Chain · Enterprise Software

Situation

e2open builds enterprise software for global supply chains. Billions of dollars of inventory decisions routed through it every day. But in the field, at the actual desks of the planners using it, something different was happening.

United States

Retail & logistics planners

Europe

Distribution & sourcing ops

China

Manufacturing & export ops

A representative sample of research sites — part of a broader multi-region engagement.

Patterns we observed

Across sites, a recurring theme emerged. The enterprise system existed. People were running operations alongside it — in spreadsheets — anyway.

How it was described

Enterprise software — deployed, technically available, the official answer to how planning got done.

How work actually moved

Custom spreadsheet grids planners had built themselves — more compatible with the actual pace and texture of the work.

Nobody described the spreadsheets as a workaround. They were just how the work got done.

One example that stuck

One extra zero

50,000500,000

At one site, a planner managed inventory across hundreds of SKUs using a custom spreadsheet. The workflow required mentally summing multiple cells and holding the result while keying an order quantity. Ten hours a day. On one order, an extra zero made it through.

Her manager's response was matter-of-fact: this is acceptable tolerances. At supply chain scale, this class of error gets priced in. The 60% off sale with no obvious reason — that is often the correction mechanism.

The point wasn't the specific order — it was what the observation revealed: a class of error that was entirely detectable before it shipped, and for which no detection mechanism existed in the platform.

A sample of what came from it

These are two examples of product directions the research informed — among a larger body of findings and opportunities.

Auto-sum grid function

Reducing reliance on mental arithmetic in table-based planning — removing one source of accumulated error from a high-volume, high-fatigue workflow.

Order volume anomaly flag

A 10x quantity spike is detectable before shipment. Surfacing it for a single confirmation step introduces just enough friction to catch a costly mistake before it becomes irreversible.

More Work