Open Source Wins Time for Innovators

In this article we explain the motivation for us to use, build and license our software under open-source licenses. While the open-source contribution to the innovation community is manifest, finding a viable business model for open-source makers has been a discovery area in the recent two decades. It is partially thanks to the shift to cloud computing, that open-source has been recognised as advantageous even by funding community. 2017 and 2018 IPOs have proven the open-source producers to be on the right spot. We wrote this article from position of a company that is basing its model on open-source and whose founders were involved in an open-source project from 2000 that ended with a successful acquisition in 2005.

We like open-source because of its profound effect on innovation. Most of innovators are hindered, time-wise and money-wise, by what we call the "platform burden". Every innovator is keen to focus on the specifics of his value proposition. However a great part of a product consists  of various "platform" components: OS, GUI, monitoring, simply all  the stuff that has nothing to do with the product's value proposition, costs 90% of development time and must be there. Linux has been a notorious example of solving the "platform burden". Innovators can use it and change it without having to implement their own OS or negotiate with an OS company about changes to their kernel. Open-source buys the innovation community time: one can use the platform software instantly and one can change it to suit his application. Time is everything for innovators and so is open source. 

While benefits of open-source to its users are obvious, the more difficult end is that of funding the open-source producers. We met this problem first time in  2000 when we started work on an open-source telephony project (SIP Express Router, AKA SER) at the research institute Fraunhofer Fokus. The development funds  were secured by  various research projects which relied on use of the software. The software matured over time and was open-sourced mainly due to concerns that lack of marketing funds  would make it short-lived. Open-sourcing indeed helped to create exposure and it soon began to be the SIP software that was deployed whenever scale mattered. Eventually we started a company that generated revenue using dual licensing for OEM customers,  closed-source extensions  and support -- the traditional methods to which most of open-source companies resort till today. Commercialisation has not been entirely easy though.  For many customers, the notion of  "open-source" evoked "free-of-charge" or even worse something "worthless". Fund raising was similarly difficult as many investors equaled open-source to IPR vacuum; remember this was years before the 2008 $1 billion acquisition of mysql by Sun.

The perception of open-source has changed since then favourably and open-source companies regularly participate in the funding process. Alone this year, 2018, Elastic made an IPO ($252m),  Microsoft acquired GitHub ($7.5 billion) and Salesforce acquired MuleSoft ($6.5 billion). In the past year, 2017, mongodb (database company) made its IPO and so did Cloudera (commercialises Hadoop).

Prevalence of clouds has improved the situation for both the open-source producers and users. There are two reasons. One of them is that offering a service requires the software to be polished and configured to perfection without any DIY effort left. This addresses one of the most frequent critique points against open-source which is additional effort needed to make it run (AKA the "Total Cost of Ownership weakness"). When the software is offered as a service, the remaining DIY effort diminishes and the TCO argument falls off. 

The other aspect that makes the open-source life-cycle easier is that intellectual property rights are becoming less relevant. Even "viral" GPL-ed software can be used without disclosing the essence of a cloud company value which has shifted from raw software to the service space where software integration and operation matters most. That is making things also easier for investors: due diligence of a service is far more telling than that of raw software.

To be fair, even in the cloud the software rights may be a business challenge.  It can be shown on the case of Elastic. Elastic has been producing  open-source database software "ElasticSearch", also offered as a cloud service, and its commercialisation efforts have been crowned by a successful IPO (94% price increase on the first trading day). The less the company spends in marketing, the more it puts in R&D. Yet in its SEC filings, the company had to warn that AWS has built a comparable cloud service based on ElasticSearch that competes against Elastic (see http://techgenix.com/elastic-ipo/). The company has been countering this situation by offering open-sourced extensions under commercial license terms.

In summary we think the Elastic experience represents well the current cloud-age open-source situation. The open-source companies cut their innovation time on the usage side. On the producing side, they save their marketing expenses at the cost of sharing some software with potential competitors. Yet the code-sharing aspect is by no means a show-stopper because in clouds the core value already resides in integration and operation -- a situation that has been been recognised even by analysts and markets and helps to secure funding. Which will add even more fuel to open-source projects.