A compiled reference of 33 verified statistics on the open-source software available to cleaning companies, covering what a GitHub search actually returns, the condition of the purpose-built projects, the general field-service platforms that a residential or janitorial operation can genuinely run today, and what every one of them leaves uncovered. Every figure was measured first-hand in August 2026 against the GitHub and Docker Hub APIs or produced by install tests in clean Docker environments, not taken from a software directory or a vendor roundup.
Key statistics
6 highlights from this report
Key statistics
Key takeaways
There is no maintained open-source product built for running a cleaning company. What exists instead is a general field-service module set for Odoo that installs cleanly, schedules recurring visits properly, and has no concept of a photo. Anyone choosing between self-hosting and a commercial subscription is really choosing between building the verification layer themselves or buying one.
The 57 repos offering cleaning, maid or janitorial business software hold 54 GitHub stars combined.
The most-starred open-source cleaning business system has 7 stars and died in January 2016.
Of 11 audited repos, one ships a container build path, and it still failed to start.
The maintained field-service option has more forks than stars: 301 forks to 193 stars.
All 35 tables of the leading open-source field-service stack contain zero photo, checklist or inspection fields.
88% of cleaning-business repos carry no license, and 67% have not been touched in a year.
How we built this report
Every figure on this page was produced in August 2026 by running commands, not by reading a directory listing. Nothing here is quoted from a software comparison site.
- Swept the field, then the adjacent field
Fifteen GitHub repository searches on cleaning, janitorial, maid, and housekeeping management terms returned 555 results across 528 unique repositories. A second sweep of twelve field-service, dispatch, and work-order queries covered the general platforms a cleaning company might adopt instead.
- Triaged by hand, not by keyword alone
A pattern match narrowed 528 repos to 219 cleaning-domain candidates, split into laundry, hotel housekeeping, and cleaning-service-business segments. We then read all 66 rows of the cleaning-business segment individually and disclose the false positives rather than quietly keeping them in the count.
- Measured each project directly
Stars, forks, licenses, archive status, contributor counts, and 52-week commit participation came from per-repository GitHub API calls. Image reach came from the Docker Hub registry API. No figure was taken secondhand.
- Install-tested in clean environments
Every credible candidate was cloned and installed from scratch on Docker 28.3.2, following each project's own documentation. Results are reported as they happened, including the errors.
- Independent review
Written by one co-founder, reviewed by the other before publishing.
Scope caveat: GitHub search is not a census of all open-source software. A project with an unusual name, hosted on GitLab or SourceForge, or described in a language other than English could sit outside this sweep, and star counts measure visibility rather than quality. The install results describe what happened on one machine on one day with each project's documented path, not a guarantee about every environment. The two halves of this report also carry different collection dates: the per-project metrics (stars, forks, contributors, commit activity, image pulls) were re-verified on August 10, 2026, while the 528-repository sweep, the segment classification, and every install test were run between August 1 and August 4, 2026. Figures move; each date on this page is the date those numbers were true.
The open-source cleaning software field, by the numbers
All 33 figures, grouped by theme, each measured first-hand in August 2026 and independently reviewed before publishing.
What a search actually returns
According to our own sweep of the GitHub repository search API (github.com), fifteen queries covering cleaning service management, janitorial software, maid service, and housekeeping systems returned 555 results across 528 unique repositories. Pattern matching narrowed that to 219 cleaning-domain candidates, which split cleanly into three different businesses: laundry and dry-cleaning shops, hotel housekeeping departments, and actual cleaning-service companies. That last segment matched 66 repositories, but reading all 66 by hand found nine that the pattern caught wrongly: waste-collection hardware, a sanitation platform, and developer tooling that happens to use the word cleaning. Removing them leaves 57 repositories holding 54 GitHub stars between all of them, an average of under one star each, and not one of them above seven. Both the raw and adjusted counts, and the reason for every exclusion, are in the segment file linked below.
Star mass, to scale
GitHub stars on one true linear scale, with no minimum bar width. The bottom two bars are drawn correctly; there is simply almost nothing to draw.
True scale Odoo = full width
The same bottom two bars rescaled so the larger one fills the width, a 278x magnification
Every open-source cleaning-business project on GitHub, added together, holds roughly one thousandth of the stars of a single general-purpose business platform, and less than a third of the stars on one community module set. This is the finding the rest of the page follows from: the category was never built, so the practical question is not which cleaning product to self-host but which general platform to bend.
