Which cleaner went to which unit, and whether they had finished, was decided and confirmed in group messages. With fifty-plus cleaners and hundreds of daily turnovers, the group chat was both the assignment system and the audit trail, and it was neither.
Six hundred units and fifty cleaners, dispatched without a spreadsheet or a chat group
A flexible-living operator that rented furnished apartments by the night or by the month moved room turnover, cleaning and linen off spreadsheets, a messaging app, a workflow tool and paper, into one system wired to its booking platform.
A flexible-living operator renting furnished apartments by the night or by the month.
Hundreds of units spread across many residential buildings, each one becoming vacant at a different hour, each one needing a cleaner, a linen swap and a readiness flag before the next guest could check in.
Flexible living
600+Units in operation, company figure from 2022
1,500Units under contract, press coverage from June 2022
50+Cleaners dispatched automatically, 2022
40h+Hours a month saved, company figure with no stated basis
600+Units in operationThe company's figure in the 2022 written case, consistent with press coverage from June 2022 (600 operating, 1,500 under contract).
50+Cleaners receiving assignments automaticallyThe company's figure in the 2022 written case.
40h+Hours a month savedThe 2022 written case's figure. No basis or method is stated for it.
66+EmployeesThe 2022 written case's figure. Press coverage four months later put the team at about 160; the case reports both and does not choose.
“At first, I was skeptical about how Jestor would handle complex tasks. However, it exceeded my expectations. The technology offers many possibilities and it's fast to build new integrations and other tech processes.”
Nino PalermoHead of Product and Tech, Nomah, in the 2022 published case
BeforeAfter
- A spreadsheet listing the units to be cleaned, updated by handCleaning tasks created from the booking platform's check-out data
- Group messages assigning cleaners and confirming completionEach cleaner receiving their own units and marking them done there
- A generic workflow tool holding the cleaning pipelineThe cleaning flow tied to the unit and to the stay
- Shared documents and paper for lost items and damageLost items and broken objects recorded in the cleaning task itself
- Pen-and-paper linen counts with laundry suppliersLinen tracked per unit and per movement, beside the cleanings
- Someone updating the booking platform that a unit was readyUnit readiness written back to Guesty automatically
- Manual reporting of what each cleaner had donePerformance data generated from completed tasks, fed to the data lake
Guesty stayed as the property management system: bookings, channels, check-in and check-out. Jestor sat between it and the cleaners. The data lake stayed as the analytics layer. Laundry suppliers stayed external. The cleaners were the same team; what changed was how work reached them and how completion was recorded.
The bottleneck was operational, not commercial
Nomah operated furnished apartments in residential buildings, rented for a single night or for months at a time, with online check-in, cleaning, maintenance and concierge handled by the operator. It managed units on behalf of individual investors and signed whole buildings with developers. In the first half of 2022 it reported 600 units in operation and 1,500 under contract, with about 90 percent of the portfolio in one city. Loft, its parent company, had put in a capital injection in 2021 for a three-year plan.
Nomah no longer operates. In August 2022 it merged with a regional short-term rental operator; the merged company shut down its operations in the country in January 2023 and dissolved entirely in August 2023. This case describes the system as it worked in 2022 and is written in the past tense for that reason.
The model puts the pressure squarely on turnover. A hotel has its housekeeping team in one building. A flexible-living operator has hundreds of units spread across many buildings, each one becoming vacant at a different time, each one needing a cleaner to arrive, a linen set to be swapped, anything broken to be logged, and the unit to be marked ready in the booking system before the next guest can check in. With bookings that could be one night long, the gap between check-out and check-in was the operation.
The bottleneck was operational, not commercial. The company was signing buildings faster than it could open them. What limited it was that six hundred units' worth of turnovers were being coordinated through a spreadsheet and a group chat.
The turnover had no single record
The written case summarizes the starting point in one line: the company replaced Google Sheets, WhatsApp and Pipefy with Jestor. Behind that line are three failures.
The booking platform knew when a guest left. The spreadsheet knew which cleaner was assigned. The chat knew whether the cleaning was done. Nobody's system knew all three, so "is this unit ready" was a question someone had to answer by hand, for hundreds of clients needing the answer in real time.
Towels and sheets moved between units, cleaners and laundry suppliers with pen-and-paper counts. Losses and shortages showed up as a guest complaint, not as a number.
The pattern behind all three is that the unit turnover, the one event every team and every guest depended on, had no single record. It was reconstructed after the fact from messages.
