TL;DR

After Harvard, I answered a post on the Harvard Business School mailing list looking for a technical co-founder and ended up building "Uber for home cleaning." Two of us coded the MVP in two weeks, recruited background-checked cleaners from Craigslist, and got our first customers from friends, family, and flyers in Harvard dorms. We were accepted into Summer at Highland, pitched dozens of investors, heard "no" relentlessly — and then two big investors wired $2 million after three months of visible, week-over-week traction. We moved to New York, hired a prominent CTO and a real team, survived buggy code and bad cleaners, and grew to a million customers a month. Here's how it actually happened, mess included.

One Email That Changed Everything

It started with an email I almost didn't open.

A few weeks after graduating from Harvard, I was sitting in my apartment with a cold cup of coffee, scrolling the Harvard Business School mailing list out of habit. Buried between event announcements was a short post: an HBS grad with a business idea, looking for a technical co-founder.

The idea was simple enough to fit in a sentence: Uber for home cleaning. Tap a button, and a vetted cleaner shows up at your door. No phone calls, no haggling, no wondering if anyone will actually come.

We met at a coffee shop in Harvard Square the next day. He slid his laptop across the table, eyes bright.

"Everyone hates two things," he said. "Cleaning their apartment, and calling strangers on the phone to clean their apartment. We kill both problems with one app."

"And the cleaners?" I asked.

"Flexible work, steady demand, instant pay. Drivers got Uber. Cleaners get us."

I stirred my coffee and felt the particular electricity you only feel a few times in your life — the sense that an idea is obvious, urgent, and somehow still unclaimed. By the time the cups were empty, I was in.

Two Weeks, Two Programmers, One MVP

There were two technical co-founders — me and one other engineer — and we made a pact: a working product in two weeks, not a pitch deck in two months.

Those fourteen days blurred together. Empty takeout containers stacked like a skyline on the desk. Whiteboards covered in booking flows and database schemas. One of us coding the customer side — booking, scheduling, payments — while the other built the dispatch logic for cleaners. We wired in Twilio early so the system could text customers and service providers automatically: booking confirmations, "your cleaner is on the way" alerts, reminders to the cleaners themselves. Those little text messages made a two-person operation feel like a real company.

At 3 a.m. on day fourteen, my co-founder hit enter on a test booking. Both our phones buzzed at the same moment with a confirmation text. We looked at each other over the laptop screens.

"It's alive," he grinned.

It was ugly. It was held together with digital duct tape. But it worked — and "works" beats "perfect" every single time at this stage.

Craigslist Cleaners and Harvard Dorm Flyers

A marketplace needs two sides, and we had neither. For supply, we went where cleaners already were: Craigslist. We posted ads, interviewed dozens of applicants in coffee shops, and ran every single one through a background check before they could accept a job. That last part was non-negotiable — we were sending strangers into people's homes, and trust was the entire product.

For demand, we started embarrassingly small: friends and family. My co-founder's aunt. My old roommate. Anyone who owed us a favor and owned a dirty apartment. Then we printed flyers and spread them through Harvard dorms — slipping them under doors, pinning them to bulletin boards between sublet ads and a cappella auditions. Students, it turns out, will happily pay to never clean a bathroom.

And here's the thing we did that most founders skip: sometimes we cleaned the customers' houses ourselves. I have scrubbed tubs and vacuumed under couches as the co-founder of my own company. It sounds absurd, but those hours on our knees taught us more than any dashboard could — how long a real cleaning takes, what supplies matter, what customers actually inspect when they come home ("they always check the sink," one customer told me, watching me polish it). You cannot build a great service business from behind a laptop.

Summer at Highland and the Demo That Opened Wallets

Around this time, my co-founders and I got accepted into Summer at Highland, the startup program run by Highland Capital. Suddenly we had office space, mentorship, and — most importantly — a steady stream of investors walking through the door.

I built the investor demo myself, and I treated it like a product launch. Not slides — a live booking, on a real phone, in front of them. In meetings, I'd hand an investor my phone.

"Book a cleaning," I'd say. "Go ahead."

They'd tap through it, and seconds later my co-founder's phone — playing the role of a cleaner — would chirp with the Twilio dispatch text. Then the investor's phone would buzz with the confirmation. You could watch their posture change. Charts get polite nods; a working machine in their hands gets leaned-in silence. That demo, more than any spreadsheet, is what ultimately convinced our investors to write checks.

But don't let the highlight reel fool you. We spoke to dozens of investors before anyone said yes. The no's had a taxonomy: the fast no, the slow fade, the flattering no ("love you guys, too early for us"), and the brutal-but-useful no.

