Internal pilot, not a client engagement. The subject is anonymised and review text is withheld on this public page. Client reports include the member records behind each finding, with authors pseudonymised.
Study summary
| Subject | A sports-retail chain in Malaysia, across several outlets |
|---|---|
| Sources | Google Maps reviews and X |
| Records analysed | 412 |
| Findings | 21: 19 recurring patterns and 2 coordinated-activity flags |
| Thresholds used | Defaults: 4+ records, intensity 0.6+, 80% concentration |
Coordinated praise suspected
- Records in cluster
- 12
- Average sentiment
- +1.00
- Weighted reliability
- 0.07
Observed
Twelve reviews with near-identical wording, all at maximum positive tone, praising the same staff member. They arrived as one burst, from sources with almost no reliability.
Inferred
These are unlikely to be independent customer experiences. The pattern fits solicited or coordinated praise.
Not known
Whether the reviews were paid, prompted by staff or written by one person. The outlet's other reviews are not affected by this finding.
Promotional spam, not customer opinion
- Records in cluster
- 7
- Topic
- Clearance sale
- Concentration
- 80%+
Observed
Seven near-duplicate posts promoting a sports clearance sale, with intense tone and concentrated in one source or one day.
Inferred
Promotional amplification. It should not be counted as evidence of how customers feel.
Not known
Who posted it. It may be the retailer's own or an affiliate's marketing, which is legitimate but still not customer sentiment.
What happens next
- Trends are weighted by source reliability, so these clusters barely move the study's sentiment trend.
- The sources that dominated each cluster lose reliability for later runs.
- Recommended: review the member records with the client before the finding is shared any further.
Want this for a brand, talent or public figure you look after? Book a reputation read, or read how each score is calculated.