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

# Reading a scan

> How to read the per-second timeline, lenses, moments, and raw output of a completed scan.

A completed scan is a **per-second timeline**, not a single number. Here's how to read each part. For the exact field-by-field contract, see [Get a scan](/api-reference/get-scan).

## The headline score

The top of the response carries an `overall_score` (`0–100`) and a letter `grade`, the single number to lead with when you need one. Treat it as a **directional average** over the whole clip, not a precise measurement: it's the same headline the dashboard and result email show. The real value is still in the timeline below; the score just tells you, at a glance, whether the piece landed.

## The timeline

`timeline` is the heart of the result, with one entry per second of the clip. Each second carries the lens `scores` for that moment, and (depending on the [options](/api-reference/create-scan#param-options) you asked for) the `raw` brain-network values, a `benchmark`, and per-second `reasons`. Read it as a curve: the *shape* across seconds tells you far more than any one second alone.

## Lenses

Each second's `scores` holds the four reads (attention, memory, purchase\_intent, manipulation), each `0–100`, where **50 is neutral**. A high attention score with a low memory score, for example, means the content grabs focus but doesn't stick, a signal to strengthen the payoff.

## Moments

`moments` names the standout seconds, where a lens **peaks** or **drops** sharply, so you don't have to scan a whole clip by hand. Each moment tells you the lens, the second, and how far it moved. This is usually where you start.

## Summary

`summary` is a plain-English paragraph: what happened, where the highs and lows were, and which read led. It's the fastest way to brief a human (or hand an agent a headline).

## Raw

`raw` is for teams that want to go underneath the lenses: the per-second activation of each of the seven [brain networks](/knowledge-base/the-7-networks). Every lens score is derived from these. Include it with `options.include_raw`.

When `include_raw` is set, each second also carries `raw_labeled`: the **same numbers** re-keyed to plain-language names (`Voice`, `Cognitive`, `Emotional`, `Memory`, `Attention`, `Salience`, `Visual`) so you don't need the neuroscience terms. `raw` is left untouched; `raw_labeled` is purely additive.

<Note>
  On short or highly-repetitive content the timeline can come back nearly flat, with every second scoring about the same. That's a faithful read, not a bug: a steady piece produces a steady brain response. The signal you act on is the **movement** across the timeline, so flatness itself is information. It usually means nothing in the content is creating contrast.
</Note>

<Card title="Create a scan" href="/api-reference/create-scan">
  Submit content and get a scan back.
</Card>