Item by item, what did each job before and what did it after
| What did this before | What did it after |
|---|---|
| A spreadsheet listing units to be cleaned, updated by hand from the booking platform | Cleaning tasks created in Jestor from the booking platform's check-out data |
| Group messages assigning cleaners and confirming completion | Each cleaner receiving their own units in Jestor, marking them done there |
| A generic workflow tool for the cleaning pipeline | The cleaning flow in Jestor, tied to the unit and the stay |
| Shared documents and paper for lost items and damage | Lost items and broken objects recorded in the cleaning task itself |
| Pen-and-paper linen counts with laundry suppliers | Linen tracked per unit and per movement in Jestor |
| Someone updating the booking platform that a unit was ready | Unit readiness written back to Guesty from Jestor |
| Manual reporting of cleaner output | Performance data generated from completed tasks, fed to the data lake |
What was not replaced. Guesty stayed as the property management system: bookings, channels, check-in and check-out. Jestor sat between it and the cleaners. The data lake stayed as the analytics layer. Laundry suppliers stayed external. The cleaners themselves were the same team; what changed was how work reached them and how completion was recorded.
The unit turnover, end to end
- 1A guest checks outExternal
The booking platform records the unit as vacated.
- 2The cleaning task is createdAutomated
Jestor picks up the vacated unit and creates the cleaning task.
- 3The task reaches a cleanerAutomated
It is assigned to a cleaner, who sees it in their own view.
- 4The unit is turned overPerson
The cleaner turns over the unit, swaps linen, and records lost items, damage and linen movement in the task.
- 5The cleaner marks it donePerson
The unit is closed off in the task, by the person who did the work.
- 6Readiness goes back to the booking platformAutomated
Jestor updates unit readiness in Guesty, so the unit is available for the next check-in.
- 7Performance data is generatedAutomated
Data per cleaner and per unit is produced and sent to the data lake.
- 8Operations reviews itPerson
Performance, losses and linen are reviewed from that data.
The hard links were 2 and 6, the two ends of the integration with the booking platform. Step 2 only works if every check-out becomes a task, with the right unit and the right time, or cleaners are dispatched to units still occupied and miss units that are empty. Step 6 is what the guest feels: if readiness is not written back reliably, the booking platform sells a unit that is not clean or holds one that is. Head of product Nino Palermo said he was skeptical at the start about how the system would handle the complexity of these flows and that it exceeded his expectations; the complexity is exactly in those two steps.
Every figure, with the basis beside it
| Indicator | Result | Where it comes from |
|---|---|---|
| Units in operation | 600 and more | 2022 written case; press coverage from June 2022 gives the same figure. |
| Units under contract | 1,500 | Press coverage from June 2022, quoting the company's founder. |
| Cleaners dispatched automatically | 50 and more | 2022 written case. |
| Hours saved per month | 40 and more | 2022 written case. No method, baseline or breakdown is given. |
| Team size | 66 and more (April 2022); about 160 (August 2022, before a reduction of about 30) | Written case and press coverage respectively. The two figures are four months apart and may count different populations. |
| Linen tracking | Moved from paper to the system | Written case; no loss figures before or after were reported. |
"40 hours a month" has no stated basis. It appears in the written case's summary line and nowhere else. Unlike other cases in this series, there is no interview describing where the hours came from, so the figure is reported as the company's claim and nothing is derived from it.
The unit count is the most solid number here. It is the one figure stated by the company in the case and repeated to the press within weeks, with the larger contracted pipeline alongside it.
The two team-size figures do not reconcile. 66 and more in April 2022 against about 160 in August 2022. The written case's figure may exclude cleaners or count a different entity. The case reports both.
What this case does not measure
Worth naming, because it is usually what gets inflated.
Nomah merged with another operator in August 2022, and the merged company ceased operations in January 2023. Nothing here describes a running system today, and the case does not claim the operating system played any role in the outcome either way.
Units, cleaners, employees and hours saved come from the written case published in April 2022 and from press coverage of the company that year. None is audited.
See above. It is the company's claim.
The company did not report what a unit turnover cost before or after.
The linen flow moved from paper to the system; no before-and-after loss rate was reported.
The case says performance data became automatic; no turnover time or units per cleaner per day was published.
The 2022 case gives a founding year and a funding figure that belong to, or are converted from, the parent company's data. This case omits both.
Unlike other cases in this series, there is no video interview for Nomah. Quotes are limited to what the written case published.
