Open and inspect your export
Open the CSV in Google Sheets or Excel. Check row count against the scrape preview in Social Scraper+. If numbers are off, you may need to expand collapsed Reddit threads and re-scrape.
Typical columns: author, body text, score, timestamp, permalink. Reddit exports may include parent comment IDs for threading; X exports include handles and engagement counts.
Filter and sort for signal
Sort by score descending to see top comments first. Filter out AutoModerator or bot accounts if they clutter results.
Use text filters or SEARCH() for keywords: competitor names, feature requests ("wish," "need," "missing"), sentiment hints ("love," "hate," "broken").
Combine multiple scrapes
Append CSVs from several threads into one sheet. Add a column for source URL or scrape date so you know where each row came from.
Pivot tables work well for counting mentions: rows = competitor name found in body, values = count of comments. Quick and dirty market map.
When CSV is not enough
Long threads with nuanced arguments may need AI summarization. Export the scrape with the AI button and paste into ChatGPT for theme clustering.
For scripts and automation, use JSON export instead of CSV. Nested reply structure survives the trip into Python or Node better.