Hospitality · Integration · Marketing · Anonymized at the client’s request
A 210-room Gulf Coast resort
Joined PMS stay history to the campaign platform, then rebuilt pre-arrival and winback on behaviour instead of send date.
- direct booking share
- +14%
- OTA commission paid
- −9%
- email revenue per send
- 2.4×
Measured across two quarters against the same quarters the prior year, adjusted for rate.
Situation
Direct bookings had been flat at 28% of room nights for three years while OTA commission grew with rate. The property had a modern PMS and a capable email platform that had never been connected, so every campaign segment was built by hand from a monthly spreadsheet export that was already stale when it arrived.
Constraint
No PMS replacement was possible; the property was mid-contract. Ownership required that the booking engine remain untouched during high season, which left a fourteen-week window. Marketing was one person.
What we built
- A nightly sync from the PMS into the CRM covering stay history, folio spend, and rate plan, with a reconciliation report
- A canonical guest record joining booking-engine, PMS, and email identities on email plus phone with manual review of near-matches
- Pre-arrival, on-property, and post-stay sequences segmented on prior spend and stay length rather than on booking date
- A winback flow that excluded guests who had complained, which the previous list-based approach had been mailing repeatedly
- A revenue dashboard reading net of commission, so direct and OTA bookings were finally comparable
Stack used
- Opera
- SynXis
- HubSpot
- Postgres
- Looker Studio
Duration
14 weeks
Team
Two engineers, one marketing lead, part-time
“The useful part was not the campaigns. It was that we could finally answer what a returning guest was worth without three people arguing about it.”
What we’d do differently
We built the full canonical guest record before shipping any campaign, which cost about three weeks. In hindsight we should have shipped the pre-arrival sequence on the simpler email-only match in week two and improved the matching underneath it afterward. The revenue would have started earlier and the matching work would have been identical.
Tell us what isn’t shipping.
Thirty minutes, no deck. Bring the thing that’s stuck and we’ll tell you how we’d approach it, whether or not you hire us.
Or email hello@fusionads.ai · Florida, US