One investor leaned back in his chair, arms crossed. "Cleaning is a commodity. What stops three Harvard kids next year from doing the same thing?"

"Nothing stops them from starting," I said. "But we'll be a year of bookings, cleaners, and lessons ahead. This business is won in operations, not ideas."

He still passed. But that answer got sharper every time we gave it.

How We Actually Raised $2 Million in 3 Months

People hear "we raised $2 million in three months" and imagine luck or connections. The real answer is a principle one mentor at Highland drilled into us: investors invest in lines, not points.

A point is a snapshot: "We have 50 customers." A line is a trajectory: "We had 12 customers in week one, 31 in week three, 78 in week six, and here's the curve." A point is an anecdote. A line is evidence — proof of momentum, and proof that the team can manufacture momentum on demand.

So we engineered the line. We worked 100 hours a week — and I mean that literally, fourteen-hour days, seven days a week, sleeping in shifts when a launch demanded it. Every week had to be visibly better than the last: more bookings, more repeat customers, more cleaners onboarded, faster dispatch times. When we met an investor twice, the second meeting's numbers had to embarrass the first meeting's numbers. That was the whole strategy.

Investors told us afterward what tipped them: traction (the line kept bending upward), a cohesive team (two business minds and two engineers who finished each other's sentences), and unmistakable hard work (we answered emails at 2 a.m. because we were awake anyway, shipping).

After three months of this, two big investors committed, and $2 million landed in our account. I refreshed the bank page three times. Then we packed our lives into boxes and moved to New York — bigger market, denser apartments, endless dirty kitchens — to grow the company for real.

Fires, Bugs, and Broken Vases

Growth exposed every weakness at once.

On the operations side, service providers were our biggest headache. Some did sloppy work. Some showed up late — fatal in a business whose whole promise is reliability. One memorable afternoon, a customer called, voice tight: "Your cleaner just broke a vase my grandmother gave me." I apologized, paid for it, and personally re-cleaned her apartment that weekend.

We developed a hard rule: fire fast. The moment a cleaner showed a pattern — lateness, complaints, carelessness — we let them go and replaced them with new providers from our constantly refilling Craigslist-and-referral pipeline. It felt harsh. It was also the only way to protect the trust that made the marketplace work. Your service providers are your brand; every great one compounds it, every bad one burns it.

On the engineering side, our two-week MVP heritage haunted us. At the beginning there were a lot of bugs, because we were moving too fast — double-booked cleaners, payment hiccups, texts firing at the wrong times. One Saturday morning the dispatch system silently failed and a dozen customers waited for cleaners who never knew they'd been booked. We spent the day calling every one of them personally and offering free cleanings.

That weekend forced a reckoning. Pure speed was breaking the thing speed was supposed to build. Pure caution would have killed our momentum. Eventually we found an equilibrium that became our engineering motto: move fast, but don't break things. Ship daily, but test the booking flow like lives depended on it. Take risks in the new features, never in the core promise.

Building a Real Team

The $2 million bought us the thing we needed most: people better than us.

We recruited a prominent CTO — a veteran who had scaled systems far bigger than ours — and watching him work was an education in itself. He took one look at our codebase, raised an eyebrow, and said, "This is impressive for two people in two weeks. Now let's make sure it survives two hundred thousand people in two years." Under him we hired multiple programmers who rebuilt our fragile MVP into actual infrastructure.

Then came customer service reps — warm, unflappable people who turned angry calls into loyal customers. Marketers who replaced our dorm flyers with referral programs, neighborhood launches, and partnerships. Operations folks who professionalized cleaner recruiting, training, and quality scoring far beyond our Craigslist origins.

That machine — solid tech, relentless marketing, ruthless quality control, and the simple magic of a service people genuinely need every week — is how we scaled from flyers under dorm doors to a million customers a month. Cleaning is a repeat business: win a customer once, serve them well, and they book again and again, then tell their neighbors. The same Twilio texts that once thrilled two sleep-deprived founders at 3 a.m. were now firing millions of times a month.

What I'd Tell a First-Time Founder

Standing in our New York office months later, watching the booking map light up across the city, I kept thinking about that coffee shop in Harvard Square — two strangers and a sentence-long idea.

If you're where we started, here's the distilled version. Answer the email; the cost of a coffee meeting is nothing. Build the working thing in weeks, not the perfect thing in years. Do the unglamorous work yourself — scrub the tub, take the angry call — because that's where the business actually lives. Guard quality brutally, even when it means firing fast. Find the equilibrium between speed and stability before a broken Saturday forces you to. And when you raise money, remember that nobody is investing in your point on the map. They're investing in your line — so put your hundred-hour weeks into making that line undeniable.

The idea was never the hard part. The line was.