# Catch products whose reviews are slipping before returns spike

> Every Monday at 8am, see which products are quietly losing stars, what customers keep complaining about, and which flaws deserve a quality ticket.

- Workflow type: agent
- Services: Yotpo, Slack, Linear
- Categories: Product, Operations
- Published: 2026-08-10

## What it does

- Walks your product catalog every Monday and compares each product's current star rating against last week's, so slow slides surface early instead of showing up as returns
- Reads the last 30 days of review text and groups the complaints into recurring themes like sizing, durability, packaging, or shipping
- Posts a Slack digest ranked by biggest rating drop, with one real customer quote per product so the problem is obvious at a glance
- Opens a quality ticket in Linear automatically when three or more separate customers report the same flaw on the same product

## What you'll need

- A Yotpo account that is already collecting reviews on your products
- A Slack workspace and the channel where your product or merchandising team talks
- A Linear workspace and the team that should receive quality issues
- A sense of which products matter most, since the weekly report covers a set number of products rather than every item in a large catalog

## Prompt

Every Monday at 8am, tell me which of my products are quietly getting worse in their reviews, before the returns pile up. Run this on a weekly cron schedule, Monday at 8am in my local timezone.

Start by walking my catalog with the Yotpo Retrieve Products operation. Cap the walk at a configurable product limit, defaulting to the first 50 products and ideally pointed at my top sellers, so a large catalog does not turn into thousands of calls. For each product in that set, call Yotpo Retrieve Bottom Line for Product to get its current average score and total review count.

Bottom Line is a per product call and the Yotpo Core API allows only 5 requests per second per store. Pace the per product calls to stay under that limit, and when a call comes back with 429 Too Many Requests, back off and retry using the number of seconds given in the RateLimit-Reset header.

Rating decline needs a prior baseline, so save each run's average score and review count per product and compare this week's numbers against the previous run. On the very first run there is nothing to compare against. In that case, state plainly in the digest that this run is establishing the baseline, report the current scores only, and skip the ranking and the ticket filing.

Next, use Yotpo Retrieve All Reviews to pull the actual review text from the past 30 days for the products that moved. Read the negative reviews, meaning 3 stars and below, and cluster the recurring complaints per product into themes such as sizing, durability, packaging or shipping. Create a new theme when the review text genuinely calls for one rather than forcing every complaint into a fixed list.

Rank the catalog by rating decline, largest drop first, and post the digest to my product channel using the Slack Send a Message operation. For each of the biggest movers, show the product name, last week's average score, the current average score, the change, the current review count, the top one or two complaint themes, and one short representative customer quote. Keep it scannable rather than exhaustive. If nothing meaningfully declined this week, say that in a single line instead of padding the message.

When the same defect theme comes from three or more separate reviewers on one product, file a Linear issue with Create Issue so quality has something concrete to chase. Make that three reviewer threshold a customizable knob at the top of the workflow. Title the issue with the product name and the theme, and put the supporting customer quotes, their review dates, the product's rating change and its current review count in the description.

Before filing, use Linear Search Issues to check whether an open issue already exists for that same product and theme combination from an earlier run. If one does, do not open a duplicate. Note in the Slack digest that the theme is recurring and link the existing issue instead.

Keep the emphasis on rating decline and complaint themes across the catalog. The Linear issue is a threshold side effect for the worst repeat offenders, not the main output. The main output is the weekly Slack digest of the biggest movers.

## How to customize

- Change the day and time, or run it every other week instead of weekly
- Set how many products to watch each run, for example your top 50 sellers
- Adjust how many separate customers must report the same flaw before a quality ticket gets filed
- Rename the complaint themes to match your product line, such as fit, battery life, or scent

## FAQ

### What happens on the very first run?

There is nothing to compare against yet, so the first report records where every product stands today and says plainly that it is establishing a baseline. From the second week onward you get real rating movement.

### Will this work if some of my products only have a few reviews?

Yes. Keep in mind that a product with very few reviews can swing a lot on a single opinion, so the digest always shows the review count next to the rating and the change. That way you can tell a real trend from one unhappy customer.

### Does it file a ticket for every bad review?

No. A ticket is only opened when the same specific complaint appears from three or more separate customers on one product, and that number is yours to change. Your quality team gets patterns rather than noise.

### Can I send the digest somewhere other than a product channel?

Yes. Any Slack channel or direct message works. Many teams post to a merchandising or quality channel and mention the product owner directly in the message.

### What if my catalog is very large?

The report covers a set number of products each run, so you can point it at your top sellers instead of every item you stock. This keeps the Monday digest fast to produce and short enough that people actually read it.

Use this prompt in General Input: https://www.generalinput.com/prompts/catch-products-whose-reviews-are-slipping-before-returns-spike