Download the full 66-row segment file (TSV)Statistic 1
15 GitHub search queries returned 555 results across 528 unique repositories.
RapidEye Research, GitHub repository search API (2026)
Statistic 2
219 of the 528 repositories matched the cleaning domain, across laundry, hotel housekeeping, and cleaning-service-business segments.
RapidEye Research (2026)
Statistic 3
The cleaning-service-business segment holds 54 GitHub stars across 57 repositories, an average of 0.95 stars each.
RapidEye Research, GitHub API (2026)
Statistic 4
Not one repository in the segment reaches 10 stars. The pattern match raised 66 repositories and 67 stars; hand review removed 9 repositories and 13 stars as false positives.
RapidEye Research (2026)
Statistic 5
The highest-starred cleaning-business system in the segment has 7 stars and has had no commit since January 2016.
RapidEye Research, GitHub API (2026)
Statistic 6
88 percent of the segment (50 of 57 repositories) carries no standard open-source license.
RapidEye Research, GitHub API license field (2026)
Statistic 7
67 percent of the segment (38 of 57) had no commit in the previous 12 months.
RapidEye Research (2026)
Statistic 8
14 of the 57 repositories were created during 2026, and not one of them has more than a single star.
RapidEye Research (2026)
What this means: The absence of a license on 88 percent of the segment matters more than the low star counts. Code published without a license is not open source in any usable sense; default copyright applies, and a cleaning company that adopts it has no right to modify or deploy it. Just 7 of the 57 carry a standard license, and the most-starred of those has been untouched since 2016, which is why the practical size of this category rounds to nothing.
The purpose-built projects, and their condition
Reading the segment by hand rather than by star count turns up six repositories that genuinely attempt cleaning-business management. Their combined output is a museum. The two most recent, a janitorial bidding tool and an ERPNext-based maid-service platform, were both created in 2026, and both were built with AI coding agents: one ships a lock file pinning a set of third-party agent skills, and the other ships an instructions file stating that it exists so any model, including a smaller or cheaper one, can implement the project safely. Neither has an audience. Of the 11 repositories we audited across this group and the field-service group, a recursive scan of every file found that exactly one ships any container build path at all, a Dockerfile and a compose file, and the other ten ship none. That is the closest available proxy for whether a project expects anyone else to run it, and the one project that clears the bar still failed to start.
Statistic 9
Of 11 audited cleaning and field-service repositories, exactly 1 ships any container build path, and the other 10 ship none.
RapidEye Research, recursive git tree scan (2026)
Statistic 10
The one open-source janitorial bidding and business management project has 1 star, no license, and a lock file pinning third-party AI agent skills in its repository root.
RapidEye Research, GitHub API and source inspection (2026)
Statistic 11
The newest purpose-built maid-service platform, created July 2026, has 1 star and hard-codes a private container registry address in its Dockerfile.
RapidEye Research, source inspection (2026)
Statistic 12
Its image builds at 4GB and its app registers inside the ERPNext bench, but its documented compose quick start fails during site creation and serves nothing.
RapidEye Research install lab (2026)
Statistic 13
The 7-star leader of the segment ships 11 Odoo modules including a domestic cleaning module, but uses the manifest filename Odoo retired in 2016.
RapidEye Research, source inspection (2026)
Statistic 14
Three further cleaning-specific projects are dormant: a cleaning field-service API last pushed December 2022, a commercial cleaner scheduling app last pushed December 2022, and a commercial cleaning task scheduler last pushed February 2012.
RapidEye Research, GitHub API (2026)
What this means: The 2026 arrivals are the interesting part. Agent-assisted development has made it cheap to generate a plausible-looking cleaning platform, and the segment now gains repositories faster than it gains stars: 14 of the 57 were created this year and not one has more than a single star. A scaffold that builds is not a product anyone can operate, and the install lab below is what separates the two.
What a cleaning company can actually run
The working answer comes from the Odoo Community Association (odoo-community.org), the nonprofit that maintains community modules for Odoo (odoo.com). Its field-service repository carries 193 stars against 301 forks, a fork-to-star ratio of 1.56 where Odoo itself sits at 0.62 and ERPNext at 0.33. That inversion is the signature of software people deploy rather than bookmark: you fork an Odoo module set to run and customize it, and you never think to star it. It is AGPL-3.0 licensed, has 98 contributors, ships 30 modules, and targets Odoo 19, the current major version. It recorded 396 commits over the last 52 weeks with 131 of those landing in the final four weeks, and its most recent commit landed on August 4, 2026. It also publishes no tagged releases at all, so installation means tracking a version branch. The two platforms that can host a stack like this are widely deployed: according to the Docker Hub registry API (hub.docker.com), the official Odoo image has been pulled more than 52 million times and the ERPNext image maintained by Frappe (frappe.io) more than 4.8 million, with Dolibarr (dolibarr.org) a lighter third option at 858,449.
