Six weeks ago my friend Sam and I graduated from college and moved to San Francisco. We joined forces with my buddy Omar and started building a company. Omar is the grizzled vet, he has a few cool projects under his belt. Sam is the engineer. And I do everything uncoded.
We didn't come to the city with a brilliant idea - but rather a philosophy. We want to iterate through ideas efficiently and methodically, until we find the one. We're operating lean and using data to judge the success of our ideas, rather than saying "lets try this - and see what happens."
The first idea we toyed with was factoring. Let's say, a baker drops off $1,000 of bread at a restaurant and then sends an invoice. The restaurant dictates the payment terms - they'll pay in 30 days. This is all well and good - unless the baker needs to pay rent (in 15 days) or employees (in 5 days). Factoring allows the baker to sell an invoice (as soon as he drops of bread) for 97 cents on the dollar.
3% is a small percentage. But annualized - it's pretty significant. If you can manage the risk, the economics are amazing. Factoring hasn't changed since it's invention in ancient Mesopotamia.
Week One: Learn
Starting out cold, we knew that the amount we could learn about factoring was limitless. So we narrowed our focus to two questions:
Q1: How do existing factoring companies work?
We camped out in the Stanford library and read. We posed as customers and acquired 5 contracts from 5 different companies. We consulted lawyers. We spoke with people who factored and worked in factoring. We drilled into the competition – bluevine, fundbox, etc.
The complexity of the processes we studied was astounding. We knew we could make the process simpler. But we wondered:
Q2: Would more businesses factor if the process were simpler?
So, we set out to factor some invoices. We knew contractors working for large companies had to wait 30 to 90 days for payment. And we knew that is was painful (we've been there). But we didn't know if these contractors would factor.
I emailed 200 freelancers asking if they wanted to factor. Many responded and several were interested. The problem was that my emails did not radiate security and privacy. Our team also knows that that expressing interest in a product isn't the same as paying for a product. So we built.
Week Two: Build.
We built a site and got it in front of 500 freelancers. Our rates were artificially low and our application was artificially easy - we were trying to minimize false negatives. To our surprise, people submitted invoices.
As invoices trickled in, we reacted. We manually approved and disapproved invoices for factoring. Two problems arose. This unprofessional system spooked our customers. And we found that all the invoices were tainted (overdue or owed by someone sketchy). We formed two more questions:
Q3: Can we collect enough data to determine the health of debtors? And can we do so through a simple process?
Q4: Will healthy businesses factor to optimize cash flow? Or will we only attract only bad debt.
To address the first question, we built a simple authorization flow. The interface was sexy and it collected only essential information. The latter question required a larger sample size. We relaunched.
People ran through our verification flow and shared their information. But the vast majority of invoices submitted were overdue or tainted. Most people we dealt with were looking for a form of debt collection. While It was tempting to jump off into debt collection - we tried to focus on vetting the factoring idea.
Week Three: Expand.
We were still excited, but our fundamental hypothesis appeared weak. We thought healthy businesses would prefer $98 today to $100 in 30 days. Yet we hadn't hadn't found this in the contracting world. We asked
Q5: Would people in other industries be interested in factoring?
We took to the phones and contacted hundreds of SMBs and accounting firms looking for a fit. Nothing.
I called 100 family friends and contacts at SMBs - ranging from bakers to lawyers. Many discussed cash flow issues, outrageous payment terms, and financing instruments that they used. The problem existed.
The problem just didn't seem to be big enough. SMB owners hated outrageous payment terms, but didn't want to give up 2% for early payment.
We found our ideal customer:
- Operates in a high margin business - so can afford 2% to get paid now.
- Has a week negotiating position compared to debtor - so is forced to accept crap payment terms.
- Has previous transaction data readily available - so we can assess risk.
Customers meeting all the criteria were the customers that already factored - truckers, shippers, factories, manufacturers, etc.
Week Four: Reflect.
It seemed that we would not bring factoring to SMBs everywhere. But perhaps there was a way to beat out current factoring companies. We marched forward asking:
Q6: Could we beat current competitors? Could we win contracts from those that already factor?
We read and spoke to people at factoring companies. Our findings were similar across verticals - it didn't seem that technology would help. Take trucking as an example:
Most truckers don't have smartphones. Most truckers don't use computers. Factoring companies currently send agents to ports, docs and trucker bars. There, the agents chat up truckers and win their business. It seemed that an iphone app or website would be tough to implement. It seemed that expanding would require agents and salesmen on the ground, armed with data input devices.
The biggest factoring company (CIT) is teetering on the edge of 50B in assets - so there is certainly a business there. We thought we could build something in the space... but we ended up reaching a final question:
Q7: Was this how we wanted to spend the next 5 years of life?
If we started a factoring company, we wouldn't be playing to our strengths. We would be building something that people wanted, but nobody loved. We would be building something that scaled with people more than technology.
We tried some interesting and creative last ditch efforts. I could write pages on each of them. But in the end we threw in the towel.
On to the next one.
TLDR: Team of three tries to reinvent 3,000 year old industry. Is methodical. Hacks aggressively. Measures carefully. Decides time is better spent elsewhere - then moves on.
Week 5: Rebirth.
A week later. We moved to the next idea. Again methodically testing. But the test got away from us. Users (paying) were signing up too fast to handle. Too fast to support. We scrambled to create a wait list.
More to come. Users means nothing. A wait list over a 1k means nothing. Unless we pull it off. Unless we're still on this idea in a month. Back to running experiments. Back to fighting fires. And deciding which fires to fight.
Ok. Maybe all that means something. We'll see.