Trust is the critical path: turning around Marriott's first cloud point of sale
The program had a $200 million price tag, a hard payment card industry compliance deadline, and the challenge every industry first carries: no one had done this before, so the early rollouts surfaced issues at a scale no pilot could have predicted, and owner confidence was being tested. The vendor was navigating a major transition of its own. I arrived new to point of sale, with limited data, expected to make a defensible course-correction call inside 30 days.
The situation
Marriott was moving thousands of hotels from on-premises point of sale to a single cloud platform, the company’s first cloud solution. The strategy was right, and the early execution was proving harder than anyone wanted, which made each next owner conversation more delicate. In hotel operations the point of sale touches nearly everything: revenue, service speed, the guest’s last impression at checkout. When it fails, owners feel it immediately and remember it for a long time.
What I did
I went to the field first. Not to audit, but to use the system the way servers and managers used it, feel the failures myself, and signal to owners that they had an advocate who would work every issue to ground. That first trip shaped everything after it.
Back at headquarters, I pulled a cross functional tiger team to map the workflows end to end, ran structured root cause sessions, and ranked the issues by impact, severity, and effort. The analysis pointed to a call that was uncomfortable to make: keep the strategy, but slow deployments by two months to fix the three highest impact issues. On paper it looked like a setback. In practice it was the fastest route to scale, because every deployment that ran ahead of the fixes made the foundation harder to rebuild.
I then got in front of owners and senior Marriott executives monthly, reporting on what we had delivered, why it mattered, and what came next. Trust gets rebuilt through visible accountability on a cadence, not through one grand apology. As the platform stabilized, we standardized training, communications, deployment templates, and integration governance, then expanded scope into payments, reporting, reservations, and loyalty. Integrations grew from 7 to 35, and each one removed manual work or unified a journey for guests and associates.
What happened
Owners who had lived through a hard early chapter agreed to scale across their portfolios, which is the trust metric I would defend above any other. The program spanned 5,000 properties in 130 countries, with 80 percent slated for conversion, and every one of them converted: more than 4,000 hotels in under three years, including 99 percent penetration in North America. Operating costs fell about 25 percent, average check rose about 20 percent, and service cycle times dropped about 30 percent. The vendor invited me onto its customer advisory board, which I took as a sign the relationship had moved from adversarial to collaborative.
The two month pause showed up as red on every status report before it showed up as green anywhere. If you measure a transformation only by deployment velocity, you will make the wrong call at exactly the moment it matters most. And the operational gains had many parents: better product, better training, better integrations. The part I own with confidence is the sequencing. Trust first, stability second, scale third.