# The commercial case (/platform/business)

## Where the margin goes [#where-the-margin-goes]

<Stats
  items={[
    ['#1', 'Fit is the most common reason for fashion returns'],
    ['2×', 'A return costs the outbound shipment again, plus handling'],
    ['3', 'Touches per return: inbound, inspection, restock'],
    ['0', 'Revenue from a return that ends as a refund'],
  ]}
/>

The exact numbers differ by category and market. The shape does not: an avoided return is worth more than a recovered one, and a return you understand is worth more than a return you merely processed.

## Pick the one that matches your bottleneck [#pick-the-one-that-matches-your-bottleneck]

<Cards>
  <Card icon={<Ruler />} title="If returns are size-driven, start with YourSizer" href="/yoursizer">
    Fewer wrong sizes leave the warehouse, and shoppers who were hesitating between two sizes buy one instead of two — or instead of none. **Measure:** product-page conversion on widget-engaged sessions, size return rate per SKU, share of orders with a recommendation.
  </Card>
  <Card icon={<RotateCcw />} title="If returns eat your support inbox, start with ResReturn" href="/resreturn">
    Self-service takes the “how do I return this” ticket out of the queue entirely, and store credit and exchange keep revenue a refund would lose. **Measure:** return tickets per 100 orders, time to resolution, share retained as exchange or credit, restock latency.
  </Card>
  <Card icon={<Camera />} title="If shooting is the bottleneck, start with NoShot" href="/noshot">
    Products go live without waiting for studio time, and a category page looks coherent because the model and light are constant across it. **Measure:** days from receiving stock to live product page, cost per product image, returns tagged “not as pictured”.
  </Card>
</Cards>

## Numbers we will not fake for you [#numbers-we-will-not-fake-for-you]

Every attribution number in our dashboards records the signal that proved it, and orders are deduplicated across signals. If a figure cannot be proven, it is reported separately rather than folded into the headline.

<Cards>
  <Card title="What we report">
    - Assisted orders, with the proving signal named
    - Recommendation distribution per SKU and per size
    - Return rate and return reasons, per product and per reason
    - Portal completion rate and outcome mix
  </Card>
  <Card title="What we tell you to check yourself">
    - Run an A/B split on your own store rather than trusting a benchmark
    - Compare return rate on widget-engaged versus non-engaged orders
    - Watch the size distribution before a production run, not after
  </Card>
</Cards>

## What a rollout looks like [#what-a-rollout-looks-like]

<Steps>
  <Step>
    **Connect the store.** App install or credentials. Catalogue and orders sync; nothing is written back yet.
  </Step>
  <Step>
    **Load your charts.** Size charts imported or ingested from images, then reviewed. This is the step that decides accuracy.
  </Step>
  <Step>
    **Pilot on a category.** One category, one placement. Verify the recommendation against products you know well.
  </Step>
  <Step>
    **Turn on measurement.** Attribution and analytics before scaling, so the rollout can be judged on your data rather than ours.
  </Step>
  <Step>
    **Scale, then add the second product.** Most brands add returns after sizing, or sizing after returns, within a quarter.
  </Step>
</Steps>

<Cards>
  <Card title="Talk to us" href="/platform/company" />
  <Card title="Check your platform" href="/platform/coverage" />
</Cards>