Introduction:
Singapore’s high-density geography and digital-first population have made it a prime
market for Q-Commerce. Platforms like RedMart, FairPrice, GrabMart, and Pandamart promise
ultra-fast grocery deliveries. But as competition intensifies, delivery time is the new
battleground.
Actowiz Solutions partnered with a leading last-mile logistics provider in Singapore
to use AI-powered web scraping and machine learning models to extract, monitor, and optimize
delivery time data across Q-Commerce platforms—by zone, time of day, and SKU type.
Actowiz’s Approach
1. AI-Powered Web Scraping Engine
Actowiz built scrapers that accessed real-time delivery ETA data from:
- RedMart: Based on product + postal code
- GrabMart: Dynamic delivery windows based on current fleet
load
- FairPrice Online: Delivery slot availability for next 6 hours
- Pandamart: ETA directly shown on product detail page
These were scraped every 20–30 minutes for top 500 SKUs across 50+ postal zones.
2. Data Captured
| Field |
Description |
| Platform |
GrabMart, RedMart, Pandamart, FairPrice |
| Product Name |
SKU being monitored |
| ETA (mins) |
Platform-reported estimated delivery time |
| Postal Code |
Singapore 6-digit code |
| Timestamp |
When the ETA was scraped |
| Category |
Frozen, Fresh, Packaged, Beverages, Essentials |
3. Sample Dataset (Singapore Postal Zones)
| Timestamp |
Platform |
Product |
Postal Code |
ETA (mins) |
Category |
| 2025-06-14 10am |
GrabMart |
Ben & Jerry’s |
239732 |
26 |
Frozen |
| 2025-06-14 10am |
RedMart |
Ayam Brand Tuna |
560143 |
40 |
Packaged |
| 2025-06-14 10am |
Pandamart |
Dettol Soap 4pk |
529538 |
18 |
Essentials |
Insight: GrabMart consistently offered faster delivery for frozen
products within city czones like Orchard and Clarke Quay.
4. AI Models for ETA Optimization
Actowiz used scraped data to train the following models:
- Regression Models – Forecast delivery times based on product,
time of day, and zone
- Clustering Algorithms – Identify delivery bottlenecks by
platform and region
- Anomaly Detection – Spot outliers in promised vs. actual
delivery ETAs
- Time-Series Forecasting – Predict future delays during rain,
lunch hours, or public holidays
Results After 60 Days
Logistics Improvements for Client (Fleet Operator):
| KPI |
Before Actowiz |
After Actowiz |
| On‑Time Delivery % |
71% |
91% |
| Fleet Overhead Cost |
High |
‑22% reduced |
| Delay Prediction Accuracy |
58% |
89% |
| Missed Frozen Delivery Incidents |
19/month |
3/month |
| ETA Variability Range |
Wide (10–60m) |
Narrowed (15–30m) |
Testimonial
“Before Actowiz, our teams had no reliable ETA intelligence. Now we can proactively plan deliveries, reroute fleets, and deliver within our promised windows—even in traffic.”
– Operations Head, Singapore Q-Commerce Logistics Partner
Conclusion
In Singapore’s tightly timed Q-Commerce race, delivery efficiency is a make-or-break factor. By scraping delivery time data from top platforms and combining it with predictive AI, Actowiz Solutions empowers logistics teams to meet customer expectations with confidence and consistency.
From postal code heatmaps to hourly delay prediction models, this solution transforms data into precision—keeping deliveries on time, every time.
