Don't Rush Outside the Building
Without First Stress Testing Your Business Model
“Get outside the building” is one of the battle cries of the Customer Development/Lean Startup movement, popularized by Steve Blank. This is a call for validating the riskiest assumptions in your business idea as early as possible with evidence-based facts you gather by interacting with customers.
But it is possible to rush outside too quickly -- only to get stuck later.
Yes, it is highly tempting for action-driven entrepreneurs to want to “rush out of the building” and immediately start testing their business model on customers. It's even possible to quickly find clusters of problems, pitch an offer, build a quick MVP, and charge your customers from day one -- which sounds exactly like what you should do.
So what's wrong with this approach?
The danger is getting stuck with a sub-optimal business model six or nine months down the line that either doesn’t match your ambition or isn’t scalable.
This is precisely the trap I found myself in with one of my early products that I will share with you in today’s issue.
The idea spark
I’m going to pick up this story in the summer of 2009 when I was launching a new product: CloudFire — a “scratch your own itch product.”
I had recently become a parent and watched my wife struggle with sharing photos and videos with family and friends (remember that this was 2009). Having already built a few products that went nowhere, I didn’t want to fall into another build trap. So, I committed to carving out two months for customer development to validate my problem/solution assumptions.
I reached out to friends who had also become new dads, like myself, who quickly pointed me to the moms doing the bulk of the sharing. With the help of my wife’s network, referrals, and a few other channels, I managed to run over 60 problem interviews over six weeks.
Almost everyone I spoke with related to the struggle of sharing lots of photos. Their workaround was to hand-pick a few pictures every week and send them over email. This only reinforced my problem assumptions, and my confidence grew.
I put together a quick demo over the next two weeks and then spent the next four weeks pitching my demo (and offer) to these parents. I managed to get over 70% of my prospects to agree to test-drive my product which I packaged as a 30-day trial and then charged $49/year. By the end of these solution interviews, I amassed an early access list of 50 parents, which seemed like a good starting batch size of customers.
I shifted focus to building out a minimum viable product (MVP). It took another two months to get the product in the hands of my first customers.
Then my focus shifted to onboarding.
I spent many in-person sessions showing parents how to use the product, recording quick-start videos, and answering support questions. All were leading up to the moment of truth: Would people upgrade at the end of 30 days?
No funnel converts at 100%. I ended up with a 50% conversion rate, which was still highly encouraging. Many parents didn’t outright cancel but asked for more time to evaluate the product citing “life getting in the way.” I extended their trials and split my time across driving more new trials, supporting existing customers, and building new features.
Being a solo founder on this project, there were days when I just sat in my armchair for an hour trying to prioritize what to take on that day. I knew I’d eventually need to bring a co-founder on board but the product wasn’t making enough revenue to justify that.
Maybe it was time to raise a seed round?
So I contacted a fellow entrepreneur, a longtime friend, and an angel investor.
I gave him a quick update and asked for his advice on fundraising. At this point, nine months since I started, I had 500 paying customers, translating to $25k in revenue.
Friend: “While you’ve made a lot of progress, you’ll need to show better metrics to get an investor's attention.”
Me: “Yes, I know the revenue number isn’t that impressive yet, but doesn’t 500 customers de-risk the product for an early-stage investor? The market for parents is big enough with lots of growth potential.”
Friend: “Yes, that’s a great start. But it would help if you showed a roadmap for how you’ll intersect your investor’s expectations at product/market fit. Right now, my instincts tell me that something is off here….”
He pulls out his notebook and walks me through a 5-minute back-of-the-envelope calculation to test the viability of my idea.
We determined that to pitch my idea to an investor, I’d need to 4x the number of customers (from 500 to 2,000) in the next three months. And then grow that to 20,000 customers by the end of year 2 and 200,000 customers by the end of year 3.
Me: “These are crazy numbers. I don't see how I'll get there unless my product goes viral. I guess I was betting on going after a big market, but I see now that simply relying on top-down sizing of any market can be misleading.”
Friend:” That’s right. ‘I just need 1% of a big market’ is what everyone uses, but the real pudding in the pie is demonstrating how you get there in 2-3 years.”
Just then, a thought comes to my mind.
A business model pivot
Me: “I can’t help but notice that pricing is one of the key metrics that destroyed my business model. If I could charge more for my product, I’d need fewer customers.”
Friend: “Pricing is a powerful lever, but can you charge parents a lot more than $50/year since their current solution is email, which is free?”
Me: “I wasn’t thinking of parents, but wedding photographers. Just the other day, a wedding photographer who is also a parent reached out over email and was wondering if I’d consider creating a Pro version of CloudFire. He loves the product and was inquiring if I’d be up for building some integrations into products he uses in his workflow, like Lightroom and Photoshop. Given it would be a B2B product, I’m guessing I could charge 3x or 4x more….”
Friend: “I think you could charge 10x more. Instead of $49/year, you could easily set pricing at $49/mo and probably go even higher. Set up a call with him immediately.”
The price of learning
On the ride home, while I’m deflated about my busy parents business model, I’m excited about the potential of pivoting to a more promising new customer segment.
I think to myself: “I’m glad I reached out to my friend when I did, or I would have kept going like this for who knows how long.”
I rationalize the last nine months as the price of learning, until I get hit with another depressing thought: “Nothing we did in that back-of-the-envelope calculation required a working product. I could have done that math nine months ago and reached the same conclusion!”
In my exuberance for “getting outside the building,” I rushed out too quickly. And even though I started with a rough sketch for my business model (using a Lean Canvas) and gave some thought to key metrics, I didn’t know how to properly stress-test my business model.
Had I known what I knew after that meeting, I would have saved nine months of my life.
That experience completely changed how I pursued future projects.
Before engaging in several weeks and months of customer validation, it’s prudent to spend a few more hours (inside the building) stress-testing your business model and tightening up any apparent cracks or flaws in your thinking.
On the surface, this may seem counter to one of the key tenets of Lean/Continuous Innovation: Using small and fast experiments to test your riskiest assumptions quickly.
If running experiments is the most effective way for testing assumptions and speed is key, then why spend more time refining your rough assumptions?
Because it is possible to effectively stress-test your assumptions using thought experiments before you test your assumptions using empirical experiments.
Before dismissing thought experiments as sub-par, remember that Albert Einstein used thought experiments, not empirical ones, to formulate the theory of relativity.
If thought experiments can work for scientific inquiry, don’t you think they can also be applied to entrepreneurial inquiry?
What I’m advocating here isn’t simply refining your business model through brainstorming but rather using well-structured thought experiments to stress test your business model.
There are at least three business model stress tests I put all my ideas through before getting outside the building. These stress tests check for desirability, viability, and feasibility. This post covered viability.
Running thought experiments is much faster than running empirical experiments saving you time.
Thought experiments help you fix fatal flaws in your thinking, saving you from having to backtrack from a dead-end.
Thought experiments help you prioritize the riskiest assumptions facing you now versus those you can defer to later, helping you focus on the right things.
Finally, it's important to emphasize that thought experiments do not replace empirical experiments.
Passing these stress tests doesn’t guarantee that your business model will work. But not passing these stress tests does guarantee that your business model will not work.
Falsifiability or testing for invalidation (versus validation) is another tenet in the Lean/Continuous Innovation Framework taken from the scientific method. In other words, the job of an experiment isn’t limited to just paving the right path. Eliminating dead-ends is progress too.
The true job of an entrepreneur is systematically de-risking their business model before running out of resources.
I wrote about this 5-minute Rapid Viability Test here.
I subsequently built a more powerful version of this stress test: the Traction Roadmap.