Let their data make the argument: breaking a service level stalemate

Across 1,200 hotels, daily data exports were failing about four days a month, and the service level agreement sat at 87 percent against a 99 percent target. The meetings about it had stopped being useful. The data the support teams could see did not match what the hotels were experiencing, so the complaints read as anecdotal from one side and the responses read as dismissive from the other, and everyone arrived defensive.

What I did

Rather than bring better anecdotes, I pulled six months of the vendor’s own daily performance reports and worked through the numbers over a weekend. The bet was simple: hard data from their own systems would land where complaints had not.

It did. They validated the analysis and confirmed the issue was real. Just as important, nobody had to lose face to get there, because the data was theirs. From that point the work turned collaborative: we root-caused the failures together and put proactive health checks and alerting in place so problems were caught before hotels felt them.

What happened

The service level reached the 99 percent target. Better still, customers began praising the proactive alerts that warned them before they noticed anything wrong. A defensive standoff had become a partnership, and the analysis that got us there cost a weekend.