Statistic 15
The Odoo Community Association's field-service module set carries 193 stars against 301 forks, a fork-to-star ratio of 1.56.
Odoo Community Association, via GitHub API (2026)
Statistic 16
By comparison, Odoo's own fork-to-star ratio is 0.62 and ERPNext's is 0.33.
RapidEye Research, GitHub API (2026)
Statistic 17
It is AGPL-3.0 licensed, has 98 contributors, and ships 30 modules targeting Odoo 19.
Odoo Community Association, via GitHub API (2026)
Statistic 18
It recorded 396 commits over 52 weeks, with 131 in the final four weeks and activity in 33 of 52 weeks.
GitHub participation statistics API (2026)
Statistic 19
32 pull requests were merged into its Odoo 19 branch in the three months to August 1, 2026, and 58 across all its branches.
GitHub search API (2026)
Statistic 20
It publishes zero tagged releases, so deployment tracks a version branch rather than a release.
GitHub releases API (2026)
Statistic 21
The host platform, Odoo, holds 53,634 stars and its official Docker image has been pulled 52,480,226 times.
GitHub API and Docker Hub registry API (2026)
Statistic 22
ERPNext, the main alternative host, holds 37,924 stars, 279 contributors, 6,806 commits across 52 active weeks, and 4,947,616 Docker pulls.
GitHub API and Docker Hub registry API (2026)
Statistic 23
Dolibarr, a lighter third option, holds 7,502 stars, 277 contributors, 4,807 commits over 52 weeks, and 858,449 Docker pulls.
GitHub API and Docker Hub registry API (2026)
Statistic 24
The equivalent field-service app for ERPNext is far thinner: 15 stars, 4 contributors, 26 commits over 52 weeks, and none in the final four.
RapidEye Research, GitHub API (2026)
Statistic 25
An MIT-licensed booking tool aimed at salons, the nearest adjacent trade because appointment-based service businesses share the same booking primitive, has 19 stars and 10 commits in its entire history.
RapidEye Research, GitHub API (2026)
What this means: A cleaning company adopting open source is not adopting cleaning software. It is adopting Odoo, then adding a community module set on top, then accepting an annual major-version upgrade cycle and a project that ships no releases. That is a real and defensible choice for an operator with technical staff. It is not a drop-in replacement for a subscription.
The install lab, and what nothing covers
We installed every credible candidate from scratch on Docker 28.3.2, following each project's own documentation. The field-service module set passed cleanly into the official Odoo 19 image with no manual fixes, and a second pass added recurring visits, routes, timesheets, worker skills, and a customer portal without a single error. That produced 35 tables. We then queried all 35 for any column relating to photos, images, attachments, inspections, checklists, or quality verification and found none. Across all 30 modules, the closest capability is customer signature capture. For contrast, ZenMaid (zenmaid.com) lists $19 per month for a Starter tier capped at 40 appointments, and Jobber (getjobber.com) places job documentation checklists on its Connect plan at $139 per month, or $99 billed annually.
odoo -d fsmtest -i fieldservice --stop-after-init
Installed clean into the official odoo:19 image with zero manual fixes. Second pass added recurring, route, timesheet, skill and portal modules: 63 modules loaded, 0 errors, 35 fsm tables created.
pnpm install && pnpm run build
Installed in 29.7 seconds and built in 1.15 seconds across 3,506 modules, by documentation alone. Ships an agent mode for AI coding tools, but has 9 commits total.
cd deploy/local && docker compose up -d
Its image builds at 4GB and the app registers inside the bench. But the documented quick start died at site creation on ModuleNotFoundError: No module named 'erpnexteasy_maid', two app names concatenated. Databases came up; no site, nothing served.
mariadb librewo < database/schema.sql
The shipped schema threw 9 duplicate primary key errors on MariaDB 11, one per table. All 9 tables were still created, since each CREATE TABLE already declares its key. Documented for repair shops, not cleaning.
Capability coverage
The 30 modules in the leading open-source field-service set, grouped by what they cover.
Scheduling and dispatch, well covered 11 modules
Money, sales and admin, well covered 10 modules
Assets, logistics and interface 8 modules
Proof the visit happened 1 module
Condition and quality verification, not covered 0 modules
Closest available capability across all 30 modules: customer signature capture. A query across all 35 installed tables returned zero columns matching photo, image, attachment, inspection, checklist, quality or verification.
