It’s Not a Feature Problem—Avoiding Startup Tarpits

Startup growth can be painfully slow. So slow that many entrepreneurs just give up and quit when their launch doesn’t go as expected. When you think about it, who can blame them for quitting. Yes, it is really hard. So hard that you may end up in a nut house if you think about it too hard. However, that’s not why I expect that most founders opt out.

They fail because they never had a chance, not when every incentive leads them in the wrong direction. The top thought leaders and business experts deceive us with shortcuts and tricks that optimize for short term success. There are no silver bullets; there are no shortcuts.

We’re dazzled with the idea of hockey stick growth — the idea that there is an inflection point in a startup’s growth that moves the rate of expansion from linear to exponential. One day you’re working in your garage on an idea; the next day you’re Facebook, Uber, or AirBnb.

We’re told this jump from linear to hockey stick growth happens when a great product meets a great market; that at this moment word of mouth will begin to propel extraordinary growth.

Products and markets do matter—they actually matter a lot. However, there’s a temptation to take this too far. To toil away at a product until you get a perfect product. That was precisely what we did when we launched our first company. It was one of our original sins.

We developed new features thinking each feature would cause the inflection point we were looking for. ‘Ticketing will be a game changer,’ or ‘this new mobile app will allow us to sell at a higher price point.’ Each time we barely pushed the needle when it came to growth. We dreamt silver bullets, but we were only pushing out what amounted to paper bullets.

It turns out paper bullets are expensive. Features cost money in the form of engineering hours. We invested all of our seed funding in developing a great product. In doing so, our feature list expanded, but our actual revenue growth never changed in a meaningful way to support that development.

Eventually we stopped making new features all together. We had no choice; we couldn’t meet payroll to continue all of this development. We were in what I call a startup tarpit. The startup tarpit is the land of inertia. We had a good enough product but not enough marketing resources to get out of our stagnation. We raised money to build a great product and when we ran out of money we had nothing to show for it; making it nearly impossible to raise additional money.

Thankfully our product was stable. Our CTO architected an incredibly resilient and scalable product. We didn’t have to shut down the company altogether; we just had to keep product going on a nominal budget. A nominal budget meant no more features.

There was a silver lining in killing our addiction to features. We started reinvesting revenue into marketing. We had never committed to marketing before. We always feared that our product was not ready to be marketed. If we pay for users they’ll just cancel their accounts early into their subscription was the thought.

When we turned on paid advertising for the first time, we had a sizable increase in signups. We always feared that a new user would just churn because of what we perceived as deficiencies in the product. While there were users who churned for that reason, it was never the nightmare scenario that we imagined. All the imperfections we noticed weren’t even a consideration for our users. The majority of our users were perfectly satisfied in the product as it was.

It became clear that we had been sitting on a business; we just didn’t know it up to that point. It made it clear that the business was more dynamic than just the product itself. The business was actually a dynamic model with the foundation of a product.

The model was really predictable. A customer finds us through Google Adwords. The cost to get that business to become a paying customer by clicking on one of our ads is on average around $70. The average user pays us about $35 a month for our subscription service. That means it takes less than two months to recover the marketing investment it took to acquire that user. After those first two months any further revenue is basically profit. It was just a matter of keeping the customer satisfied, which was easier than we thought.

Bar none marketing was the single best investment we made. The small iterations of the product just didn’t compare to the return on our marketing investments. I imagine this is the case with most startups. Withstanding some meaningful pivot of an idea, the first iteration of a product delivers the kernel of value that delivers for years to come. Small iterations in the product just deliver marginal returns.

This is why the product development is not everything. It is certainly the foundation of the business, but without ever getting users in a meaningful way you’re never giving your business the chance to survive.

The implications of marketing early are really easy to understand. Imagine if we thought of our seed funding as a means to propel revenue and not our list of features. If we hired one less engineer and took what would be that person’s salary and invested it into our marketing budget for example. It would have fundamentally changed the trajectory of the business.

Say we took 12k each month and invested into Adwords. We know it cost us on average $70 to acquire the customer. That means with 12k monthly budget marketing we would buy about 170 users a month. A user on average pays us $35 per month. This means we are adding $6000 to our monthly recurring revenue. The marketing budget pays for itself within two months.

With that budget, when you factor in customer churn, it should take about 16 months to hit a 1m run rate. Create a marketing engine early and the fruits of that work will pay off forever. A million dollar run rate gives you some peace of mind that you are not going out of business any time soon and it also allows you to continue to develop your product. Get to a million dollars in revenue and you have a business and you don’t have to be dependent on VC funding to survive—that’s the startup promise land.

Now imagine if you build your product on your time. Imagine that your only product costs are nominal hosting and operational costs. Imagine then you raise your round and invest early into customer acquisition instead of features. You can have a business that can sustain itself forever.

There are also additional byproducts of this approach. In our case marketing lead to an actually improvement in our product in a measurable way.

Prior to our marketing initiative we had made development decisions on the basis on a small user base that found us through free acquisition channels (blogs, word of mouth, et). This is how a lot of startups end up in their own version of a startup tarpit; they assume word of mouth and organic growth will be the primary driver of growth.

The amount of traffic we generated from those channels was anemic, which I imagine is the case for most startups. As a result, we only really listened to the needs of a small user base. Listening to those users led us to believe we had a product problem. Our stagnation was because we had not built something feature rich enough.

In actuality one of the biggest impediments to our growth was not that our product lacked features, but instead it was really hard to actually signup for our service. When we turned on ads this became apparent to us. We would spend thousands of dollars a month to get users to a signup page that was a bottleneck to getting to the product.

Meaningful improvements in a product are created through feedback on a larger scale. At scale you see the fissures in the business and product. In our case, we were able to decrease the cost to acquire a customer from an average of $130 a user to $70 with small changes to the signup. We did this all while staying lean.

Staying lean is in fact a critical piece to avoiding a startup tarpit. It is something I wished we focused more on. In fact, I would never raise money to propel more development again. I would raise money with a functioning product already developed. The alternative is just too risky. If you raise money and make no progress except in expanding your head count and list of features, there is no way to raise additional funding—you’re dead. Why would an investor invest in a company that hasn’t shown any ability to hit some meaningful metrics?

Going head first into marketing is scary. I have come to believe it is essential though. The physics of a startup are important. To stave off inertia you have to be methodical about your resources. If you can invest early into customer acquisition you can create a sustainable business from the outset you create an engine of growth that is almost self executing and propelling you past the startup tarpit.

I hope you enjoyed this article. If you did, please share it on Hacker News or your social media feed.

If you’re interested in connecting about my new AI focused health care project Zen Patient, please email me at daniel@zenpatient.com.

For more updates from me, follow me on Twitter at @dantawfik