HomeJoy was a home-cleaning marketplace that raised $40 million and shut down on 31 July 2015. Too many customers and cleaners met through the app and then dealt directly, cutting HomeJoy out, while it also fought worker-classification lawsuits and failed to raise more money. The lesson: a marketplace that only makes the introduction gets bypassed, so keep adding value after the match.
Key takeaways
- HomeJoy raised $40 million, then shut down on 31 July 2015 after failing to raise a big enough round while fighting worker-classification lawsuits.
- Disintermediation means buyers and sellers meet on your platform, then leave to deal directly. On repeat services like cleaning, the incentive to leave is huge.
- As HomeJoy closed, rival Handy offered HomeJoy cleaners a $1,000 signing bonus. Handy was later acquired by ANGI Homeservices in October 2018.
- Modern marketplaces fight leakage with software: booking, payments, reminders, reviews and guarantees that make staying easier than leaving.
- Charge for new connections, not for every repeat job, and the incentive to go around you mostly disappears.
Imagine you book a cleaner through an app.
She's brilliant. On time, thorough, friendly. You book her again. And again.
By the fourth visit, she says quietly, "You know, if you just pay me directly, I can do it for a bit less. And I'll get more too."
You both save money. The app gets nothing.
Now multiply that conversation by thousands of customers, and you have the story of HomeJoy.
What did the book say about HomeJoy in 2022?
In Chapter Six of The Subscription Playbook, I called disintermediation the biggest challenge facing any marketplace network. It's "a fancy word" for what happens when a supplier and a customer connect through your platform, form a relationship, and then leave to work together directly.
They cut out the middleman. You.
I used HomeJoy as the warning: a home-cleaning startup that raised $40 million in funding "before going under because too many of its users ended up cutting it out of the process."
What's happened since?
HomeJoy is a closed story. But looking at how it ended, and what followed, makes the lesson sharper.
- HomeJoy shut down on 31 July 2015. Reporting at the time said it had failed to raise a large enough funding round while fighting lawsuits over whether its cleaners should be classified as employees.
- As HomeJoy closed, rival Handy offered HomeJoy's cleaners a $1,000 signing bonus to switch.
- Handy itself was acquired by ANGI Homeservices (the HomeAdvisor and Angi group) in October 2018.
So disintermediation didn't act alone. It weakened the business, and the legal fight made raising more money harder. Each problem made the other worse.
Then the cleaners walked straight over to a competitor. Which tells you something: the platform owned very little that the supply side couldn't get elsewhere.
Why do service marketplaces leak?
Let me show you the math. These numbers are illustrative, not HomeJoy's.
Illustrative: why both sides win by leaving
A weekly clean costs the customer $120 through an app. The app takes 20%, so the cleaner receives $96.
Now they agree to go direct at $110.
- The customer saves $10 a visit, or $520 a year.
- The cleaner earns $14 more a visit, or $728 a year.
- The app loses $24 a visit, or $1,248 a year, for that one customer.
Both people the platform serves are better off without it. That's the problem in one box.
The more repeat business there is, the bigger the incentive. A one-off job isn't worth the awkward conversation. A weekly service for years definitely is.
And here's the cruel part: the better your marketplace is at matching people well, the more likely they are to want to keep working together, and the more likely they are to leave you.
Why does this still matter in 2026?
Because it changed how service marketplaces are built.
A decade on, the platforms that survive give providers software they can't easily walk away from: booking, payments, reminders, reviews, insurance and guarantees. The introduction is just the start. Value keeps flowing after the match, so staying on the platform is easier for both sides than leaving.
The best example is StyleSeat, the hairdresser marketplace I also wrote about in the book. It gave stylists free booking and payment software, and it now charges a fee only on brand-new clients the marketplace sends, never on their repeat visits. That removes most of the reason to go around it. I cover the details in how StyleSeat started charging for new clients.
Could your network hold together?
Score your business against the six network effects, and spot where people could go around you, with the free Network Effects Scorecard.
Six ways to stop people going around you
These come from the book's examples plus what's worked since.
- Hold the money. Upwork's payment protection means freelancers get paid and clients only pay for work done. Without the platform, both lose that safety net.