Statistic 26
The field-service module set installed cleanly into the official Odoo 19 image with zero manual fixes.
RapidEye Research install lab, Docker 28.3.2 (2026)
Statistic 27
Seven modules installed across two passes produced 35 tables, with 63 modules loaded and zero errors.
RapidEye Research install lab (2026)
Statistic 28
The resulting order table carries 38 columns, the location table 24, and the worker table 15.
RapidEye Research, PostgreSQL introspection (2026)
Statistic 29
Recurring visits are modelled properly, with a frequency set, a maximum order count, start and end dates, and a scheduled duration.
RapidEye Research, PostgreSQL introspection (2026)
Statistic 30
Across all 35 installed tables, zero columns relate to photos, images, attachments, inspections, checklists or quality verification.
RapidEye Research, PostgreSQL introspection (2026)
Statistic 31
Across all 30 modules, the closest capability to condition verification is customer signature capture.
RapidEye Research, module manifest review (2026)
Statistic 32
ZenMaid lists $19 per month for a Starter tier capped at 40 appointments, then $39 and $49 per month for unlimited tiers.
ZenMaid published pricing (2026)
Statistic 33
Jobber lists its single-user Core plan at $49 per month, or $29 billed annually, and places job documentation checklists on its Connect plan at $139 per month, or $99 billed annually.
Jobber published pricing (2026)
What this means: Open source solves the half of the problem that is a calendar. Nothing in it solves the half that is proving the work was done to standard, which is exactly where disputes with clients and property managers start. Whichever way an operator goes, the verification layer is bought or built separately, and that is worth pricing into the comparison before treating self-hosting as the cheaper option.
Cite this study
Academic or press use: copy a ready-made reference. RapidEye is the publisher.
Quick FAQ
Is there an open source alternative to ZenMaid or Jobber?
Not as a purpose-built cleaning product. Across 528 unique repositories swept in August 2026, 57 offer cleaning, maid, or janitorial business software, and they hold 54 GitHub stars between all of them. The most-starred has 7 stars and has not been committed to since January 2016. The realistic open-source path is a general field service platform: the Odoo Community Association's field-service module set, which is AGPL-3.0, has 98 contributors, and installed cleanly into Odoo 19 in our test with no manual fixes.
What is the best open source field service management software for a cleaning company?
The Odoo Community Association's field-service repository is the only maintained field-service platform that installed cleanly in our August 2026 lab. It carries 193 stars against 301 forks, ships 30 modules including recurring visits, routes, timesheets, worker skills and a customer portal, and recorded 396 commits over 52 weeks with 131 of those in the final four weeks. It publishes no tagged releases, so it is installed from a version branch rather than a release.
How many open source cleaning software projects are actively maintained?
Almost none. Of the 57 cleaning-business repositories identified in the August 2026 sweep, 67 percent had no commit in the previous 12 months and 88 percent carried no standard open-source license. Not one repository in the segment reaches 10 GitHub stars, and only 7 carry a standard open-source license at all.
Does open source cleaning software handle photo verification or quality inspections?
No. After installing seven field-service modules into Odoo 19, we queried all 35 resulting fsm tables and found zero columns for photos, images, attachments, inspections, checklists or quality verification. Across the 30 modules in the repository, the closest capability is customer signature capture. Scheduling and dispatch are well covered in open source; condition verification is not covered at all.
How much does commercial cleaning business software cost compared to self-hosting?
ZenMaid lists $19 per month for its Starter tier, capped at 40 appointments per month, then $39 and $49 per month for its two unlimited tiers. Jobber lists its Core plan at $49 per month, or $29 billed annually, for a single user, and places job documentation checklists on its Connect plan at $139 per month, or $99 billed annually. Self-hosting the open-source field-service stack removes the license fee but adds an Odoo server, upgrades across annual major versions, and the integration work that the commercial products ship finished.
Data sources
Every figure traces to one of these named public sources, each checked against the original before publishing.
GitHubRepository search, contents, contributors and participation statistics APIsgithub.com
Docker HubRegistry API image pull countshub.docker.com
Odoo Community AssociationField-service module set, manifests and branch documentationodoo-community.org
OdooOfficial container image and platform repositoryodoo.com
FrappeERPNext platform repository and container imagesfrappe.io
DolibarrPlatform repository and container imagesdolibarr.org
ZenMaidPublished subscription pricing pagezenmaid.com
JobberPublished subscription pricing pagegetjobber.com