Pilot results
11.5× less time on expiry control. Write-offs −59%.
A two-store pilot in a Ukrainian grocery chain, November 2024. Baseline measured before go-live, result measured after one month of use. The full method is below.
The chain is not named: we publish client names only with written consent.
- 11.5×
- less time on expiry control
- 18.5 → 1.6 person-days per store per month
- −59%
- write-offs
- across the 8 product groups measured
- +0.38%
- of turnover
- extra sales from markdowns
- 16,935
- tasks in one month
- created by Freshto, completed by staff
Pilot in a Ukrainian grocery chain: 2 stores, 3,684 SKUs in 11 categories, November 2024. Baseline measured before go-live, result after one month of use.
Read the pilot resultsContext
- Format
- Grocery chain, two self-service stores
- Assortment under control
- 3,684 SKUs in 11 categories
- Go-live
- 25 October 2024
- Baseline
- 24–25 October 2024, timed on the floor
- Result period
- 11–30 November 2024 (20 days), scaled to a calendar month
- Devices
- Staff smartphones
- Integration
- Excel assortment upload, no ERP integration
- Implementation effort
- 30 hours of IT and 80 hours of consultants; 840 hours of continuous service operation
Before the pilot
Expiry control meant walking every shelf, writing dates in a logbook, checking the logbook daily and removing products according to it. We timed it: 68 seconds per product in one store and 26 in the other, for assortments of 4,261 and 2,911 products in the measured groups. With the daily logbook work and removals, that added up to 21.7 and 15.3 person-days a month, 18.5 on average.
Method
- 1
Baseline
On 24–25 October 2024 we timed the existing routine in both stores (checking a group of 52 products, logbook work, removals) and collected write-off data for the preceding period.
- 2
Setup
The assortment (3,684 SKUs) was uploaded from Excel. Markdown steps were agreed per category with the chain. Staff entered the nearest date for each product once.
- 3
Live period
From 11 to 30 November 2024 the stores worked exclusively from the Freshto morning list. Every task was logged with time and employee, which is where the "after" timing comes from.
- 4
Measurement
Time spent, write-offs and markdown sales were measured for the live period and scaled to a calendar month for comparison with the baseline.
Time on expiry control
| Before | Store 1 | Store 2 |
|---|---|---|
| Seconds per product checked | 68 | 26 |
| Products in measured groups | 4,261 | 2,911 |
| Shelf checks with logbook entry, min / month | 3,317 | 2,266 |
| Daily logbook review, min / month | 900 | 900 |
| Removals according to logbook, min / month | 1,944 | 1,255 |
| Total, person-days / month | 21.7 | 15.3 |
| After (20 days, scaled to a month) | Store 1 | Store 2 |
|---|---|---|
| Tasks completed in 20 days | 1,665 | 1,075 |
| Seconds per task | 45 | 29 |
| Total, minutes / month | 1,221 | 348 |
| Total, person-days / month | 2.5 | 0.7 |
Average: 18.5 → 1.6 person-days per store per month, 11.5× less. Note that a single check did not get faster (45 and 29 seconds per task against 68 and 26 before). The saving comes from checking only the products on the list.
Did the checks find anything?
Before the pilot, a timed check of 52 products in one category found 0 products requiring action and 26 that needed to be watched again. In the pilot month, every one of the 2,740 tasks on the list required an action: a new date, a markdown or a removal.
- 52
- Before: products checked
- 0
- Before: products requiring action
- 2,740
- After: tasks on the list
- 100%
- After: tasks requiring action
Results
| Before | After | Δ | |
|---|---|---|---|
| Time on expiry control | 18.5 person-days / month | 1.6 person-days / month | 11.5× less |
| Write-offs of expired goods (8 product groups measured) | Baseline | −59% | −59% |
| Sales of marked-down goods | Not tracked | +0.38% of turnover | +0.38% |
| Tasks created and completed | Logbook and chat | 16,935 in a month | ≈ 68 per store per day |
| Supplier reminders | Calculated by hand | 394 SKUs | automatic |
All numbers: Freshto pilot report, 2 stores, November 2024. Write-offs were measured across 8 product groups (soft drinks, beer, confectionery, snacks, grocery, frozen, bread and eggs, household chemicals). No other performance figures are published.
What we learned
- Most of the time saving comes from not walking the whole floor: about 68 tasks per store per day replaced full shelf checks.
- Markdown steps turned products that used to be written off into sales worth 0.38% of turnover.
- Supplier reminders were an unexpected benefit: 394 SKUs with short remaining shelf life on delivery were reported automatically, a calculation staff used to do in their heads.
- Initial entry of nearest dates is the one real effort; after that the daily routine is minutes.
Run the same pilot in your chain
Same method, your stores, your numbers. We will plan it with you.