# Otelier's answer to a 1% hotel profit forecast starts with the night audit

By MarketScale Newsroom · Published 2026-11-19 · Hospitality on MarketScale
Canonical: https://www.marketscale.com/industries/hospitality/oteliers-answer-to-a-1-hotel-profit-forecast-starts-with-the-night-audit

> With U.S. hotel GOPPAR forecast to grow 1% in 2027, a new Otelier report looks at which back-office tasks are worth automating now.

## Key points

- Otelier's report covers which accounting tasks are ready to hand off today, from night audit to bank, credit card and OTA reconciliation. One reason they may suit software is that the records either match or they don't.
- With profit per room forecast to grow 1% in 2027, the hours spent reconciling bank, card and OTA statements are among the few costs a hotel can cut without a guest ever noticing.
- Before an AI agent is allowed to place orders or approve invoices, a hotel needs written answers on how much it can spend, when it needs approval and who owns a wrong decision.

U.S. hotels have their revenue back. Holding on to profit is harder. CoStar and Tourism Economics expect gross operating profit per available room, or GOPPAR, to grow just 1% in 2027, because labor, insurance and energy costs keep rising faster than inflation.

Hospitality data platform Otelier cites that forecast in announcing a new report, "A Back Office That Runs Itself: Automating Hotel Operations in 2027." The argument is blunt: revenue growth will not deliver the next round of margin gains. Those gains will instead be found in work that guests never see, such as night audits, reconciliations, reports and the handoffs between a property and corporate.

The report draws on recent conversations with those currently running hotel finance, operations and technology. Featured voices include Jan Freitag of CoStar Group, Crescent Hotels & Resorts CFO Rob Smith, MCR Hotels' Kyle McCaig and Mandarin Oriental's Todd Wood. Its roadmap for back-office automation is simple: begin small, prove the hours and build from there.

The report is aimed at readers who run operations across a growing portfolio or who oversee a management company's finances. A recent LODGING Magazine piece makes the same point about pacing: hand software limited responsibilities at first, then widen what it can do on its own as trust grows.

Block Field

## Why reconciliation goes first

The report identifies the accounting tasks that are ready to be handed off now, ranging from night audit to bank, credit card and OTA reconciliation. These tasks seem to have one thing in common: each has a clear right answer. Either two records agree or they don't, which suggests exceptions could be simple to flag.

Chip Fritsch, COO and co-founder of Hotel Investor Apps, laid out the finished version of this work in a Hotel Management column in March. In his picture, the property management system closes the day, AI reconciles every transaction, posts the journal entries and sends leadership a plain-language daily report before 7 am. Bank reconciliation works the same way: pull the daily feed, match each transaction against the general ledger, produce a clean rec. He also noted that much of the groundwork is already common, including end-to-end accounts payable, smart bank reconciliation, daily PMS reconciliation and automated approval workflows.

His long-term accounts payable picture follows the same logic. Vendor invoices are extracted, matched against purchase orders and contracted rates, coded to the correct GL account and queued for payment, with exceptions flagged for human review. Every one of these jobs ends in either a match or an exception, which is exactly the trait that puts them at the front of the queue.

Block Field

Vendor products shown this year are more cautious than that end state, and the difference tells a buyer something. While showcasing at HITEC 2026 in San Antonio in June, Inn-Flow listed Daily Activity 2.0 among its recent accounting enhancements, saying it streamlines the review and approval of nightly audit data, according to Hospitality Technology. Reimbursements 2.0 supports inter-company and inter-property expense allocation while improving traceability and reconciliation. In both cases, software seems to prepare the work for a person to review and approve. That could be the realistic 2027 version of a self-running night audit.

With profit per room forecast to grow 1% in 2027, the hours spent reconciling bank, card and OTA statements are among the few costs a hotel can cut without a guest ever noticing. They are also the easiest place to follow Otelier's advice to prove the hours, because the task happens every day and the result is easy to check: the records matched, or they produced an exception.

## From monthly reports to early alerts

Otelier's second theme is report overload. The report describes operators cutting the stack of reports that reach general managers and regional leaders down to fewer, sharper signals. It also says hotel business intelligence is moving from dashboards to answers, which only works on a solid data foundation.

Inn-Flow's AI Insights offers one example of what a sharper signal could look like. Its initial capabilities include three alerts. Over Budget Alerts identify expense categories trending beyond budget before month-end. Labor Costs at Risk combines actual punches and schedules to project labor spend before payroll closes. Missing Expenses identifies recurring invoices or accruals that have not yet been received. The alerts are built on each property's own operating patterns and data.

A report tells a GM about a month that's already over. Alerts like these could still leave time to do something.

The answers stage goes one step further. Fritsch's example is a CFO preparing for a board meeting who asks an AI assistant why fourth-quarter EBITDA came in 8% under forecast, then has it pull ERP actuals, compare them to budget, flag properties where labor ran above 35% of revenue and cross-check PMS occupancy. He expects the Model Context Protocol, or MCP, which links an outside AI assistant directly to financial databases, to turn that multi-hour, multi-login analysis into a three-minute conversation. The catch is Otelier's point about foundations. That question only works if the ERP, labor and PMS data already line up.

Fritsch's nearer-term examples are humbler: asking an in-system assistant to list the 10 most recent open invoices, pull last year's room revenue or name the departmental expenses with the highest increases year to date. Each of those questions lives inside one system. That suggests single-system questions are the realistic starting point, while cross-system answers wait on the data work.

