Hello, e-Commerce enthusiast! Contribute to the 2023 edition happening this month. Click.

e-Commerce Advent Calendar

The e-Commerce geek's favorite time of year
2023 Edition

Summarize Positive And Negative Feedback From Product Reviews With AI

by Kalen Jordan
ABOUT THE AUTHOR
Kalen Jordan

Kalen is a Shopify automator at MESA. He is also a founder of MageChat, Commerce Hero and co-host of Magetalk.

This is a re-post from the MESA blog. The original post and other original post could be found here https://www.getmesa.com/blog/summarize-negative-and-positive-feedback-from-product-reviews-with-ai/


These workflows aggregate product reviews into our database and then feed them into ChatGPT in order to get a concise, bulleted list of positive and negative feedback for you to review.

In this example, we’re using the Okendo reviews API to summarize reviews for a brand called Carnivore Snax.

Here is an example of the summary:

Positive Feedback:

  • Many customers love the taste of the snack, with frequent descriptions such as "flavorful,", "delicious", "melt in your mouth" and "addictive".
  • Customers appreciate the high-quality, lean protein content of the snacks and regard them as beneficial for their diet and lifestyle.
  • The texture of the snacks is often highlighted as satisfying and the right balance of crisp, chewy, and fatty.
  • The snacks are perceived to be worth their cost due to their high quality and taste.
  • Some customers mention the benefit of ideal fat and protein ratios and the overall nourishment provided by the snacks.
  • Some reviews highlight the versatility of these snacks, suitable as quick bites or as meal replacements.
  • Some customers enjoyed the simplicity of the ingredients and respect the company’s mission to deliver healthy and nutritious snacks.
  • Many customers also mention that they keep reordering due to the overall quality, taste, and convenience of the snacks.

Negative Feedback:

  • Some customers found the snacks to be too greasy, leading them to wipe them down before eating.
  • A few customers mentioned dissatisfaction with the texture of the snacks, with complaints about them being too hard, dry, or chewy.
  • Some found specific flavors or cuts of meat not to live up to their expectations, and that certain snacks had more fat than they preferred.
  • The price of the snacks was a point of contention for some customers, with some feeling that they could be costly for regular consumption.

The first workflow saves the reviews:

And the second workflow pulls the reviews from the database and passes them over to ChatGPT for a summary:



How do I use this?

  1. Add MESA app. From the Shopify App Store, install MESA on your store. Start for free or enjoy a 7-day trial on any plan.
  2. Import this download to My Workflows. Navigate to your automations then follow the prompt to import a workflow. Upload the downloaded ZIP file from this demo.
  3. That's it! Personalize any step if you want. You got this.
  4. Questions? Shoot me an email: kalen.jordan@getmesa.com

Kalen Jordan

Interested in submiting an article?

Please check our contribute page in case you are interest to submit an article.