> ## Documentation Index
> Fetch the complete documentation index at: https://syncly-a76215af.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.

# Glossary and metric interpretation rules

> How metrics are calculated, when they're recorded, and how to read them.

Always read numbers in context: which Queries, date range, filters, and platforms they cover. Current totals and numbers recorded at a point in time answer different questions.

## Posts and mentions

* **Post:** A single piece of content published on a social platform.
* **Mention:** Where a brand, product, or topic you track comes up in a post.

One post can match more than one Query if it fits several of their criteria. When you add up numbers across Queries, keep in mind that the same post may be counted more than once.

## Views and impressions

* **Views:** The view count each platform reports.
* **Impressions:** Syncly's standard measure of exposure, which evens out the different ways platforms report visibility. It shows how much a post was seen, not how many unique people saw it.
  * **Facebook and Instagram posts:** Likes plus comments.
  * **Reddit:** Upvotes plus comments.
  * **Other platforms:** Views.

> [!NOTE]
>
> If a platform doesn't provide a metric, it may show as 0. Keep each platform's data limits in mind when comparing.

## Engagements and engagement rate

* **Engagements:** A post's saves, comments, likes, and shares added together.
* **Engagement rate (%):** Engagements ÷ Views × 100. It's 0% when a post has no views.

Instagram feed posts don't show engagement rate on the post card or in post detail. You can still see it in table views and XLSX exports.

## Sentiment

Syncly's AI reads a post's text and context and classifies it as positive, neutral, or negative.

* **Per Query:** Sentiment is set **for each Query**. The same post can be positive for one brand and neutral for another.
* **Reasons and edits:** You can see why a post was classified a certain way. Editors and Admins can change it, or switch it back to the AI result. Changes apply only to that Query. See [post detail](https://syncly-a76215af.mintlify.site/en/analysis/post-details.md).
* **Sentiment score (Share of voice):** For each Query, (positive posts − negative posts) ÷ posts with a sentiment × 100. It ranges from -100 to +100, and is 0 when there are no posts with a sentiment.

With only a few posts, one or two can move the score a lot, so check the original posts before making big decisions.

## Promotion type

**Promotion type** shows whether Syncly's AI thinks a post is **Likely sponsored** (it found signs of a paid partnership, such as #ad, discount codes, or brand tags) or **Likely organic** (no signs of sponsorship). It was previously called AI Sponsorship. You can see why each post was classified a certain way, and edit it, in [post detail](https://syncly-a76215af.mintlify.site/en/analysis/post-details.md).

## Latest and Day metrics

* **Latest:** A post's most up-to-date totals.
* **Day 1, 2, 3, and 7:** What a post had 24 hours, 48 hours, 72 hours, and 7 days after publishing. Choose a Day in Query explorer's **Posts** tab to compare early performance.

> [!NOTE]
>
> **Blank vs. 0**
>
> * **Blank:** That point hasn't been recorded yet, or the post hasn't reached it.
> * **0:** It was recorded, and the value really was zero.

## Share of voice and change units

* **Share of voice:** Each brand's or product's share of a metric within the Queries you compare. It shows standing within your chosen group, not the whole market.
* **Change units:**
  * Share changes are shown in **percentage points (pp)**.
  * Sentiment score changes are shown in **points (pt)**.

See [Share of voice](https://syncly-a76215af.mintlify.site/en/insights/sov-analysis.md) for more.

## Collection period, noise, and limits

* **Collection period:** Syncly collects posts that match your criteria within the period you choose, starting from when matching posts actually exist.
* **Noise:** Posts that don't fit a Query are marked as noise and left out of that Query's results. Noise is set per Query, so a post can be noise for one Query and count for another. See examples in the **Noise posts** tab in Query detail.
* **Daily collection limit:** Check limits and usage under **Query usage** in workspace settings. You'll get an alert if a Query reaches 90% of its daily limit.

## XLSX Export and dashboard PDF

> [!NOTE]
>
> * **XLSX Export:** Built from the latest data that matches your filters at the time the file is created.
> * **Dashboard PDF:** Exports a dashboard's charts, notes, and post lists as a PDF to share.

## MCP

* **MCP (Model Context Protocol):** Connects AI assistants like Claude to your Syncly data using the official server address and your Syncly sign-in. What it can access follows your role in each workspace (Viewer, Editor, or Admin).

Next, [run your first analysis with a prepared Query](https://syncly-a76215af.mintlify.site/en/quick-start/getting-started.md).


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