Editor’s note: This is post 3 of 3 for the Teaching Startup “Growth Series.” Teaching Startup is published each Tuesday. Author Joe Procopio is the founder of teachingstartup.com. Joe has a long entrepreneurial history in the Triangle that includes Spiffy, Automated Insights, and ExitEvent. More info at joeprocopio.com.

DURHAM – If you want to be a successful entrepreneur, you need to find a big problem and solve it. But if you stop there, you might just wind up being the inventor of an amazing solution that no one ever uses.

That’s probably not your goal.

There’s never any guarantee that your startup is going to scale to a massive valuation, or for that matter even get to the growth stage. But one surefire way to invite failure is to rest on the ingenuity of the solution and assume that growth will take care of itself.

Photo courtesy of Joe Procopio

Joe Procopio

You can’t just stop at a great solution. You have to turn that solution into an effective machine.

Now, I’ve never run a billion-dollar company, but that hasn’t stopped me from swinging for the fences. Here are the basic steps I’ve learned to get from viable solution to high-growth company.

Step 1: Perfect the scalable solution

To begin mapping the path from solution to valuation, I’m going to use two high-growth startup examples from my recent past, plus a third example you’ll be more familiar with.

Robbie Allen’s StatSheet started life as a college basketball data analytics website. Within five years, a bunch of use had turned it into Automated Insights, a 75-person pioneer in Natural Language Generation technology that produced billions of reports across multiple industries. We sold to a private equity firm back in 2015, and that story continues today.

Three years after we sold Automated, I joined up with Scot Wingo, who had previously taken eCommerce giant ChannelAdvisor from startup to public company. I wanted to learn from him as well as help him scale his new company, Get Spiffy, beyond its beginnings as a mobile car wash.

Oh, and back in 1995, Amazon started selling books online. That’s the third example, and we all know how that ended up.

Amazon and Spiffy came into existence to solve a big problem, and roughly the same problem, a lack of strategic consistency in mobile product delivery and mobile service delivery, respectively.

Automated Insights, on the other hand, was a solution in search of its big problem. We had developed the science to automate the creation of insightful written stories from data, but the question always hung over us — “Exactly who needs this tech?” That’s definitely a backwards approach, but it drives home the point that solving a big problem isn’t the only ingredient for success, or even the first ingredient in some cases.

At each of those three startups, a perfect solution was developed and waiting for traction. Amazon got a foothold in eCommerce by perfecting the delivery of books and CDs, much like Automated Insights cracked NLG by perfecting the generation of insights from college basketball data, and Spiffy is revolutionizing mobile service delivery by perfecting a few use cases in auto care and maintenance.

Step 2: Carve out a new market

All three companies started with an opportunity to take what they’d learned at a low level and solve a much larger problem.

Back to Automated Insights and our search for a big problem — it turns out that problem sort of developed around us: The overload of new data from various new sources — including IoT, mobile devices, and wearables — was creating a massive new need for data science. Data scientists were expensive, and we automated a lot of what they did with our tech.

So we started addressing a new market. We were no longer automated writers on demand, we were automated data scientists on demand. This was Natural Language Generation before we knew what NLG was. This was a much bigger challenge with a much bigger addressable market, but we had already been inadvertently addressing that market for a while, vertically, within sports.

We pivoted, and made a horizontal move to other industries, providing automated data science in finance, marketing, insurance, and on and on. Now when we asked the question, “Exactly who needs this tech?”, the answer wasn’t a single industry, it was a market segment in every industry: Anyone who had to analyze a lot of data quickly. In the mid-2010s, that turned out to be everyone.

Amazon and Spiffy approached their new markets in a much more proactive fashion. That approach included a vertical expansion into their current lines of business — fulfillment, technology, and logistics for Amazon and auto maintenance for Spiffy — as well as a horizontal expansion into other lines of business.

Amazon decided to sell anything and everything over the Internet, including bringing in third-party vendors to their platform. Spiffy began developing its own spinout software platform that could make mobile service delivery available to any service in any industry.

Step 3: Build off your intellectual property in the new market

Anyone can say they’re using technology and innovation to revolutionize an industry, including the industry incumbents. You might remember the recent craze to add machine learning to just about any product, regardless of whether it’s useful or whether it’s even viable.

But what most large companies are terrible at is focusing advancements in technology into a cohesive product strategy.

That strategy begins with streamlining and expanding advancements in every function of the business, like better ways to fulfill orders at Amazon eventually becoming 2-day shipping.

At Automated Insights, we created entire libraries to speed the development of custom automated stories, and we constantly focused on reusability where others were coding solutions from scratch every time. At Spiffy, we have a skunkworks of next-generation projects we call Automotive 2.0. These projects anticipate sweeping changes in four areas: autonomous vehicles, electric vehicles, connected car, and vehicle ownership.

These advancements aren’t scattershot, rather, they’re calculated initiatives to make sure we’re marching toward solving the big, big problem. They revolve around concepts like MVP, revenue, CAC, LTV, and market share. They measure the results, build on what works, and discard what doesn’t.

Step 4: Maintain your lead

Every successful first-mover startup usually finds itself entering its growth phase with a 6-to-18-month head start in creating and dominating their new market. The pain and struggle they go through in the previous three steps results in scars that are expensive that build moats that are wide.

But that’s not enough.

That lead must be maintained and extended by constantly cycling through those advancements in intellectual property, extending the resulting innovations to other markets and industries, and pivoting off of them into riskier but hugely rewarding initiatives.

It’s why Amazon repurposed Amazon Web Services, an internal technical initiative to help get them through their busy holiday season, and spun that out into a service offering for other businesses. Amazon wasn’t in Microsoft’s or Google’s face ten years ago. Now they’ve got a huge lead in cloud computing.

And it all started by solving a problem with shipping books.