The Ups and Downs of Creating An API Company

Matthew Broderick once quipped in Ferris Bueller’s Day Off that “life moves pretty fast. If you don’t stop and look around once in awhile, you could miss it.” Sure, he wasn’t talking about API companies but he might as well have been.

At our 2016 Platform Summit, Johannes Ridderstedt of 46elks spoke about how much an API company can change in five years. Namely, he described his journey implementing temporary phone numbers at Blocket, a Swedish Craigslist equivalent. He addressed privacy issues when selling goods all the way through to building an API company using that technology.

Five years is a long time in tech, and nowhere is rapid change more real than in the world of APIs. Take, for example, the huge growth in popularity of REST between 2006 and 2011:

Source: InfoQ

While industry standards might change, many of the business problems faced by those building an API company (or for whom an API represents an important part of the company) are similar to how they were five years ago.

In this post we’ll be looking at a few of those API product issues and how best to address them, with a little help from Ridderstedt’s own experiences.

Watch Johannes Ridderstedt present at the 2016 Nordic APIs Platform Summit:

Download slides here

Building an API product is not the same as building an API company

It might sound obvious, but Ridderstedt highlights in his presentation that “building an API product and building an API company are two completely different challenges.”

Building a solid API product requires the creation of a service that does something – ideally something useful – and has solid documentation to explain functionality. But there’s much more to building an API company, or indeed any type of company, than that. One also needs:

  • An efficient team, with good leadership
  • Customers
  • Sales, marketing, strategy
  • Billing, usage reports
  • Time and persistence (which requires money)

The latter, he jokes, often translates to “suiting up” and pitching to investors. Perhaps that’s not surprising since Ridderstedt tells us that, 10 months after launching the 46elks alpha, they had “500 trial accounts created in the platform, but only one customer. And a total revenue of 100kr.” That’s about 12 bucks, in USD.

We’ll talk more about this below, but something that Ridderstedt highlights in his talk is the frustration among API providers that people don’t always see the value in their product: “Why don’t you get the value of our API product? It’s really, really good!” That frustration will be present regardless of whether you’re convincing investors or potential users.

For whatever reason, while just about every API provider understands the need for an efficient team and customer growth, many neglect some (or all) of the other aspects of building an API company mentioned above – sales/marketing, strategy, security, an effective billing solution etc. – in favor of focusing on creating a killer product. That’s an understandable mistake, but one that must be avoided to build a successful API company.

Build it and…they won’t come

As we’ve previously written about, to really get the most out of an API, you need to find people – either within the company or externally – to rave about your product.

This is a lesson that Ridderstedt, after creating his temporary phone number offering, learned the hard way. His thoughts soon turned to how he could sell the solution that worked so well at Blocket to other companies. However, he initially struggled to generate interest.

He attributes this to the fact that telco solutions are complicated and, particularly in the case of custom solutions, are expensive. Offering the suggestion of adapting temporary numbers for virtual dating, he explains how “suddenly there are different call patterns and different communications to users. It’s a totally different problem and requires a totally different solution.”

In this case, the solution was to remove as much of the complexity and cost as possible by offering an API instead of selling the solution as a custom product. But that still didn’t mean that the product would simply fly off the proverbial shelves of its own accord.

These things take time (and attention)

If you’ve come to this post looking for a quick fix for your API company, you won’t like these next two sections. Sorry! Ridderstedt describes a pivot undertaken by 46elks, during which they spent two years looking for a niche industry they could dominate.

He compares the less than optimal move to a hammer manufacturer focusing their attention only on a very particular type of carpenter and recognises, in hindsight, that the correct thing to do would have been to continue targeting developers more broadly.

However, he acknowledges that there are instances where a narrow focus can be more helpful. For example, during the company’s alpha testing stage, the product only worked in Sweden. So, it made perfect sense that their advertising and marketing was focused solely on Swedish developers.

When the company began to focus their attention in such a way, they found that 25 trial accounts led to one customer. Still only generating revenue of 12 bucks, but a much better conversion rate. In other words, metrics are important.

The Up and Downs of Creating an API Company

Ridderstedt shows a graph that outlines the “up and down” nature of the company in its early days. This was not only off putting to investors, who feared the company may be struggling, but made it difficult to predict future activity.

After identifying customers who spend a minimum of 2 EUR per week for 7 weeks as “continuing customers,” and using this as a true measure of their customer base, they were able to retain 95% of their revenue but without all the peaks and troughs. Undertaking this type of analysis makes it easier to predict growth and retention, refine marketing campaigns, and so on.

Don’t lose hope

Easier said than done, we know. In January 2014, Ridderstedt was grappling with 400,000 SEK of outstanding loans but had a product that tons of customers loved.

By June 2014, thanks to high retention rates and growth predictions, he was able to secure 2.6 million SEK in investment. Turns out that he must be pretty good at that whole “suiting up and pitching to investors” part of the job!

He also talks about the importance of recruiting leaders, particularly the balance of finding people with the attitude of “I want to do this” vs “I accept the task.” But how do you find those people?

Ridderstedt isn’t one to understate the importance of getting to know team members – he advocates spending weeks getting to know your co-founder before the process of creating any products truly begins: “Running a company is like a marriage. It’s not the same…but it’s close.”

Final Thoughts

One of the most interesting things about the graphs Ridderstedt brought up at our event is how steady and methodical the growth they picture is:

There’s a tendency among startups, and that includes API startups, to massage their stats to look as sexy as possible…particularly when there’s an audience! And yet, that’s not the case here. In fact, if you connected all of the dots on the graphs shown between 2013 and 2016, what you’d see is remarkably close to a straight line.

There’s a couple of big takeaways from 46elks’s story. One is the importance of targeted marketing, with Ridderstedt explaining how narrowing the company’s focus to Swedish developers (rather than an arbitrary business niche) resulted in steadier and more sustainable growth. Another is that – harking back to the assertion above – success in business takes time.

It’s evident from Ridderstedt’s presentation that two things building an API company can bring are maturity and humility. 46elks are mature enough to realise that their slow, organic growth towards the 1,000,000kr per quarter mark is something to be proud of AND humble enough to understand that they’ve faced plenty of mistakes (as all API companies will) that could have easily derailed them in the last six years.

We hope this case study acts as another validation of the power APIs have to create new business models. What has your journey toward an API company been like?