Reddit keyword research
Type a seed phrase. We'll mine Reddit titles for the buyer language around it.
- Mentions in sample
- CSV export
3 of 3 free scans left.
Also: Opportunity finder · Mention finder
How this Reddit keyword research tool works
Type a seed phrase (niche or product). We run search.json three ways — raw seed, quoted seed, and seed plus (recommend OR alternative OR vs) — up to 100 posts from the last year, then tokenize titles + selftext into 2–3 word n-grams, strip stopwords, rank by count, and keep the top 25. The top 8 carry mention counts (posts in sample) plus an example post and the subreddits they came from. Arctic Shift fills the sample when Reddit search is thin. An optional OpenRouter clustering pass groups phrases when a key is configured.
How to use the table
- 1
Seed honestly
Use the words buyers type ("cold email tool"), not your brand poetry.
- 2
Read buyer language
Phrases with "recommend / alternative / vs / anyone used" nearby are money — those rows deserve a thread open.
- 3
Export
CSV download is client-side from the returned JSON. Chips mirror the buyer-language queries we ran.
Who this is for
- SEOs mining Reddit phrasing for landing pages.
- Founders naming features after customer words.
- Writers briefing titles that sound native.
Limits
One of 3 scans. Counts are "mentions in sample", never Google volume. Coverage is what Reddit search returns — imperfect, English-heavy, and blind to deleted threads.
Is this search volume?
No. It counts how often a phrase appeared in ~100 sampled Reddit posts. Treat it as buyer-language signal, not demand sizing.
Why so few results for my brand?
Young or uniquely-named products have thin samples. Try the category ("email warmup") instead of the brand.
Can I cluster the phrases?
When OpenRouter is configured we group them; otherwise the raw ranked list still ships.
Want comments in these subreddits?
We can make comments and posts on your behalf using high-karma, high-CQS aged accounts and ensure maximum visibility.
Order comments