Freemium and open source business models. Friends or Foes?

Basics of the freemium model

The freemium model started being popularized in the 1980s, frequently under the term “shareware”. The term freemium model became better known in the past decade. Freemium stands for the combination of “free” and “premium”. It refers to a licensing model that starts at free in the hope to gain broad traction and then convert a subset of users to a for-pay, premium version of the software or service.

In the era of information overload and too much choice, the freemium model promises an awareness and marketing boost by enabling a broad and lower friction adoption model. Generating the necessary pull to make a freemium model work is critical and requires providing enough value in the free edition to drive broad adoption and preferably word-of-mouth based adoption.

To then encourage free to paid conversions it is important to expose some of the value-add premium capabilities within the free product. This is usually accomplished by exposing premium functionality with some limitations. The ability to use the product perpetually for free is what makes a freemium model different from a free-trial model. In a free-trial model the buyer knows they will need to make the buying decision within a matter of weeks.

Making a freemium model successful

The trickiest aspects of making a freemium model successful is to get a number of levers right:

  1. Provide enough value in the free version to ensure broad (preferably viral) adoption.
  2. Have clear cross over points from free to paid so that the target prospect will have strong motivation to pull the trigger on going premium. There are many ways to differentiate free and paid including storage space, bandwidth consumption, features, support SLAs, and capacity.
  3. Implement effective premium tiering to drive up the average selling price and tailor to multiple prospective buyers and use-cases.

To motivate a content free user to move from free to paid it is important to have big, must-have reasons to make the purchase. Preferably the user hits some limitation which is important to them which creates an impeding event. Incremental motivating drivers (should-haves or nice-to-haves) may not create enough urgency when the free version is “good enough”.

While in the free-trial model the trial expiration is a hard limitation which forces a buying decision, the same does not exist in a freemium model. This is why freemium models tend to work best when there are some significant hard, must-have reasons to upgrade such as more storage space needed (e.g. Dropbox or Box), data volume limits (e.g. Splunk), and number of servers to meet a needed amount of capacity.  When such clear reasons exist it tends to be easier for the vendor to give away more features for free to drive the broad adoption of the product. Also, in such situations there is less cannibalization risk when the target audience are the ones making significant use of the product.

The quality of implementation of the freemium model within the product is also a critical success factor. Not only is it best to find ways to expose the value-add within the free version but there should also be a smooth, low friction path from free to paid. To ensure that happens, the product itself should be upgradable, preferably in product, so that the user can continue their work right away. This is easily achieved when delivering a software-as-a-service, but in the shrink wrapped software world it typically requires the free version to ship with the premium features ready to be turned on with a change in license key. Needless to say it can get quite tricky to implement an effective freemium model within the product and there are many nuances making it successful.

Freemium as part of an open source business model

Monetizing open source projects via freemium models is even more complicated. Many open source vendors monetize by developing proprietary value-add functionality to the open-source software. The value-add in conjunction with support SLAs is what drives conversions from open source to paid. The open source business model on its own is not a true freemium model as the open source project usually does not expose the vendor value-add. In a significant amount of cases the open source vendor is not the one who controls the distribution of the open source project and/or cannot for political reasons bundle proprietary value-add within the open source project. Therefore, the open source software is not equal to the free version in a traditional freemium model.

Hence, in order to implement a true freemium model in an open source market the vendor often needs to have a “community edition” (CE) which is a free, value-add version of the open source project. The goal would be for the CE version to be the way the user adopts the open source project and at the same time expose some of the premium value-add of the premium version.

However, this exposes a number of challenges which are unique to freemium models as part of an open source business model:

  1. There is one more level of differentiation that the vendor needs to tailor to which can ultimately dilute the value-add of the premium version. The vendor is not only contributing functionality to the open source project to ensure its ongoing success (the foundation of the business) but also, in addition, they need to give away functionality that differentiates the CE version from the open source project to create enough motivation for the user to obtain the OSS project via the CE product. In this open source differentiation model there is one additional “free” tier that needs to be differentiated which means that the vendor’s value-add functionality is now spread across three tiers (open source, free and premium) which makes it harder to retain enough must-have value for the premium editions.
  2. As noted earlier, the most effective premium value-adds in freemium models tend to be related to capacity whether storage, servers, bandwidth or other hard limits on workload or data. These hard limits create a compelling event where the target user needs to make a buying decision. However, in an open source model the often most effective workload related limitations are irrelevant because the base open source project can be deployed in an unlimited fashion, think of MySQL (unlimited storage), Hadoop (unlimited compute), or Lucene (unlimited indexing). This is probably the biggest reason why freemium models are difficult to implement in an open source market. There can be workload related limitations to the proprietary value-add (e.g. APM limitations, servers under management limitations) but the base workload which has made the open source project so successful cannot effectively be limited as they are not limited in the open source version.

While I am not saying that a freemium model can never work in an open source business model, most of the successful open source companies have implemented some form of free trial model vs. true freemium model.. This approach creates clearer differentiation between the premium product and open source project. This is especially important as workload based limitations do not typically work as well in open source freemium models so all the possible value-add needs to be focused at differentiating the premium version from open source vs. diluting the value-add with a second free version between open source and premium.

By example, in 2003 Red Hat eliminated Red Hat Linux (their free distribution) in favor of Red Hat Enterprise Linux (subscription-only binaries).  With that they eliminated that middle free tier between open source Linux and a premium Enterprise offering. There was no longer such a thing as a commercially-blessed but not supported Red Hat distribution. That change enabled Red Hat to become a billion dollar, high margin company. Red Hat must have realized that it is too difficult in an open source model to differentiate your product twice. It worked!

All this does not say you cannot build a very lucrative business on open source. Many have done so including Red Hat, Cloudera, MySQL, Zend and others, but I do believe the freemium model may not be the best fit for many open source companies. In most cases, open source companies will be best off focusing on two initiatives – making the OSS project successful and concentrating their value-add on the premium offering. And in cases, where it makes sense to deliver some of the value via a service, even better. In those cases, workload capacity can be an effective limiting factor.

Would love to hear your thoughts in the comments below. In what cases do you believe a freemium model does or does not work as part of an open source business model?

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s