How to Sell Open Source

You've heard this sentence so many times: "Free and open source software like Linux doesn't need a marketing department!" To some extent, there's some truth to it: People learn about FOSS thanks to word-of-mouth and grassroots movements. Many of us discovered Linux from friends and colleagues or from reading about it in online discussions, rather than from flashy TV adverts. We keep using it because we like it, not because our brains are being toyed with. We'll keep using, promoting, and supporting Linux regardless of what happens to the companies developing and using it.

On the other hand, Linux and FOSS exist in a highly competitive (and often unpleasant) market. If you want it to grow, you need to consider the marketing strategies and approaches that the competitors are using and combat the "fear, uncertainty, and doubt" that's often spread by groups and individuals that want to cause FOSS harm. You want to present the software you love in a positive light, even when things are not going so well (e.g., with the Heartbleed security vulnerability).

Marketing open source is something everyone can do, whether it's using social media effectively, creating videos or infographics to explain FOSS, or attending events and giving talks. Over the next few pages, I'll examine some of the traditional approaches to software marketing, look at examples from Firefox and IBM, and provide some tips that you can use.

Open vs. Closed

Imagine you're a newly appointed marketing manager for a proprietary software company. What's your responsibility? Well, you'll want to examine your current product portfolio and see how those products fit into the market (e.g., who they're targeting, whether there are growth opportunities, and whether the pricing is right). You will also look at products in development and see what opportunities are ahead. You can then work on marketing materials to describe, promote, and sell your products to the right people. As you become more experienced, you can work more closely with product developers to shape the final result.

Now, imagine you have the same role, but within an open source project (or a company that sells FOSS). Some of your responsibilities will be the same, but others will differ significantly. For instance, when an open source program is being developed by a large community of volunteers around the world, you can't easily shape its development. Sure, you can do market research and find out what features people need – but then what? Post them on a mailing list and see what happens? When developers are scratching their own itches and not being paid by you, it's harder for you to push development in the direction you want.

Similarly, if you're not selling the software as a product but rather support contracts and other services around it (like many notable FOSS-friendly companies), then your approach has to change. And on top of that, you don't just want end users to try your product, you want potential contributors to get involved, as well. So your marketing job ends up stretching out into community outreach and management.

Classic Example 1: IBM

Back in the late 1990s, GNU/Linux was emerging as a decent server operating system and gaining some use in small companies and ISPs, but it still lacked major commercial backing. Various companies in the FOSS ecosystem supported it – such as Red Hat – but they were small fries at the time. Linux was still very much the domain of geeks and hackers, and it didn't have the image or money behind it to really take off.

That is, until IBM got involved. In 2000, IBM made a major announcement: The company would spend $1 billion on Linux development over the next year [1]. The reason? According to IBM chief executive Louis Gerstner at the time: The company "is convinced that Linux can do for business applications what the Internet did for networking and communications." IBM already had its own gamut of operating systems, including a flavor of Unix called AIX, but it was showing serious commitment to the open source fledgling by making the decision to back Linux.

While that $1 billion went toward improving GNU/Linux and other FOSS tools, IBM also started marketing Linux. Most notably, in 2004, the company produced a 90-second video advertisement called "Prodigy" that can still be found on YouTube today [2].

In the video, a small boy is sitting alone in a white room (see Figure 1). Over the 90 seconds, various famous people approach the boy and offer him advice in all walks of life – even Muhammad Ali makes an appearance, telling the kid: "Speak your mind. Don't back down." While all of this is going on, two unidentified observers are making notes. At the end of the video, one asks the other: "What's his name?" The response is: "His name is Linux"; then the advert ends.

Figure 1: IBM's "Prodigy" advert in 2004 introduced Linux to millions of people across the USA.

The video is striking and creative, and it takes a novel approach to marketing software. IBM didn't shy away from positioning Linux as the young and relatively inexperienced contender, but it focused on its ability to learn and absorb information (i.e., through the open source community).

What's especially notable is the positioning of this advert. This wasn't some cute in-joke for geeks at a conference or shown internally at IBM meetings. No, it was shown during the 2005 Super Bowl – one of the most viewed events in the world. IBM was reaching out way further than pointy-haired IT bosses. IBM wanted everyone to know that it was serious about Linux.

Classic Example 2: Firefox

Still, IBM was focused on large Linux deployments in enterprises. If you want an example of consumer-oriented FOSS marketing, you can turn to the Mozilla Foundation's spectacular push for the mass market in 2004. Firefox 1.0 had just been released, Internet Explorer was still hugely dominant, and Firefox was seen as a way to introduce the masses to FOSS – ideally setting them on the path to explore more and maybe even try Linux in the end.

The Mozilla Foundation set up a site called Spread Firefox and sourced donations to be used for marketing. The end result was a two-page advert in The New York Times (see Figure 2) that asked readers if they were "fed up" with their web browser – and showed them an alternative. Finally, they added some quotes from users, a snippet of information about the software, and a download URL.

Figure 2: Mozilla reached out to disgruntled web surfers with a two-page advert in The New York Times.

The advert obviously had an effect: Firefox tripled its market share over the following 18 months, breaking the 10 percent mark in June 2006. (It peaked at 32 percent in 2009.) Of course, other factors were involved as well, but it's a great example of how a FOSS project can reach out to a wider audience in a clear, targeted, and effective way.

