> **For coding agents and LLMs:** This is one Social Fetch marketing page (markdown export). For API orientation start with [`/llms.txt`](https://www.socialfetch.dev/llms.txt); for auth and crawl rules use [`/agents.txt`](https://www.socialfetch.dev/agents.txt); for every operation use [`/llms-endpoints.txt`](https://www.socialfetch.dev/llms-endpoints.txt).

## This page

- **On-site (HTML):** <https://www.socialfetch.dev/use-cases/comment-buyer-intent-api>
- **Markdown (.mdx) URL:** <https://www.socialfetch.dev/use-cases/comment-buyer-intent-api.mdx>

---

# Buyer intent from comments

**How do I find buyer intent in social media comments?** Search for the words buyers use, like "alternative to" plus a competitor's name. With Social Fetch, Reddit comment search finds matching comments, then the comment routes for Reddit, YouTube, TikTok, Instagram, X and LinkedIn return the whole thread. One API key, one response shape, and you pay per page of results.

Find the people asking for an alternative to a competitor, a tool recommendation or a fix, in Reddit threads and under videos and posts. Then read the whole thread.

## A real run, measured and priced

Find people asking for an alternative, then read the thread, run on 21 Nov 2026 against live public data. 22 people weighing a switch in the last month, and the thread where one asked.

| Request | Result | Credits | Time |
| --- | --- | --- | --- |
| `GET /v1/reddit/search?query=alternative to hubspot&type=comments&sortBy=new&trim=true` | 22 comments from 28 Oct to 21 Nov, 12 of them in r/CRMSoftware. | 1 | 2.2 s |
| `GET /v1/reddit/posts/comments?url=https://www.reddit.com/r/CRMSoftware/comments/1xxtfx5/` | A 41-comment thread; 25 top-level comments from the last five days. | 3 | 4.7 s |

Total: 4 credits, 7.0 s of server time. Starter pack ($14 per 1,000 credits): $0.056. Scale pack ($1.64 per 1,000 credits): $0.0066.

Intent lives in the comments. Buyers say what they want there, not in your form. A Reddit thread titled "Thinking about moving away from HubSpot" gathers 41 comments in five days, and vendors who never see it miss the deal.

Post search misses most of it. The intent is in the replies, and every network nests replies, pages them and hides older ones differently.

You need comment search. Search for the phrases buyers use, and get full threads you can read and score, in one shape across networks, on a schedule you control.

## Endpoints

| Method | Path | What it does |
| --- | --- | --- |
| GET | `/v1/reddit/search` | Site-wide Reddit search. Set `type=comments` to search comments, not posts, for phrases like "alternative to". |
| GET | `/v1/reddit/posts/comments` | The full comment tree of a Reddit thread, for reading what people recommended and why. |
| GET | `/v1/youtube/videos/comments` | Comments under a review or comparison video, where viewers say what they bought instead. |
| GET | `/v1/tiktok/videos/comments` | Comments on a TikTok video, for consumer products and brands with a young audience. |
| GET | `/v2/instagram/media/{media}/comments` | Comments on an Instagram post or Reel, including a competitor's own posts. |
| GET | `/v1/twitter/tweets/replies` | Replies to a post on X, such as a competitor's outage or pricing announcement. |
| GET | `/v2/linkedin/activities/{activityId}/comments` | Comments on a LinkedIn post, where B2B buyers ask peers for recommendations. |

## Workflow

1. **List the phrases buyers use.** "Alternative to", "switching from", "anyone using" and "recommend a", each with your competitors' names and your category. Keep the list in your own config.
2. **Search comments on a schedule.** Run Reddit comment search for each phrase, sorted by newest. Store comment ids so the next run only handles new ones.
3. **Read the thread.** For each match, pull the thread's comments. The question is often in the post and the intent in the replies: what people tried, what they paid, what annoyed them.
4. **Score with your own model.** Pass the text to your classifier or an LLM: is this a buyer, which competitor, how urgent. Your thresholds and labels stay in your product.
5. **Alert the right person.** Send strong matches to Slack or your CRM with the link, so a person can reply in their own words. Watch competitor posts and videos for their comments too.

## Example: newest Reddit comments asking for an alternative to HubSpot

```bash
curl -sS \
  -H "x-api-key: YOUR_API_KEY" \
  -G "https://api.socialfetch.dev/v1/reddit/search" \
  --data-urlencode "query=alternative to hubspot" \
  --data-urlencode "type=comments" \
  --data-urlencode "sortBy=new"
```

## Why intent tools read comments here

- Comment search, not just posts. Reddit comment search finds the question even when the post title says nothing about it.
- Six networks, one shape. Reddit, YouTube, TikTok, Instagram, X and LinkedIn comments load through one parser.
- Pay per page of results. No seats, no keyword caps, so you can watch every competitor at once.
- You only pay for completed lookups. Requests that fail on our side cost nothing.
- You own the scoring. We return the text, authors and timestamps; your model decides what counts as intent.

## FAQ

### How do I find buyer intent in social media comments?

Search comments for the phrases buyers use, such as "alternative to" with a competitor's name, then read the full thread. `GET /v1/reddit/search?type=comments` finds the comments, and the comment routes for each network return the threads in one shape.

### Which networks can I read comments from?

Reddit, YouTube, TikTok, Instagram, X and LinkedIn. Reddit also supports comment search across the whole site; on the others you start from a post or video, for example a competitor's own.

### Should I reply to these comments automatically?

No. Use the data to find conversations and alert a person, who replies in their own words where the community allows it. Automated replies get accounts banned and buyers annoyed.

### How fresh are the results?

Sort by newest and the latest comments come first. In the run above, the newest matching comment was posted the same day.
