# List an experiment's results

`GET /api/v1/experiments/{experimentId}/results`

The stored verdict: one row per arm per slice. IT SERVES THE LAST COMPUTED SNAPSHOT AND RECOMPUTES NOTHING, so computed_at is the whole measure of how current the answer is and the merchant's dashboard is what refreshes it. Recomputing would be a write, and this family holds no write grant. DECIDE ON always_valid_p_value, not on p_value. The second is only honest at a pre-declared sample size, and a client polling this endpoint is monitoring continuously by definition. All three statistics are null on the control arm, which carries no test result by construction; reading a null as zero calls a winner on the baseline. CHECK srm_ok FIRST. False means the arms did not receive the traffic they were allocated, which invalidates every other number in the row. It is computed within the slice, and per-slice SRM is noisy on thin cells, so treat a false on a country row as a reason to distrust that row rather than the test. revenue_minor, profit_minor, ppv_minor and aov_minor are MINOR UNITS WITH NO CURRENCY: the rollup does not group by currency, so on a multi-currency store they sum more than one. That is valid for COMPARING arms, which are randomly assigned and therefore carry the same currency mix, and it is not a money value. Do not render either as a price. An unknown experiment id is a 404 rather than an empty page, so a typo can never be read as "this test produced nothing". An experiment that has never been analysed answers an empty page, which is a real state.

## Path parameters

- `experimentId` (string, required)

## Query parameters

- `segment` (string) — One slice of the verdict. overall is the whole experiment, device:mobile, device:desktop and device:tablet are the device classes, and country:XX is an ISO alpha-2 country. Omit it for every row, in which case the overall rows sort first. Anything outside this vocabulary is a 400 rather than an empty page, so a typo is never read as "this slice has no data".
- `limit` (integer) — Rows per page. Out of range is a 400 rather than a silent clamp, so a client asking for more than 100 learns it did not get it.
- `after` (string) — The next_cursor from the previous page. Opaque: decode nothing from it and construct nothing by hand, since its encoding is not part of this contract. Omit it to read the first page.

## Responses

- `200` — Success
- `400` — The request was refused before any state changed. `code` is one of:

`invalid_body`, a write body this route will not take. `reason` partitions it
and `field` names the key when one key is at fault.
`invalid_query`, a query parameter, including limit and after.
`invalid_text`, a NUL byte or bytes that are not valid UTF-8 anywhere in the
path, the query or the body. Strip control characters before sending.
`idempotency_key_required`, a write sent without the Idempotency-Key header.
`invalid_request`, an Idempotency-Key longer than 255 bytes.

Routes add their own codes for rules only they know. Switch with a default arm.
- `401` — No credential, or one this API does not accept. `code` is always
`unauthorized`.

THE BODY IS DELIBERATELY UNINFORMATIVE. An expired key, a revoked key, a
publishable key, a key belonging to another merchant and a key that never
existed are all refused with the same bytes, so this response cannot be used to
probe which keys exist. Check the key's state in the dashboard rather than
inferring it here.

Send the key as `Authorization: Bearer <key>` or as `X-API-Key: <key>`. It is
never accepted in a query string.
- `403` — The key lacks the read_experiments scope. `code` is `insufficient_scope` and the message names the scope to ask the merchant for.
- `404` — The resource this url addresses does not exist, or never did on this store.

`resource` names the kind of thing that was missing and is the half to branch
on: the code is `not_found` on newer routes and a per-entity spelling on
routes that shipped before that rule, and the two mean the same thing.

A 404 is ALWAYS about the url. An id inside the request body that resolves to
nothing is 422 `unprocessable_reference` instead, so the two cases stay
distinguishable. An id that belongs to a different merchant reads exactly like
one that never existed.
- `429` — Too many requests. `code` is `rate_limited`.

Two limits apply independently: one on the credential and the route family,
one on the client address. The headers describe whichever has less left, so
honouring Retry-After always clears the window that bound.
- `default` — Any status this operation does not list, in the same envelope.

A 5xx means the request may or may not have applied. Retry it with the SAME
Idempotency-Key: that is the only way to find out without risking a duplicate,
and it is what the key is for.

A few 4xx conditions arrive here rather than as a listed status because they
depend on the merchant's plan or on a module being wired: 402 when a quota or a
plan limit is reached, and 503 when a capability the route needs is not
configured on this deployment. Both carry a `code` naming which.

## Example

```bash
curl --request GET \
  --url 'https://api.mercemur.com/api/v1/experiments/{experimentId}/results' \
  --header 'Authorization: Bearer <token>'
```