Risky Approaches

When advocating and marketing open source software, it's very tempting to make bold claims and sweeping generalizations. How many times have you heard the statement "Open source is more secure than proprietary software" (or some variant on that), for example? You know there's some truth in that statement; history has demonstrated again and again that proprietary software is often riddled with security holes that are exploited by not-so-nice people for years before they become public.

However, you have to be careful. You can say "open source is more secure," and then something like Heartbleed [3] happens (Figure 3). If you're out of the loop, Heartbleed is a whopping security vulnerability in the OpenSSL crypto library, which is used pretty much everywhere. The bug was disclosed in 2014, and vast numbers of websites were affected. It was so bad that the OpenBSD folks forked the library to do a massive cleanup, in the form of LibreSSL.

Figure 3: Be careful when marketing FOSS using "it's more secure" arguments! Things like Heartbleed can still happen.

Most importantly, though, those of us who'd trotted out the mantra "open source is more secure" ended up with egg on our faces. Yes, we still believed that statement, on the whole, but Heartbleed provided nay-sayers with a big chunk of ammunition: Open source is also insecure, open source has had vulnerabilities sitting in the code for years, and so forth.

Therefore, it's important to state that open source itself doesn't just magically make things secure; rather, it's the development process. You could even argue that OpenSSL was barely open source to begin with – yes, the code was there, but hardly anyone was working on it, and those who poked around inside tended to run away screaming. The oft-quoted line "Given enough eyeballs, all bugs are shallow" didn't really apply here.

With this example in mind, you shouldn't say "open source is more secure"; instead, clarify it somewhat: "Open source development processes generally lead to more secure software." It's not as snappy, and it leads to more questions, but it's more honest. I think the same phrasing should be applied in terms of reliability as well – Linux adherents know that GNU/Linux is pretty rock solid, but they need to remember that other users may be affected by corner case bugs. As tempting as it is to poke fun at Microsoft for "blue screens of death," Linux is not competing against Windows ME any more. Windows has plenty of problems, and I'm glad Linux can avoid them, but for many users, it's pretty reliable.

Money, Money, Money

Another area where caution is required is price. Yes, GNU/Linux is free – as in beer – to obtain, but many other factors come into play when determining how much it really costs. For instance, if you need to replace a piece of hardware for something that's Linux compatible, that bumps up the price. If you want to deploy Linux (or other open source software, such as LibreOffice) in a large company or government body, end users might need training – which also affects the end price.

For this reason, many IT purchasing people talk about total cost of ownership (TCO). How much will it actually cost to use program X, given the required hardware changes/upgrades, staff retraining, and support costs? In the short term, this doesn't always look so great for Linux and FOSS – if a large company is already using a proprietary system, perhaps getting large discounts for bulk purchases, then a switch to a FOSS solution could initially be very costly.

So you need to bear this in mind when marketing FOSS. Don't just say "Linux is free," because pretty much every large deployment will require support of some kind, whether from the likes of Red Hat, SUSE, or Canonical or from in-house IT staff. A better approach is to say: "Over the long run, Linux/FOSS can reduce the total cost of ownership, with no need for license fees, along with improved reliability and security." Again, you may have to qualify the "reliability" and "security" parts somewhat, but it's the right way to sell FOSS.

Also on the subject of support: A good marketing strategy is to highlight that, with FOSS, you often have many more choices for support. As an example, if a company has thousands of PCs running Microsoft Office and a serious bug is affecting many users, what can the company do? Call Microsoft, maybe spend some more money, and hope that it gets fixed in the next release? If that doesn't work, they're up a certain creek without the slightest hint of a paddle.

Contrast this with a company that deploys LibreOffice: If a bug is affecting workers, the company can choose from many certified developers [4] and hire one to (one hopes!) fix it. Sure, it still costs money, but the company is not reliant on a single vendor and can shop around for solutions (including local ones). It's the free market at work – so much for FOSS being anti-capitalist, as some people used to say!

What Can You Do?

Many FOSS projects are in desperate need of marketing. Maybe their website is subpar, with no proper information on what the app actually does (alarmingly, a common occurrence). Maybe the project is making great progress but is not communicating this effectively with the outside world. Perhaps the project could reach out to new users and potential contributors via social media, Reddit, or Hacker News, but nobody involved has the skills or know-how.

Of course, most FOSS projects don't have the budget to hire a full-time paid marketing person, but if you're interested in this field, dive in and offer to help. As with all things in FOSS, you don't need to be an expert in the field – any help is appreciated. Plenty of books are out there that provide an introduction to marketing, and you can take those basic skills and techniques and apply them to the open source projects you use and want to support.

You can also look at the materials and processes used by large and established FOSS projects. For instance, the Mozilla community has a marketing guide [5] and other materials, including web and print materials [6] (Figure 4). Fedora, meanwhile, has a marketing team that's based around a mailing list [7]; in LibreOffice, there's a small community working on presentations, press releases, and other materials (Figure 5), and it organizes through calls once a month [8].

Figure 4: Mozilla's marketing community even has its own mascot.

Figure 5: Look at existing FOSS projects for inspiration – for example, the leaflets produced by LibreOffice's marketing team.

Of course, in the process, you'll build experience that you could potentially use in a paid marketing job some day, and if you end up rich with your own private island, don't forget about the humble Linux Magazine writers who set you off on your new career path.