Product research with AI: real complaints and wishes from Reddit and X
Your AI reads what people say about a product category in threads and posts, and comes back with the complaints and wishes that keep coming up, each with the post it came from. It uses the Comment analysis and Brand buzz tools and 8 interfaces. A real run recorded 2026-10-18 answered “What do people complain about with standing desks on Reddit and X?” with 3 requests for 88 credits (≈ $0.18–$0.29). Failed calls are free.
What you get
Your AI reads what people say about a product category in threads and posts, and comes back with the complaints and wishes that keep coming up, each with the post it came from.
The useful complaints sit deep in long threads. Reading enough of them to tell a pattern from one loud voice takes hours.
What it uses
The ready-made tools do the whole job in one request: SocialToAI runs the searches, reads the posts and comments, and returns one result with its sources. Each is priced before it runs.
Your AI can also call the interfaces one at a time. These are the ones this job reads; each costs a fixed number of credits per call.
| Platform | Interface | What it returns | Price key | Credits |
|---|---|---|---|---|
| Search | Find posts by keyword: text, author, comment count and publish time. | reddit.search | 2 | |
| Comments | The comments under one post: text, reply count and time. | reddit.comments | 2 | |
| Community posts | Get a page of the posts of a Reddit community by its name, hot, newest or top score first. | reddit.subreddit_posts | 2 | |
| X | Search | Find posts by keyword: text, author, likes, comment count, views and publish time. | x.search | 2 |
| X | Thread context | Get the posts of an X thread by the ID of any post in it. | x.thread | 9 |
| YouTube | Comments | The comments under one video: text, likes and reply count. | youtube.comments | 2 |
| Xiaohongshu | Search | Find notes by keyword: text, author, likes, comment count and publish time. | xiaohongshu.search | 12 |
| Xiaohongshu | Comments | The comments under one note: text, likes, reply count and time. | xiaohongshu.comments | 12 |
A worked example
A real run on the live service, recorded 2026-10-18: the question, the calls the AI chose and what it answered. The credits are what the run was charged.
What do people complain about with standing desks on Reddit and X?
| Platform | Tool | Sent | Requests | Returned | Credits |
|---|---|---|---|---|---|
search | query: "standing desk" | 1 | 7 posts | 2 | |
| X | search | query: "standing desk" | 1 | 11 posts | 2 |
comment_insights | targets: ["https://www.reddit.com/r/BuyItForLife/comments/1ntedtv/do_standing_desks_actually_help_or_are_they/","https://www.reddit.com/r/WFH/comments/1q1ia5q/i_bought_an_expensive_standing_desk_6_months_ago/","https://www.reddit.com/r/StandingDesk/comments/qd8b8v/was_your_standing_desk_worth_it_in_the_long_run/"] · focus: "what people complain about with their standing desk" · scale: "small" | 1 | 173 comments | 84 | |
| Total | 3 | 191 items | 88 | ||
I searched both, then read 173 comments under the 3 most-discussed Reddit threads. 55 of the 162 that could be judged complain about something; the complaints cluster here:
- Back and neck pain12%“Either position for 8 hours solid leaves my back hurting.”20 comments
- Forgetting to switch between sitting and standing11%“I forget I have it most of the time...”18 comments
- Sore feet from standing8%“My feet do get tired by the end of the day it’s better than the back pain.”13 comments
The posts it rests on, and the same calls as API requests, are in the whole recorded run.
88 credits is $0.29 at 300 credits per dollar (purchases from $10) and $0.18 at 500 per dollar (from $200). Failed calls are free, and a search that runs and finds nothing is charged as a call.
Give this to your AI
Connect SocialToAI once: in Claude, ChatGPT, Claude Code, Cursor and Codex you sign in and press Allow, with no key to paste; other clients are on the connect page. Then paste this and replace the parts in angle brackets. Never paste a key into a chat.
Use the SocialToAI tools (https://mcp.socialtoai.com/mcp) to research <product category> before we build one. Search Reddit and X for posts about it, read the comments under the most discussed ones, and group what people complain about and wish for. For each group, give how many posts mention it and two quotes with links. Tell me the credits you spent.Last updated: