Your reviews are data you have never read systematically
A product with 40 reviews contains hundreds of signals: specific feature praise, recurring complaints, edge cases, and comparison mentions. Most store owners skim the lowest-rated ones after a bad week and call it done. The pattern that appears in 30 of those 40 reviews stays invisible.
You can read them one by one. That takes an hour, produces no structure, and leaves everything in your head. The next person who needs to know must start over.
What most people do instead
A better way: structured analysis in seconds
Run tp analyze reviews -product_id=312 in the TrueCommander navigator. The command pulls up to 50 of the most recent approved reviews, includes each star rating, computes the average, and sends the full set to AI with a prompt built for structured feedback analysis. The navigator renders the themes, praise summary, and complaint summary in one clean response. It reads nothing except what is already in your database, it writes nothing back.
Read-only, always. The command fetches approved reviews and sends them to AI. It does not post, edit, approve, or delete anything. Your reviews stay exactly as they are.
How it works
WooCommerce is required. The command calls wc_get_product($product_id) to confirm the product exists, then fetches up to 50 of its most recent approved reviews via get_comments with status=approve and comment_type=review. Each review body is truncated to 600 characters. The star rating is read from the rating comment meta field. The average is computed from all returned ratings. The full set goes to the AI proxy with a prompt requesting structured output: key themes, what customers praise, what they complain about, and any notable patterns.
/wp-admin/edit.php?post_type=product or in the URL when you open a product to edit.| Parameter | Value |
|---|---|
-product_id (required) | Numeric WooCommerce product ID |
| Reviews fetched | Up to 50 most recent, status approve, type review |
| Review body limit | 600 characters per review before sending to AI |
| Ratings | Included per review; average computed and sent with the prompt |
| Requires | WooCommerce. Returns an error if the plugin is deactivated |
| No reviews case | Returns a clear message if the product has no approved reviews |
| AI provider | TrueCommander's server-side proxy, no API key required |
| Can be used in |
Real example
Your flagship jacket has 40 reviews and a 4.0 average you haven't interrogated in months. You type tp analyze reviews -product_id=88. The analysis comes back in seconds: customers consistently praise the hood adjustment and the water resistance, but 11 reviews mention the same thing, the left zipper catch sticking in cold weather.
You didn't know it was 11. You knew it was "a few." The AI grouped them, counted the pattern, and surfaced the language customers use. You forward the finding to your supplier contact the same afternoon and update the product description to note the fix is coming. One command, one concrete action.
Goes further with TrueCommander
analyze reviews to understand sentiment, then suggest upsell to see what complementary products customers might also want. Two commands, complete picture.summarize to condense the output into a short brief you can drop into a shared doc or product management ticket.