As a startup, your choice of programming language may not be top of mind for potential investors, but it can affect funding potential under some circumstances. We asked tech investors about factors to consider when making that choice.
For Federico Wengi, a venture capital associate with Paua Ventures in Berlin, startups should consider their comfort level with the language, where the market is going and if the language they choose is the best one to reach their goals. While VCs usually don’t try to change decisions around programming languages, the way programs are written could affect investment decisions.
“If we don’t like what we see in the tech [due diligence], we walk away,” Wengi said. “It is important that we can establish proof that the team has built a sustainable backbone of tech on top of which the first customers can be won.”
A CTO, for example, might be comfortable using Scala, and that language might also be suitable to reach the startup’s goals. “But once you scale who are you going to hire?” said Wengi. “According to GitHub, Scala is one of the least popular languages. On the other hand, if you are building a robotic prototype you have little use of a language like Java, regardless of the quantity of talent you might be able to acquire.”
Get the TechPortfolio briefing
The month’s fintech and funding news, plus new tools for founders, in one short email.
For Leo Polovets, a partner at Susa Ventures and former software engineer at Google and LinkedIn, the programming language isn’t a factor for his fund unless the choice is clearly inappropriate. “If someone is trying to build cutting-edge deep learning APIs, then PHP is not the right language for that,” he said. “But PHP, while somewhat out of style, is perfectly fine for many other applications.”
The most popular programming languages for startups are Javascript, Ruby, Python and Java, according to Polovets’ analysis of AngelList data. He found that JavaScript is by far the dominant programming language choice for startups, followed by Ruby, Python and Java; Ruby on Rails is the top choice for front-end technology.
Polovets also correlates stronger startups with modern or functional programming languages, such as Go, Scala, Haskell, Erlang and Clojure. PHP, on the other hand, is on the decline, and several sites list developer pain points, such as PHPWTF and PHP Sadness.
Overall, though, VCs generally have no say over a startup’s tech stack, said Polovets. Most VCs are adept at evaluating founders, markets and business models, but don’t have the tech knowledge to recommend programming languages.
“Most investments happen after there’s already some code in place, or even a live product,” Polovets said, “Changing programming languages at that point would be a huge, unnecessary time sink.”
Frequently Asked Questions
Does a startup's choice of programming language affect investors?
Usually not directly, but it can in some circumstances. Federico Wengi of Paua Ventures said his firm walks away if it dislikes what it sees in technical due diligence, because the team must show a sustainable tech backbone.
Why might choosing Scala cause problems when a startup scales?
Wengi said a CTO may be comfortable with Scala and it may suit the goals, but hiring becomes hard at scale because GitHub data showed Scala is one of the least popular languages.
When does Susa Ventures consider programming language a factor?
Leo Polovets of Susa Ventures said language only matters if the choice is clearly inappropriate. For example, PHP is wrong for cutting-edge deep learning APIs but perfectly fine for many other applications.
Which programming languages were most popular with startups?
Polovets's analysis of AngelList data found JavaScript was by far the dominant choice, followed by Ruby, Python and Java. He also linked stronger startups with modern or functional languages such as Go, Scala, Haskell, Erlang and Clojure.
Why don't venture capitalists usually tell startups which language to use?
Polovets said most VCs are good at evaluating founders, markets and business models but lack the technical knowledge to recommend languages. Most investments also come after code or a live product exists, when switching would be a huge time sink.