Two use cases — same platform
A sports retail chain — 8 stores
The network manager receives Google reviews from all 8 stores but has no time to read them. The national average score is 4.1 — everything seems fine.
After 3 months of TextIQ + PredictIQ analysis:
- Store 4 has 73% of negative mentions on checkout wait time — invisible in the overall score
- 12% of customers mentioning 'out of stock' do not return within 60 days (average PredictIQ score: 68%)
- The Bordeaux store has the best 'satisfaction' emotion score on sales advice — its practices are rolled out to other locations
- 3-month trend: frustration score dropped from 34% to 21% after targeted training
A home décor e-commerce site
400 reviews per month across Google, Trustpilot and product pages. The team reads the 1-star reviews but ignores the 3-star ones — which contain the richest weak signals.
After 6 months of TextIQ + VisualIQ + PredictIQ analysis:
- 67% of 3-star reviews mention 'damaged packaging' or 'fragile product' — never flagged to the carrier
- VisualIQ detects on SAV photos that 89% of damage affects one packaging format (picture frames)
- Customers mentioning 'delivery time' in their first review are 3× more likely not to reorder (PredictIQ)
- 6-month trend: SAV return rate reduced by 23% after changing frame packaging
