Startup Shutdown
What I learned from launching a startup: Assumptions kill products
Reflecting back, now ten years later…
In late 2013, I met two former colleagues for coffee to brainstorm business ideas. We were all hungry to start something on our own.
We settled on the home remodeling industry — an industry we knew nothing about.
We reached out to a few contractors and started living and breathing home remodeling. We traveled in their trucks. Shipped prototypes while camped out at job sites. Watched them plan projects on whiteboards.
We formed relationships with a group of contractors we felt represented our ideal customer: Typically had 3-4 open jobs at a time, used QuickBooks, spreadsheets, their smart phone, a whiteboard for scheduling, and a decade or three of muscle memory.
Our hypothesis: If we could make it easier for the contractor to manage their projects and reduce friction in communicating with the homeowner, then the homeowner would have a better experience, resulting in more referrals and business for the contractor.
Our back-of-the-napkin assessment in 2013:
Opportunity: 300,000 residential construction firms in the U.S., with less than 5% using competing software. Room to expand by use case and vertical (pools, roofing, landscaping), as a marketplace platform for construction supplies and tool rentals, and introduce premium tiers with wifi cameras, and more.
Problems: Wrangling sub-contractors, change orders, materials delays, and keeping homeowners in the loop (and off their backs).
Competition: Tougher competition in commercial and higher end residential. Two of the closest competitors were only targeting new construction with a starting price point of $100/month/job. But of course, the actual competition was muscle memory, and pen and paper workflows.
Acquisition: Contractors are always on the phone. Because of that, we found that cold calling (leveraging purchased permit data), paired with a free trial CTA worked well with almost startling predictability.
Conclusion: Perfect for a low-priced, simple SaaS offering.
By February 2014, we had bootstrapped Contractor View: Scheduling and communication software for the independent general contractor. A lightweight, project planning tool specifically built for home remodeling use cases — complete with canned templates and starter plans — for web and iOS/Android. Contractors could build their high level project plans, assign subcontractors to jobs, and send automated text messages as reminders. Subs could upload pictures and ask for clarification while on site. All starting at $30/month.
And after six months, we shut it down. Here’s what I learned: Assumptions kill products.
The Assumptions
Assumption 1: Positive early feedback would predict adoption. We made it as easy as possible for the group of 30 beta contractors to start. We’d fill in current projects manually to ease adoption (our “custom” onboarding!). We’d sit next to them on site as they logged in: “Wow! Can’t wait to use this!” But as the weeks rolled by, we noticed a growing dissonance between their reaction, born from aspiration, and their actual behavior, constrained by day-to-day chaos. Our assumption was that the positive feedback they gave us after a demo would carry through to purchasing and adoption behavior.
Assumption 2: Behavior change is easy. We finally realized the massive shift in behavior that Contractor View demanded of contractors. Everyone else in the contractor’s orbit — sub-contractors, suppliers, etc — were still using the lowest common denominator (usually an urgent phone call). We were asking a lot. In our Product bubble, swapping out tools and adopting new software and workflows was a constant though; Our assumption was that contractors would be as quick to adopt new tools as we were in our own jobs.
Assumption 3: Everyone wants to scale. On the surface, the value proposition resonated with contractors. We assumed that contractors 1) did not want to spend their time managing minutia and putting out fires, and 2) would rather be finding more crews and sourcing more jobs. As it turned out, we were mostly incorrect on both. In one of our last interviews, a contractor put it like this: “What am I supposed to do if the app does all that stuff for me? Just do QuickBooks back at the office?” Most contractors enjoyed the work we thought they’d be eager to offload, and they rarely had grand visions of scaling, given how difficult it was to grow a business that depends on labor and materials — two very challenging ingredients! Our assumption was that contractors thought about scaling their businesses in the same way we did.
Assumption 4: Raising money would be easy, and six months would be plenty of time. Maybe it was ignorance — or arrogance — but we had traction with our beta, had gotten accepted into Capital Factory (a startup incubator in Austin, TX), and were talking to angel investors weekly. Yet, every conversation ended with: “Stay in touch” (code for: Not a giant market, not a giant risk, and home remodeling isn’t that exciting). We weren’t ambitious enough! And it was too early. We were unable to raise before our self-imposed runway ran out, which limited our product market fit iterations. Our assumption was that six months was plenty of time. While this tweet didn’t exist in 2013, it certainly hits the mark looking back:
The Conclusions
As a product manager, guarding against biases and assumptions is crucial. I’ve carried these learnings with me, and Contractor View fundamentally changed how I approach product management.
Hunt for assumptions and root them out. The most dangerous assumptions are the ones we don’t know we’re making because they influence the product without even recognizing it. I’ve focused on giving teammates permission to question and challenge so I can pause and question. Slowly over time, I’ve started to self-serve and been able hunt for those assumptions myself.
Test for assumptions and be willing to change my mind. With Contractor View, we researched and tested the UX, the value prop, the price point — everything. Looking back through our interviews, we learned a lot about home remodeling workflows. But we never tested our deeper assumptions about motivation and aspiration. And we were determined to get our product to the next stage. I’ve added a few more questions when I talk with users, like: “What would you stop doing if you used this?”
For those starting the startup journey now:
Your cofounders will be everything — I feel extremely fortunate for mine.
It will all take longer than you think.
You will learn and grow faster than any other time in your life.
Remember: Assumptions kill products.
A version of this post, First Time Startup Cofounder, was originally published in Austin Startups on Medium in 2015.