Regular readers will know that I believe the price of content, including software, is trending towards the marginal cost of delivery, which is zero (note that doesn’t mean it will necessarily reach zero). Innovative companies are responding with innovative pricing models, of which ‘freemium’ is perhaps the most common. This is now a pretty well trodden path for consumer applications where free to use has been very common in social media and content for sometime and following the success of anti-virus companies is increasingly the default model for software as well. The logic of free applies equally in the B2B market but the inappropriateness of advertising and many other ‘third party pays models’ combined with typically the more limited market size has led to slower takeup of freemium models in the business to business sector.
That is now changing rapidly, and many B2B software businesses have now either adopted or are actively considering a model which makes at least part of their offering available for free, including those of a number of companies that we have invested in and sit on the boards of. This is particularly true of those which are still young and in the phase of seeking to maximise growth and dominate their market. Many VC backed businesses, and companies seeking to raise venture fall into this category.
There is a great post on Techcrunch this morning which sets out a number of the considerations for freemium models. The list below takes the highlights of that post from a B2B perspective:
- A truly great product is a pre-requisite for a successful freemium model (that said, a truly great product is increasingly a pre-requisite for success, full-stop – see Brad Feld this the centrality of user experience)
- Value perception – Freemium works best if customers believe more in the value of a product the more they use it (and it would be a disaster if customers believe more in the value at the point of initial engagement – in the consumer segment many health services are like this, e.g. gyms)
- Value of free users – the free users should have a value that can be calculated, even if it is just a reduction in the cost to acquire a paying customer
- Cost to serve a free user – a freemium model will be challenging to make work unless the cost to serve a free customer is minimal – i.e. limited to server and bandwidth costs and without excessive storage or any content/licensing expenses
- Market size – the total number of users needs be large enough that if a high percentage of them use the service for free there will be enough payers left, either that or the free version offering isn’t appropriate for the target paying users (see below on revenue impact)
The major additional consideration that wasn’t mentioned in the TC post is what the competition are doing. If the competition has a free offering which includes your most compelling feature, or even just 80% of it, then you should think carefully about whether you are going to be able to sustain a paid model. Note also that the customer often sees more value in the basic features than their technology supplier, and correspondingly less in the advanced features that often form the basis for differentiation.
Companies with existing customers and revenues will need to consider the revenue impact of adopting a freemium model. On the debit side, some paying customers will immediately switch to the free version and some new customers who would have paid will opt for the free version. On the positive side the cost to acquire paying customers should drop as potential customers should be easy and cheap to recruit to the free version and then enough of them will be wowed by the product to upgrade to paying. The freemium model is only worth adopting if the positives are greater than the negatives and there is is sufficient resource in the business to cover the period of time between the initial fall in revenues and the freemium model starting to work properly.
Finally, as Eric Ries and the lean startup movement has taught us, the process of building the perfect startup is one of continual iteration and measurement. The adoption of a freemium model should be regarded the same way. It is a big decision for a startup to go down this route, but it isn’t an irrevocable one, rather the benefits should be measured, and if they fall short of expectations then the model should be changed again, and that might mean dropping freemium.