Housekeeping is where the labor gap is
| Indicator | Figure | Source |
|---|---|---|
| Room attendant time per occupied room | 24.61 minutes in 2025 | HotelData and Actabl, Q4 2025 Hotel Labor Costs Report, March 2026 (vendor, about 5,000 US hotels) |
| Total labor cost per occupied room | 48.32 USD in 2025, up 12.8 percent on 2024 | Same report |
| Hotels reporting staffing shortages | 65 percent; housekeeping the most-cited gap at 38 percent | AHLA and Hireology survey, January 2025 (282 hoteliers) |
| Median wage, maids and housekeeping cleaners | 17.07 USD per hour; 860,670 employed | US Bureau of Labor Statistics, OEWS May 2025 |
| Linen lost or discarded before end of life | 15 to 20 percent | TRSA via Hotel Executive, January 2015 (industry association; widely re-cited without its date) |
| Tech-enabled operator at scale | About 9,000 rooms across 140 properties at the time of its liquidation filing | Sonder, Chapter 7 filing, November 2025 |
| Tech-enabled operator at scale | About 15,000 apartments in 32 markets; 560 million USD revenue in 2023 | Blueground, Series D announcement, March 2024 (self-reported) |
Two thirds of hotels report being short-staffed and housekeeping is the department they name first. For a dispersed operation with fifty cleaners across hundreds of units, the constraint is not only finding cleaners but not wasting the ones you have on units that are not ready or already done, which is what a dispatch system built on check-out data fixes.
The 15 to 20 percent loss figure comes from a 2015 article by a laundry industry association and has been recycled by suppliers since without its date. There is no recent independent study. It is included as an order of magnitude only, and it is the reason a paper linen count is a cost, not a detail.
The two best-known operators in this category ended 2025 in opposite places: one in liquidation with 9,000 rooms, the other reporting 560 million USD in revenue. Nomah's own story ended in a shutdown in 2023. Turnover mechanics are not what decides these outcomes, but they are what the outcome is built on: a unit that is not ready is revenue not earned on a lease already paid.
The system is custom-built and stays our responsibility
Nomah's product team built its turnover system in a short time and wired it into the booking platform themselves. When the company merged and then wound down, the system went with it. That is the year-two question every custom system faces in some form: who owns it when the people who built it are reorganized, reassigned or gone, because a system nobody can touch becomes a spreadsheet again within three years.
One builder owns each request end to end, with one always in progress.
A new flow, a new automation or a new report enters through the same channel, without becoming a new project.
If what was built is not right, it is rebuilt. Unlimited revisions within the subscription.
Seats are never the billing unit, which matters when fifty cleaners, laundry suppliers, operations staff and product all touch the same operation.
People learn to use their app the way they learn any app, by opening it. Building, configuring and maintaining stays on our side.
Full export at any time, by CSV and API. SOC 2 compliant, no exit fee. Pause in one click and the systems keep running.
Methodology and sources
Units in operation, cleaners, employees, hours saved, the tools replaced, the cleaning and linen flows and the integration with the booking platform come from the written case published in April 2022 (updated March 2025) and the quote from its head of product and technology, Nino Palermo, whose title in 2022 is confirmed by a public organization directory. A second quote in the written case is attributed to a product manager whose name and title could not be confirmed in any other public source and is not used here. There is no video interview for this company.
The company's model, its 2021 capital injection, the 600 operating and 1,500 contracted units, the team size in August 2022, the merger in August 2022 and the shutdown in January and August 2023 come from press coverage between 2020 and 2025. The founding year and funding figure stated in the 2022 written case are omitted because press coverage attributes them to, or derives them from, the parent company.
HotelData and Actabl Q4 2025 Hotel Labor Costs Report (March 2026); AHLA and Hireology staffing survey (January 2025); US Bureau of Labor Statistics OEWS (May 2025); TRSA via Hotel Executive (January 2015); Sonder Chapter 7 filing coverage (November 2025); Blueground Series D coverage (March 2024). Vendor and association sources are labeled in the table.
The 2022 case's founding year and funding figure, for the reason above. The product manager's quote. Any derived per-cleaner or per-unit productivity number, because no baseline exists. Any claim linking the operations system to the company's later outcome, in either direction.
Nomah started as Uotel in 2016 and was acquired by Loft in 2020; in March 2021 it was relaunched as Nomah with a capital injection from Loft for a three-year plan, and the case Jestor published in 2022 described the company as recently acquired by Loft. In August 2022 Loft transferred its Nomah shares into the merger with Casai and became the largest shareholder of the combined company, which ceased operations in the country in January 2023 and was dissolved in August 2023. The operation described here ended with it. Loft itself is still a Jestor customer, with other teams and other processes: what shut down was Nomah, not the relationship.