> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pocketsflow.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Benchmarks

> Your conversion rate, average order, refund rate and email open rate next to anonymised stores of a similar size, with one next step for each weak spot. Pro plan and above.

**Benchmarks** (Pro plan and above) compare your store with stores like it.
Open **Analytics → Benchmarks**, or look at the Benchmarks card on your
dashboard.

## What is compared

All four use the last 90 days of live sales.

| Metric | Definition |
| - | - |
| **Conversion rate** | Paid orders ÷ unique visitors to your store, product and checkout pages. Needs 100 visitors. |
| **Average order** | What a paid order is worth on average, order bumps included. Subscription renewals are left out. Needs 5 orders. |
| **Refund rate** | Refunded orders ÷ paid orders. Lower is better. Needs 20 orders. |
| **Email open rate** | People who opened your newsletter emails ÷ emails sent. Needs 200 emails sent. |

For each metric you see your value, the **median** of similar stores and the
**middle half** of them (25th to 75th percentile), and where you stand: top
25%, above the median, below it, or bottom 25%. Every metric below the median
comes with one concrete next step and a link to the Pocketsflow feature that
helps (abandoned checkout emails, order bumps and upsells, clearer product
pages, or resending to people who did not open).

## Who you are compared with

Stores are grouped by **revenue over the last 90 days** (under $1k, $1k to
$10k, $10k to $50k, $50k to $250k, $250k and up) and, when there are enough
of them, by **kind of store**: digital products, courses or subscriptions
(whatever brought in most of your revenue). Average orders are compared in
US dollars and shown in your account currency at the day's rate.

## Privacy

No store ever sees another store's numbers:

* a group is only shown when it has **at least 10 stores**, and each metric
  only when at least 10 of them have it — otherwise you see "Not enough
  similar stores yet";
* only the 25th percentile, median and 75th percentile are ever computed and
  stored — never a store's own value, a minimum or a maximum;
* store counts are rounded down to a multiple of 5;
* test-mode sales, banned stores, stores confirmed as fraud and our own test
  stores are left out of every group.

Benchmarks are recomputed every night.

## From the API and AI tools

`GET /benchmarks` with an API key, or the MCP tool `get_benchmarks`. See the
[endpoint reference](/api-reference/endpoints#benchmarks). Below the Pro plan
it answers `403 FEATURE_LOCKED`.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.