- Give suppliers tools they'd otherwise pay for. Scheduling, invoicing, reminders, reviews. If leaving means losing the tools that run their business, they'll think twice.
- Mask contact details. Uber and Lyft hide drivers' phone numbers. You can't call someone off-platform if you don't have their number.
- Taper the fee. Tutor.com started at a 50% take and reduced it with each transaction. The longer the relationship, the less reason to leave.
- Charge for the introduction, not the repeat. A one-time "finder's fee" for a new customer, as StyleSeat does, means repeat business costs the supplier nothing extra.
- Guarantee the result. A satisfaction guarantee or insurance cover only applies to jobs booked through you. That's the customer's reason to stay.
And then there's the sneaky option. Hired.com sent job seekers a $150 bottle of champagne when they landed a job, then checked whether the employer was a customer trying to dodge its fee. Clever. Maybe a bit too clever for most of us (but I still love that story). If you do lean on contract terms to stop clients being poached, check them against Australia's unfair contract terms rules first.
Network Effects ScorecardWhich of the six network effects could you build?Open the free tool →What you can steal
You don't need to run a marketplace for this to apply. Any business that introduces clients to contractors, subcontracts work, or refers jobs has the same risk.
- Own the booking, payment and guarantee. If the relationship runs through your systems, you stay in the loop.
- Keep your take low on repeat work. A high commission on every repeat job is an invitation to go around you. Consider a modest monthly subscription for the provider instead, or mix the two as I describe in the five pricing models.
- Make your guarantee the reason customers stay. "Booked through us, guaranteed by us" is a promise the provider can't match alone.
- Ask yourself the leaving question. If your best supplier and best customer went direct tomorrow, what would they lose? If the honest answer is "nothing much," start building that something.
If you're running a cleaning business rather than a marketplace, flip this around: a direct subscription with your own guarantee and easy rebooking is exactly what platforms are trying to copy.
The lesson
A marketplace that only makes the introduction will be bypassed.
Network effects are the strongest moat I know, as I explain in my guide to network effects. But a network only protects you if people stay inside it. Keep adding value after the match, charge in a way that doesn't punish loyalty, and you turn the introduction into a relationship you're part of.
For the wider picture of defending your business, see my guide to digging your moat, and for a marketplace that got the balance right, read how Zillow turned Trulia's network into a profitable business.
Frequently asked questions
What does disintermediation mean in simple terms?
It means cutting out the middleman. A customer and a supplier meet through your marketplace, get to know each other, and then agree to deal directly so neither has to pay your fee. You did the hard work of the introduction and lose all the future revenue. It's the biggest structural risk for any marketplace built on repeat services.
Which businesses are most at risk of disintermediation?
Marketplaces for repeat, local, personal services are the most exposed: cleaning, tutoring, dog walking, beauty, trades and freelancing. The customer wants the same person every time, the job is simple to arrange by text, and both sides can save money by skipping the fee. One-off or complex transactions, like buying a house, are much harder to take off-platform.
Is it legal to stop suppliers taking clients off a platform?
Many platforms include terms that restrict taking clients off-platform for a period, but how enforceable those terms are depends on the contract and the law where you operate. In Australia, unfair contract terms rules can apply to standard-form small business contracts. Value-based fixes like payments, guarantees and tools tend to work better than legal threats. Get advice before relying on restrictive terms.
Did HomeJoy fail only because of disintermediation?
No. Reporting at the time said HomeJoy shut down after failing to raise a large enough funding round while fighting lawsuits over whether its cleaners were employees or contractors. Disintermediation weakened the business, and the legal risk made new investors nervous. The two problems compounded each other.
Get both books free
The Subscription Playbook (ebook + audiobook) and Feed a Starving Crowd, instantly. In Australia? Claim the free hardcover.
Robert Coorey, MBA
Robert is the author of The Subscription Playbook and Feed a Starving Crowd and a Director at BPAG Accountants in Sydney. More about Robert.
Sources
- Vox/Recode, Homejoy shuts down after battling worker classification lawsuits (17 Jul 2015)
- TIME, Homejoy Cleaning Company Shuts Down (18 Jul 2015)
- Wikipedia, Handy Technologies
Facts and prices checked 27 October 2026. Spotted something out of date? Tell us.