Evidence

Client stories

Specific notes from kitchens that asked us to read their delivery-app signals.

These notes name real constraints — tablet placement, menu photos, branch definitions — rather than star ratings. Names are used with permission; some outlet brands are generalised where owners preferred it.

Short notes

Amira K., multi-outlet nasi lemak group (Petaling Jaya)
“They caught that our Petaling Jaya branch was accepting orders twenty seconds slower after 8pm — not because of staffing, but because the second tablet sat under a tray stack. Fix was a shelf, not a hire.”
Engagement: Operations analytics review

Daniel T., dark kitchen, Subang
“The write-up named three menu photos that drove cancels once riders saw the real portion. We reshot two; the third still underperforms, which they warned us about.”
Engagement: Operations analytics review

Siti R., family restaurant with delivery add-on, Shah Alam
“Peak-hour observation was uncomfortable at first — someone with a notepad by the bagging table. By the end of dinner they had timed the printer gap we kept shrugging off. We still argue about whether to cut the slowest curry from the night menu.”
Engagement: Peak-hour desk observation

Marcus L., three-outlet chicken rice group
“The scoreboard session stopped us comparing apples to oranges. One branch had been padding ‘ready’ times by marking food ready before lids went on. Mild gripe: the half-day felt packed; we could have used another hour for franchisee questions.”
Engagement: Multi-outlet scoreboard session

Longer story — Friday nasi lemak cluster

A four-outlet cluster near Damansara asked for a full operations review after partner cancel rates climbed through Ramadan evenings. Exports showed Item X (a large set) converting well at lunch and collapsing after 7pm. Floor observation found the set required a garnish station that only lunch crews staffed.

The brief ranked three actions: remove the garnish from night listings, move the secondary tablet to eye height, and stop accepting the large set within 40 minutes of closing. Two weeks later, night cancels on that item fell; overall accept speed improved once the tablet move stuck. The owner still declined to drop a second slow item we flagged — a fair commercial call, and we noted it as unresolved in the follow-up call.