← Systems

/

Submission Workflow: Three Iterations

Submission Workflow:

Three Iterations

Workflow

Submission Workflow: Three Iterations

The honest version of this story includes the failure in the middle. Here’s the annual-window model that created contributor pressure, next to the rolling quarterly pipeline that finally balanced our predictability with contributor flexibility.

The honest version of this story includes the failure in the middle. Here’s the annual-window model that created contributor pressure, next to the rolling quarterly pipeline that finally balanced our predictability with contributor flexibility.

Type

Workflow

Format

Before/after flow comparison

From

Submission redesign · three iterations over six years

01 · Context

Version one was unstructured — artists could submit anything, anytime, with the company covering all costs; demand was unpredictable and quality inconsistent. Version two swung too far: a single annual window created predictability for us and pressure for contributors — miss the window, fall behind a year. Version three, diagrammed here, is a rolling quarterly model that balanced operational predictability with contributor flexibility. The mechanics were never the hard part; adoption was. Rolling a change across 150–200 contributors took time every single time. What kept the workflow changing was demand-side pressure — clients needing fresher offerings, curation filling gaps, and a standing feedback loop with the artists themselves.

Version one was unstructured — artists could submit anything, anytime, with the company covering all costs; demand was unpredictable and quality inconsistent. Version two swung too far: a single annual window created predictability for us and pressure for contributors — miss the window, fall behind a year. Version three, diagrammed here, is a rolling quarterly model that balanced operational predictability with contributor flexibility. The mechanics were never the hard part; adoption was. Rolling a change across 150–200 contributors took time every single time. What kept the workflow changing was demand-side pressure — clients needing fresher offerings, curation filling gaps, and a standing feedback loop with the artists themselves.

02 · The artifact

V1 · Unstructured

Unstructured — submit anytime, we cover cost

V2 · Failed

Annual window — predictable for us, pressure for them

V3 · Held

Rolling quarterly — balanced both sides

Before · V2

Before · V2

Annual window, manual review

Annual window, manual review

Annual window, manual review

Single annual submission window — miss it, wait a year

Single annual submission window — miss it, wait a year

↓

Email submissions → manually logged to a spreadsheet

Email submissions → manually logged to a spreadsheet

↓

Serial review meetings; decisions bottleneck on scheduling

Serial review meetings; decisions bottleneck

on scheduling

↓

Outcomes communicated ad hoc; contributors chase status

Outcomes communicated ad hoc; contributors chase status

Friction

Volume spikes overwhelmed capacity · contributors submitted before ready · infrequent submitters compounded behind — the structure created the imbalance.

After · V3

Rolling quarterly pipeline

Rolling quarterly pipeline

Rolling intake: up to 10 pieces per quarter, submit anytime

Rolling intake: up to 10 pieces per quarter,

submit anytime

↓

Structured form → auto-triggered review task sequence

Structured form → auto-triggered review task sequence

↓

Async review on a cycle; heads see a filtered set, advisors give market-framed feedback

Async review on a cycle; heads see a filtered set, advisors give market-framed feedback

↓

Decisions + targeted feedback on a predictable cadence

Decisions + targeted feedback on a predictable cadence

Fixed

Dec–Jan closed to match vendor schedules and capacity — the calendar tells the truth. Review timelines cut 50%+ while quality held.

↳ recreated & sanitized · stage names generalized · no client data shown

03 · Outcome

50%+

50%+

reduction in review and decision timelines

3

3

full iterations before the structure held — each driven by feedback and data

200+

200+

contributors across 36 states operating on one pipeline

The middle iteration failed because it optimized for our predictability at the contributors’ expense. The version that worked distributed both the flexibility and the constraint. When a workflow keeps breaking, the problem is usually who’s absorbing the variance — not the tooling. And the redesign is only ever half the job; the other half is adoption, which is slow by nature and never fully finished.

The middle iteration failed because it optimized for our predictability at the contributors’ expense. The version that worked distributed both the flexibility and the constraint. When a workflow keeps breaking, the problem is usually who’s absorbing the variance — not the tooling. And the redesign is only ever half the job; the other half is adoption, which is slow by nature and never fully finished.