Built the analytical infrastructure and ran weekly delivery for a 21-week guest experience programme at one of Dubai's flagship lifestyle destinations — closing the loop between guest sentiment, operational decisions, and external reputation.
About this engagement — I joined this project at the invitation of Emma Logan of CX Collective, who designed the Voice of the Customer methodology and shaped the programme's analytical framework. My remit was to automate the data pipeline, run the weekly client delivery, and report on programme performance to the senior team on a biweekly cadence. What follows describes the work I led within that scope.
The Square at Nad Al Sheba Gardens is one of Shamal Holding's flagship lifestyle destinations — a premium outdoor venue offering curated F&B, live entertainment, family programming and seasonal cultural activations. After the venue's first months of operation, the Shamal leadership team needed answers that no operational dataset alone could provide.
How did guests actually feel about the experience? What was working and what was friction — by component, by guest segment? Was the venue building loyalty or churning through novelty traffic? Where should operations invest the next dollar of effort, and which fixes would create the most leverage on guest sentiment?
The brief was to build, run and report on a continuous Voice of the Customer programme that would answer those questions every single week — in time for operational decisions to be made within the same operating week. In Dubai's competitive lifestyle and hospitality sector, the inaugural seasons are a critical window for establishing reputation, repeat visitation, and word-of-mouth — all of which compound or decay quickly.
Five interlocking components, designed to turn raw guest sentiment into operational decisions inside the same week it was captured.
In-field collection at the moment of exit. Trained Customer Feedback Liaisons on-site every operating day, equipped with a tablet-based, 10-question conversational survey designed to be completed in under 90 seconds. Sampled across restaurant departures, playground exits, valet returns and general venue exits — to ensure no single experience type dominated the dataset. Branded QR totems captured guests who preferred to self-respond. Both English and Arabic were handled natively.
A live Power BI environment for real-time exploration. Built the full data pipeline — Power Query (M) for ingestion, DAX for NPS, CSAT Top-2-Box, and week-over-week deltas with significance testing. Embedded Python inside Power BI for Arabic-to-English translation of free-text fields and keyword-based topic classification of open-ended responses against a frozen taxonomy of CX themes.
A weekly Voice of the Customer report — 21 of them, on cadence, no skipped weeks. A 15-page branded PowerPoint following a standardised structure: cumulative dashboard, NPS-by-week timeline annotated with operational milestones, themes-in-dynamics for each component, returning-guest deep dive, segment focus, and a CX Improvement Action Plan with editable Decision / Owner / Status columns that operations actually used.
A separate CEO-level executive summary every week. One to two pages of structured text designed to be readable in five minutes by senior leadership — performance snapshot, what guests loved, where to do better, opportunities ahead, strategic context, and verbatim guest messages.
Eight-page interactive dashboards for self-service exploration. Built in HTML/React with Recharts and Tailwind, colour-matched to the Shamal identity. Stakeholders who wanted to explore the data themselves had Executive Summary, NPS Deep Dive, CSAT Component Analysis, Component Dynamics, Guest Improvements, Returning Guests, Performance Drivers, and Recommendations views — all live, all explorable.
Across 21 weeks and 1,121+ guest responses, the analysis surfaced a series of operationally and strategically significant findings. The headline arc — the NPS journey itself — tells the programme's story.
Each bar reflects the average NPS across the named phase; the highlighted Week 13–14 bar represents the programme high (Hag Al Laila weekend). The cumulative programme NPS settled at +61 by Week 21.
Children's Activities was the most persistent friction theme of the programme — feedback consistently centred on pricing for paid play activities and the case for a free play zone. Surfacing this in writing every week led to specific operational planning by the venue team.
One of the more actionable findings of the programme: returning guests gave a higher NPS but lower component CSAT than first-timers. The interpretation was clean — returning guests have stopped being impressed by the novelty and are now judging the everyday execution. That's healthy maturity. It tells operations exactly where to focus the continuous-improvement budget.
