Understanding Factory Scorecards: What Actually Matters

Why most brands evaluate factories wrong

Most apparel brands have some form of factory evaluation process. They visit facilities. They review certifications. They check audit reports. They ask for references from other brands. And then, after the season ends, they fill out a supplier performance review — on-time delivery, quality pass rate, communication responsiveness — and file it somewhere that makes it unlikely to be consulted when the next sourcing decision is made.

This is not negligence. It is the predictable outcome of a process built for a different operating environment. When a brand works with the same fifteen factories for twenty years, institutional knowledge about supplier performance lives in the heads of the sourcing team. The account manager knows which factory to call for a complicated knit. The production director knows which supplier cannot be trusted with a tight timeline. This knowledge is real and valuable — and it is completely invisible to anyone who hasn't worked at the brand for a decade.

The problem surfaces when teams turn over, categories expand, or brands need to source from factories outside their established network. The knowledge walks out the door with the people who held it, and the new team starts from scratch — making sourcing decisions with less information than their predecessors had.

The metrics that actually predict performance

Not all factory metrics are equally predictive. Experience working across hundreds of production runs points to a consistent hierarchy of indicators — the ones that correlate most reliably with production outcomes.

First-sample accuracy is the single strongest leading indicator of overall production performance. A factory that consistently returns accurate first samples has a production process that is well-managed from the beginning of a run, not just at the end. Brands that track first-sample accuracy by factory — rather than by season or style — accumulate a predictive signal that most sourcing teams don't have access to.

Communication responsiveness, measured specifically as time-to-substantive-response on technical queries, predicts how well a factory will manage exceptions during production. A factory that answers questions quickly and accurately during the pre-production phase is a factory whose team is engaged and capable. The opposite is an early warning sign.

On-time delivery rate matters — but only when tracked at the style level rather than the order level. A factory with a 90% on-time delivery rate at the order level can have individual styles running dramatically late, masked by an otherwise reliable operation. Style-level delivery tracking surfaces the categories or construction types where a factory's performance degrades.

Defect rate at first quality inspection, tracked separately from post-correction defect rate, reveals whether a factory catches its own errors or ships them. These are very different operational characteristics, and they predict very different outcomes when things go wrong during a production run.

Building a living scorecard

A factory scorecard that is completed once per season is a historical document. A factory scorecard that is updated continuously as production unfolds is a management tool.

The distinction matters because factory performance is not static. A supplier that performed exceptionally for three seasons can have a difficult fourth season — a change in management, a production capacity constraint, a quality control process that degraded during rapid expansion. A brand relying on historical scorecard data will not see this change until it has already cost them a production run. A brand with a live scorecard will see it in the metrics before the first sample comes back.

Building a living scorecard requires two things that most brands currently lack: a structured data model for factory performance that is consistent across seasons and categories, and a process for updating that data continuously rather than periodically. The first is a design problem. The second is an operational discipline problem — and it is the harder one to solve, because the data entry has to happen as part of the production workflow rather than as a separate administrative task.

What Calico tracks automatically

Calico's Produce product solves the operational discipline problem by making factory scorecard data a byproduct of the production process rather than a separate activity.

Every factory interaction that flows through Calico — every communication thread, every sample review, every delivery confirmation, every quality inspection — contributes to the factory's live performance record. First-sample accuracy is captured when samples are reviewed and logged. Communication responsiveness is measured against the timestamp of the original query. Delivery performance is tracked at the style level, not the order level. Quality outcomes are recorded with enough specificity to identify whether issues cluster around particular construction types or material categories.

The result is a factory intelligence layer that grows more accurate and more useful over time — not because anyone is filling out forms, but because the work is being done inside a system that captures performance data as a natural byproduct. Sourcing teams gain access to the kind of institutional knowledge about factory performance that used to live only in the heads of long-tenured employees — and they gain it whether the person who built it is still at the company or not.