Customer Story

How The Johnstown Collection went from two curious users to ten across five departments, and found revenue nobody was looking for.

AI adoption in a hotel rarely starts with a transformation programme. At The Johnstown Estate it started with two people trying to answer practical questions faster.

The estate has an unusually broad operating mix for an independent property: hotel bedrooms, self-catering lodges, a destination spa, weddings, sports teams and a large conference and events operation. With Trim Castle Hotel, one central commercial team is also looking across two properties.

The data existed. The inefficiency sat in the work required to reach it. Opening reports, exporting files, building helper tabs, checking totals, and repeating the same process the following morning.

Life before Otel: manual, workable, and repeated every day

The daily pickup process began with a CSV export from Opera, pasted over the previous day's tab in Excel and recalculated. Seán Reid estimated it took roughly 40 to 45 minutes per property, every morning. For Johnstown and for Trim.

The weekly forecast required separate Opera reports for hotel rooms, lodges and the combined position, followed by a helper tab to join half-month periods. Group blocks and event movements were tracked in a manual document.

The process worked. It also consumed skilled time and created room for inconsistency and human error.

It started with two users. The value spread through the hotel

In December 2025 it was just two users, asking questions. Adoption then expanded as individual teams found work that mattered to them. There was no central instruction to use AI. It was a response to useful answers, reports and controls.

By August, 23 recurring reports, alerts and control checks were running in the background. Competitor-rate checks, new block alerts, forward views, rooming-list checks, wedding and sports reports, GM briefings and low-inventory alerts.

The questions got broader as adoption grew

The clearest evidence of embedding is not a log-in count. It is the range of real questions different teams now bring to the same connected hotel data.

Part of the reason the range is so wide is the operation itself. A wedding, a spa package, a corporate rate and a sports block are commercially different problems, and each generates its own questions.

Where the efficiency return comes from

The return is not about replacing people. It is capacity returned to revenue, reservations and management teams, plus faster intervention when something changes.

The daily pickup saving was one of the earliest wins rather than the ceiling. As more systems are connected and new use cases are developed, the efficiency gain is expected to increase significantly.

• Recurring reporting. New reports and controls are validated against source systems before they become part of operational reporting, then delivered on the required schedule.
• Ad hoc analysis. A question that previously required several exports can be answered while the meeting or operational decision is still live.
• Cross-system context. One question can combine PMS restrictions, RMS floor rates and rate-shopping market information.
• Exception monitoring. Teams do not inspect every record to find the small number of dates, blocks or reservations that need attention.

Rate decisions previously meant moving between systems: one place for historical and market analysis, the PMS for restrictions, the RMS for floor rates, rate shopping for current market position. With those sources connected, the team asks one question and receives the full context together.

The same principle applies beyond revenue. A reservations check can move from a high-level exception down to the individual booking. An events report can show that a function moved from 100 to 128 attendees and quantify the revenue effect. A GM brief can bring last night’s performance, today’s arrivals and the most important forward risks into one view.

Answers arrive while the decision is still being made

One of the most valuable benefits is the ability to answer questions while the discussion is actually taking place. Traditionally, a question arising in a meeting that fell outside standard reporting meant somebody saying they would have to come back to you on that. Pull the reports afterwards, reconcile the data, return later with an answer. By that point the meeting is long over, and the opportunity to influence the decision has passed.

Seán described a meeting where a weak ADR was questioned. Instead of leaving the room to open and reconcile several reports, he asked what was dragging the rate down and had the answer in about ten seconds. The same now applies to pickup trends, pacing concerns and commercial anomalies, investigated live, so decisions are made more quickly and on real data.

The commercial opportunities report is now circulated in advance of the weekly sales, marketing and revenue meetings for Johnstown and Trim. Its findings are used to shape revenue actions, marketing campaigns and sales priorities.

The part that wasn't on any report schedule

Some of the most valuable work was not on a daily or weekly schedule at all. Seán asked Otel to review the estate's GDS activity, agency records and account profiles. The kind of housekeeping exercise that rarely reaches the top of anyone's list in a busy hotel.

• Duplicate profiles. The same account appearing more than once, so its full value was harder to see in one place.
• Incomplete records. Bookings arriving without a company or agency attached, making corporate production harder to attribute.

These are the kinds of things that accumulate quietly in any hotel database, in any system, as bookings arrive through multiple channels over years. They are rarely anyone’s fault, and almost never anyone’s job.

The value was in turning them from a general sense that the data could be tidier into a defined piece of work, with a clear scope, for the revenue, reservations and sales teams. This is a different form of efficiency: not doing an existing report faster, but making visible the work that never had a schedule, and giving it an owner.

From efficiency to commercial opportunity

While efficiency was the first measurable benefit, the longer-term value has been commercial. Otel has helped identify opportunities and issues that were previously hidden inside large volumes of operational data.

• Revenue leakage identified through upgrade activity, rooms given away that could be recovered and resold.
• Earlier visibility of block pickup changes, and what each change is worth.
• GDS housekeeping opportunities that had not previously surfaced.
• Duplicate account profiles and missing company or agency attribution.

That single question took two minutes and put sellable inventory back on the market. Nobody had asked it before, because answering it by hand meant opening bookings one at a time.

A new way of working for an independent hotel group

Large hotel brands can put internal analysts, data teams and engineers behind bespoke reporting and automation. Independent hotels rarely have the same in-house resource, even though their operating complexity is no smaller.

At Johnstown, the person closest to the work describes the report, alert, audit or analysis they need in normal language. The hotel team keeps the judgement and decides the action. Otel removes the technical barrier, and the repeated assembly work around it.

What the story now says

1. Adoption grew from 2 direct users to 10 as the product became useful to more hotel functions, pulled through by teams rather than pushed by a mandate.
2. A daily reporting process is estimated to return around 1.5 hours of team capacity across two properties.
3. Some live questions move from multiple reports and 30 to 45 minutes of work to an answer in seconds or minutes.
4. Recurring controls apply the same logic consistently and surface exceptions before they disappear into manual trackers.
5. The team uses Otel for known recurring work and for investigations that previously had no owner or schedule.
6. Beyond time saved, the platform has surfaced commercial opportunities and data quality issues, including duplicate accounts, missing attribution and upgrade leakage, that nobody was actively looking for.