Track Sentiment Shifts Across Review Batches
Spot when customer mood changes before metrics catch up
Star ratings stay stable while customer tone quietly sours. By the time your dashboard flags a problem, you have lost weeks of trust.
That is the gap this drop closes. It is a marketing skill built for Kimi, and it takes about ten minutes to set up the first time. After that it runs in under a minute.
Who should use it
Marketing managers monitoring brand health across multiple touchpoints or product cycles
How it works
The skill file does five things, in this order.
1. Split reviews into time windows. Group your reviews by month, campaign period, or product release cycle so you can compare periods directly.
2. Ask for sentiment scoring per batch. Request a sentiment score and recurring theme summary for each time window separately.
3. Identify new complaints. Look for themes that appear in recent batches but were absent or minor in earlier ones.
4. Map changes to business events. Cross reference sentiment drops with launches, price changes, support policy updates, or supplier switches.
5. Build a trend brief. Summarize each period in two lines: prevailing sentiment and biggest new issue.
What comes back
A timeline that shows exactly when customer mood shifted and what triggered it
The mistake to avoid
Run this every quarter and archive the output. Patterns that repeat seasonally or after specific changes become predictable, giving you lead time to prepare responses.
Running it
Paste the prompt block into a new Kimi chat and attach your files in the same message. Kimi's long context means you should upload everything at once rather than in batches, splitting the input is what causes it to lose cross-references between documents.
Where this fits
On its own, one skill saves an hour a week. The compounding happens when three or four of them run in sequence on the same input, the same transcript that produces a scope of work also produces the follow-up email and the project brief. That is the point at which it stops being a prompt and starts being an internal tool. If you want that wired into the systems your team already uses, that is the work 67 Digital does.
In the file
I have customer reviews spanning {time_range}. I have grouped them into the following periods:
{list_each_period_with_date_range}
For each period separately:
1. Assign an overall sentiment score from -10 (very negative) to +10 (very positive)
2. List the three most frequent complaint themes
3. List the three most frequent praise themes
4. Flag any new themes that did not appear in earlier periods
Then create a comparison table showing:
Period | Sentiment Score | New Complaints | New Praise | Mood Shift
In the Mood Shift column, describe in one sentence what changed compared to the previous period.
Finally, if any sentiment score dropped by 3 points or more between consecutive periods, highlight that period and suggest which business event or change might correlate with the drop based on the themes.
Reviews by period:
{paste_reviews_grouped_by_period}