108 sites.
3 countries.
12 months.

A global F&B operator needed one POS across 108 hybrid restaurant-and-retail outlets in India, the UAE and the UK. Their own first attempt had stalled in pilot and cost a day of trading. We ran the second attempt from the client's side. Every outlet came up overnight and opened clean.

Client

$1B+ global F&B operator. 108 hybrid outlets (restaurant + retail counter), company-owned and franchise.

Platform

Restroworks POS on an existing SAP S/4HANA backbone. Payments, aggregators, KDS, e-invoicing.

Role

Implementation PMO, end to end: design & blueprint, SIT/UAT, phased rollout, hypercare.

Timeline

12 months, pilot plus 4 waves, against an 18-month internal plan.

Outcome Measured, sources in the tables below
96%first-time-right cutovers, 104 of 108 sites
0lost trading days across 108 cutovers
~4 hrsper-site overnight cutover, down from a 2-day pilot scramble
<0.4%failed-transaction rate in the first 72 hours per site
The situation A failed first attempt

Every till was different. One had already failed.

The client had expanded internationally through acquisitions and its own store footprint, and the result was a patchwork of legacy tills. Every outlet runs a hybrid model: a restaurant and an integrated retail counter selling packaged food under the house brand. One POS has to handle dine-in, packaged goods and delivery orders, with each sale taxed correctly, because packaged snacks, dine-ins and deliveries can attract different tax treatment.

The estate: roughly 90 sites in India, 12 in the UAE, 6 in the UK. Company-owned and franchise. Several on patchy connectivity. Menu, SKU, recipe and pricing data lived in inconsistent shapes across the legacy systems, and cross-region reporting was hand-assembled and nine days late.

An earlier attempt to self-run the rollout had stalled in a single-market pilot. One botched cutover cost a full day of trading. With a loyalty and menu program depending on one clean operating spine, a second failed rollout across 100+ outlets was not an option.

What we did Design once, test hard, repeat 108 times
  • Prove once, repeat 108 times. A clean re-pilot on 3 company-owned Delhi-NCR outlets hardened the runbook and the hybrid config before any scale.
  • Region-clustered waves. Dense metro clusters first, company-owned before franchise, India before international, festival peaks blacked out.
  • Store-in-a-box. Till, KDS, payment terminal, weighing-scale integration and label printer pre-staged and pre-configured before shipping. On-site work was install and validate, not build.
  • Overnight cutovers. Each outlet cut over after close and before open. Median window about 4 hours. No site lost a trading session.
  • Offline-first go-lives. Every till validated to trade through a dropped line and sync on reconnect, because several sites could not assume a stable connection.
Master wave plan for the 108-site POS rollout: pilot and four waves across a 12-month axis with festival blackout bands
Exhibit 1Master wave plan, anonymized
Wave sequencing
WaveScopeSitesWhy sequenced here
PilotDelhi-NCR, company-owned3Harden the runbook and hybrid config; rebuild trust after the failed first attempt
Wave 1North India metros~32Densest cluster, company-owned, best support coverage
Wave 2West & Central India~30Scale the machine; first mixed owned + franchise wave
Wave 3South & East India~25The longest tail, franchise-heavy
Wave 4UAE, then UK~18International last, once the playbook was proven
Per-site lifecycle, run inside every wave
StageWhat happenedExit gate
Design & blueprintCurrent-state mapping across dine-in, counter, online and delivery; one Guest/Food/Money system of record; menu, SKU, recipe and combo rationalization; per-category tax mapping; integration design across SAP, payments, aggregators, KDS and e-invoicingDesign sign-off
Build & configureOne golden standard config, with market localization layers for tax, language and receipt formatsConfig baseline frozen
Data migrationMenu, pricing, recipe and loyalty data extracted from legacy tills, cleansed, loaded, reconciled per siteMigration reconciled
SITEnd-to-end POS to SAP to payments to aggregators to KDS, payment and UPI certification, e-invoice validation, festival peak-load testSIT pass
UAT40+ scripted scenarios run by store operations and finance; defects logged, triaged and retested to closureUAT signed off
CutoverPer-site readiness checklist, dress rehearsal, overnight runbook, rollback plan on standbyGo-live sign-off
HypercareRole-based training, train-the-trainer for franchise, war-room triage with a first-day defect SLAStabilized & handed off
Measured results Before / after
MetricBeforeAfterHow measured
Full rollout18-month internal plan12 months, 4 wavesPMO wave tracker
First-time-right cutovers1 failed pilot96% (104 of 108)Cutover sign-off log
Per-site cutover window~2 days in pilot~4 hours, overnightInstall crew logs
Lost trading days1, at the failed pilot0 across 108 cutoversTrading and cutover logs
Defects at go-live1.4 per site average, all closed within 48hDefect log
Cross-region reporting~9 days, hand-assembledNext morning (T+1)BI refresh logs, 60 days post-wave

“Our first switch to the new POS was not a success. Our staff struggled to program-manage it and it ran well over time. We decided to bring in an outside PMO and saw the results. This time every outlet came up overnight and opened clean, on a schedule we could actually see. Our franchise partners noticed the difference.”

COO, global F&B brand

Anonymized case study. Client-identifying details redacted. Reference available on request.

Next case NetSuite implementation: live in 11 months, 4% under budget →