> For the complete documentation index, see [llms.txt](https://ailyze.gitbook.io/ailyze-docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://ailyze.gitbook.io/ailyze-docs/common-research-workflows/analyze-social-media-discussions.md).
# Analyze Social Media Discussions
This page shows you how to collect what people say on social platforms — comments, posts, and threads — and turn it into [themes](/ailyze-docs/reference/glossary.md#theme), quotes, and a summary you can share.
This workflow has two parts: first you **collect** the conversation, then you **analyze** it.
{% hint style="info" %}
**What you need:** the Free Plan or higher, plus the public URLs you want to collect from (a video, a post, or a search or hashtag link).
**What you will have at the end:** a spreadsheet of comments, plus themes with [frequencies](/ailyze-docs/reference/glossary.md#frequency) and quotes.
{% endhint %}
## Step 1: Collect the conversation
Use **"Get Social Media Web Data"** to pull comments and posts. Evidano can collect from many popular platforms — including YouTube, Facebook, Instagram, TikTok, X/Twitter, LinkedIn, Reddit, and Threads, among dozens of others.
1. Open **"Get Social Media Web Data"** (the page is titled **"Extract Data from URLs"**).
2. Use **"Auto-detect Platform"** (recommended), or pick a specific source (for example, *YouTube Comments*, *Reddit Comments*, or an *X/Twitter* search).
3. Add your URLs (a video, a post, or a hashtag or search link), or upload a list.
4. Run the extraction, then click **"Download ZIP File"**.
{% hint style="info" %}
**Heads-up:** Collection settings like date ranges go *inside* the URL (for example, an X/Twitter search with `since:` and `until:`), not in a separate form. See [Collect social media and web data](/ailyze-docs/collect-research-data/collect-social-media-and-web-data.md).
{% endhint %}
## Step 2: Analyze the discussion
The ZIP holds a spreadsheet per source (comments, authors, dates, likes, and so on). Unzip it, then:
1. Open **"Analyze xlsx/csv files"** and upload the spreadsheet. This is exactly the handoff the **"Select Analysis"** prompt offers after your download.
2. At **"Choose the type of Analysis"**, pick **"Analyze text responses in selected columns"** and select the comment or text column.
3. If your export includes useful columns (likes, date, author type), tag them under **"Segments"** so you can compare — for example, themes in highly-liked vs low-engagement comments.
4. Click **"Create Project and Start Analysis"**.
## What your results show
* **Themes** in the discussion with [frequencies](/ailyze-docs/reference/glossary.md#frequency) — what people actually talk about, ranked.
* **Quotes** (real comments) to illustrate each theme.
* Optional [**segment**](/ailyze-docs/reference/glossary.md#segment) **comparisons** (for example, by engagement level or time period).
* **A coded table** of every comment for your own analysis.
See [How to read your results](/ailyze-docs/analyze-data/how-to-read-your-results.md) for a tour of every panel.
## Turn it into a deliverable
* Click **"Chat to generate report or chart"** and ask: *"Summarize the main reactions to this launch, with the 3 most representative comments per theme."*
* **"Download as DOCX"** or **"Download as CSV"** to share with your comms or research team.
{% hint style="info" %}
**Heads-up:** Your first 100 URL extractions are free. For what happens beyond that, see [Plans, Limits & Credits](/ailyze-docs/reference/plans-limits-and-credits.md).
{% endhint %}
## Important notes and limits
* **Respect platform terms and privacy laws.** Only collect publicly available data, and use it in line with each platform's terms and the law.
* Some sources come back as spreadsheets; **generic web pages** come back as Word documents instead — analyze those with **"Analyze docx/pdf files"**.
## A realistic example
> **Scenario:** A product-launch video with **3,000 YouTube comments** and a related **Reddit thread**. You collect both, combine the comment columns, theme them, and compare reactions between the two platforms. You get the top praise and concerns, quantified and quoted — useful for a launch retrospective.
## Next steps
* [Collect social media and web data](/ailyze-docs/collect-research-data/collect-social-media-and-web-data.md)
* [Analyze product reviews](/ailyze-docs/common-research-workflows/analyze-product-reviews.md)
* [How to read your results](/ailyze-docs/analyze-data/how-to-read-your-results.md)
* [Social media and web data issues](/ailyze-docs/troubleshooting/social-media-web-data-issues.md)
---
# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.
## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.
Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:
```
GET https://ailyze.gitbook.io/ailyze-docs/common-research-workflows/analyze-social-media-discussions.md?ask=&goal=
```
`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.
Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.