The returning-guest share itself moved decisively over the programme: from an early-programme average of ~45% to a late-programme average of ~67% — a +22 percentage point uplift in five months. A venue can drive footfall once with marketing; only a great experience drives the second visit.
A more granular cross-tabulation surfaced one further insight: Female First-Timers were the lowest-scoring single segment, but a quarter of the entire respondent base — making focused improvements to their first visit the highest-ROI opportunity in the experience.
To validate that on-site sentiment was reflected in external reputation, I scraped and analysed all 521 Google Maps reviews of the venue from October 2024 to February 2026 — comparing the year before the programme to the year it ran.
A meaningful uplift on a 5-point scale where movement is typically slow. A sustained 4.5★ average compounds over time — it raises booking conversion, increases inclusion in "best of Dubai" media coverage, and lowers customer-acquisition cost across all paid channels. The programme paid back outside the venue gates as well as inside them.
"This is the external proof point that the on-site improvements being driven by the weekly Voice of the Customer programme were translating into the most public-facing measure of guest sentiment that exists in modern hospitality."
The clearest illustration of the programme's value isn't in the headline numbers — it's in the specific, week-by-week course corrections it enabled. Parking & Access was the single most operationally critical theme of the programme — for an outdoor lifestyle venue in Dubai, the experience starts before guests even step out of the car. We tracked it relentlessly week after week, with focused incremental improvements logged in every report.
In Week 2, Parking & Access scored 37% — a clear launch-phase infrastructure problem. Guest feedback was specific: access lanes, valet pricing, sandy surfaces, lighting, congestion at peak. The weekly report flagged this as a HIGH-priority operational issue with the qualitative evidence attached.
By Week 4, after Operations reconfigured access lanes and added overflow capacity in response to the Week 2 report, Parking had jumped to 92%. It oscillated through subsequent weeks as new sub-issues surfaced (sandy surfaces, valet pricing, lighting all surfaced in turn) — and each time, the weekly cadence let operations isolate, address, and verify the fix in two-to-three week cycles.
By Week 21 — Eid weekend, with the highest sample volume of the programme, Parking & Access scored a perfect 100%. The first perfect score on any major component in the programme. A 21-week journey from the worst-performing component to the programme high.
This is what the programme was built to do. Not to produce a single annual report, but to give operations a steady weekly drumbeat of decision-grade evidence — and a closed-loop framework for acting on it inside the same operating week the feedback was captured.
NPS journey from +32 to a peak of +85, with cumulative programme NPS settling at +61.
Returning-guest share grew +22 percentage points over the programme — a clear loyalty signal.
Parking & Access transformed from the worst-performing component to a perfect score by Eid weekend.
Externally validated by a +0.17-star Google Maps uplift year-on-year (4.34 → 4.51).
Programme delivered on cadence — 21 weekly reports, 21 CEO summaries, and three interactive dashboards, with no skipped weeks.
In our view, six things made this engagement successful — and they're the things we'd repeat on any future programme of this kind.
Weekly, every week, for 21 weeks. No skipped weeks, no late reports. CX intelligence loses 90% of its value if it doesn't arrive in time for operational decisions.
Tablet-led, at the moment of exit, with trained collectors. The combination produced both volume and quality — and most importantly, qualitative depth.
The Priority Score formula, the cumulative-plus-incremental approach, the frozen category taxonomy — written down, applied identically every week, handed over for ongoing use.
The detailed PowerPoint for the operations review, the 5-minute executive summary for senior leadership. Different audiences, different formats, same underlying truth.
The Action Plan slide had Decision / Owner / Status / Mitigation columns that operations actually used. The data didn't just inform — it tracked.
Tracking Google Maps reviews in parallel with on-site survey data meant we could prove the programme's external impact, not just its internal numbers.
AUTODOC AG · 27 countries
Read Case StudyU-LEAD with Europe · Ukraine
Read Case StudyIf your venue, retail estate, or service property could benefit from continuous, decision-grade guest feedback — let's talk.