The LODGING author explains why that data work is hard. Hotels don't run on one system but on dozens, spanning property management, accounting, procurement, inventory, payroll, point of sale, banking and forecasting, and many of them don't talk to each other. "The software stores the information. The human does the work," the author writes. Inn-Flow Chief Product Officer Lee Bridges made a similar diagnosis at HITEC, saying "too many teams are still working across disconnected systems and spreadsheets."

Block Field

## When software holds the company card

The third step changes the job. A hospitality technology provider writing in LODGING Magazine this month described a 200-room hotel facing 92% occupancy tomorrow, with a large group arriving, higher breakfast demand expected and the kitchen short on several items. Today a person connects those dots. In the author's scenario, an AI agent reads the forecast, checks inventory and past consumption, compares supplier prices and places the orders itself, within the hotel's purchasing policies, budget limits and approval rules.

The chef still owns the menu and the quality but gets back the 45 minutes the author says go into moving information between systems. Invoices follow the same pattern: match the purchase order and receiving records, assign GL codes, approve automatically when everything fits policy, and ask a person when it doesn't.

The author is frank that trust is harder than the AI. A recommendation is one thing. An agent that buys $15,000 of supplies, codes expenses to the wrong accounts or approves an invoice it should have rejected is another. In a hotel, a missed delivery can affect breakfast service and an incorrect payment can create accounting problems. The piece lists the questions a hotel should answer before it lets an agent transact:

- What the agent is allowed to do
- How much money it can spend
- When it needs approval
- What happens when information is missing or conflicting
- Who is responsible when something goes wrong
- Whether the hotel can explain exactly why the agent made a given decision

The author's suggested model is to treat agents like new hires: no unlimited spending authority on day one, defined responsibilities, monitored results and authority that grows as trust is earned. The stated goal isn't to remove people from every decision but to keep them involved in the decisions that actually need them.

Inn-Flow Procurement, the result of Inn-Flow's acquisition of hotel procurement technology provider Lilo announced in early May, appears to stop just short of that line. Its AI cart optimization compares pricing across vendors and recommends lower-cost alternatives, and budget visibility and policy controls apply before purchases are made. Its AI-assisted inventory module, previewed at HITEC and slated at the time for broader availability at the end of the third quarter, offers reorder suggestions that push directly to cart. That suggests software fills the cart while a person still places the order.

For a management company, those questions belong in purchasing policy now, even if no agent will spend a dollar this year. The vendor tools already enforce budget and policy limits, and those limits need to be written down before they can be enforced.

## What stays with the finance team

Even the most optimistic roadmap here doesn't promise a back office without people. Fritsch argues that AI doesn't solve accountability, because someone still has to stand behind the numbers. "The back office isn't disappearing. It's being elevated," he writes. He says the next three to seven years may bring some of these end-to-end automations to early adopters. Read that way, Otelier's title describes a direction. For operators planning 2027, that suggests the matching jobs are the likeliest parts to run on their own.

The same line shows up on the commercial side. A guest blog on RoomPriceGenie's site says routine, data-based work such as daily price adjustments, offer follow-ups and pre-arrival emails should run automatically in 2027. It says hotels should keep their pricing strategy, their most important guests and everything that happens on site. It also cites RoomPriceGenie's 2026 report, which found 66% of hotels in the DACH region struggle to hire qualified staff and only 46% have a dedicated pricing role. For hotels where one GM covers several jobs, recovered hours count more than they would at a fully staffed property.

Its advice on handing over pricing mirrors the new-hire model. Set minimum and maximum prices once, then watch the system's recommendations for a week before letting them apply automatically. Pricing and purchasing are different jobs, but the sequence is the same: fixed limits first, observed results second, autonomy last.

For finance teams setting 2027 budgets, that points to a likely order. Start by measuring how many hours reconciliation and report-building take today, then automate the tasks that have a clean right answer. Spending limits for agents could come later. Otelier's report also looks at what still needs a person and how teams change when the busywork goes away, and its featured voices include Kelly Baker, Accounting & Operations Training Manager at Stonebridge. Fritsch describes finance teams moving from receiving help from AI to supervising it as it runs reporting and workflows. The finance staff who used to do the matching could become the ones checking the exceptions.

## Sources

- [A Back Office That Runs Itself: Automating Hotel Operations in 2027](https://www.hospitalitynet.org/whitepaper/4135554/a-back-office-that-runs-itself-automating-hotel-operations-in-2027) (Hospitality Net)
- [How AI can change the hotel back office | Hotel Management](https://www.hotelmanagement.net/tech/assist-automate-elevate-ai-revolution-hospitality-back-office) (Hotel Management)
- [Inn-Flow Introduces Expanded AI and Automation Platform ...](https://hospitalitytech.com/inn-flow-introduces-expanded-ai-and-automation-platform-capabilities-hitec-2026) (Hospitality Technology)
- [The Day When AI Runs Hotel Operations Is Just Around the Corner](https://lodgingmagazine.com/the-day-when-ai-runs-hotel-operations-is-just-around-the-corner/) (LODGING Magazine)
- [Hotel Automation 2027: What Should Still Be Done Manually?](https://roompricegenie.com/hotel-automation-tasks-to-automate/)

Tags: Otelier, Inn-Flow, Hotel back-office automation, Night audit, GOPPAR, AI agents

---
Source: MarketScale, https://www.marketscale.com/industries/hospitality/oteliers-answer-to-a-1-hotel-profit-forecast-starts-with-the-night-audit. Published for AI indexing and citation; cite the canonical URL. Site guide for agents: https://www.marketscale.com/llms.txt
