Thursday, July 17, 2014

The Freemium Business Model: as an affordable loss strategy of founders

The Freemium business model is one the most commonly used revenue realization streams by the companies offering web-services. The promise of a large user base, and cross-diffusion of free users to the paying customers; makes freemium the most sought after business model in the internet domain. The utility of this model, however, is not without concerns - Ash Maurya  a lean methodology proponent, in his land mark book highlights the drawbacks of using this model by a start-up. In this blog, I extend the conversation on Freemium as a business model beyond the drawbacks suggested to look at it from the lens of effectuation theory, and use the affordable loss principle to explain the wide spread usage of this model in the web-service industry. 

The term 'Model' refers to an abstraction of the a real life phenomenon. The "business model", is then a abstraction of the way a business operates. Wikipedia[1] defines Business model as: "A business model describes the rationale of how an organization creates, delivers, and captures value". A business model is simply a plan that the company executes to generate its revenues! There are number of business models that companies could adopt from - I have tried compiling some of these here <<more on business models>>, the explanation of the freemium model elsewhere is here

It is easy to note that one often finds multiple business models being used by a single business. We donot include other models in our discussion but acknowledge the possibility and only introduce a plausible reasons for such an approach: Companies operate in a sea of market uncertainty - and plan to reduce this uncertainty in a sytematic way. The companies always attempt to reduce the uncertainty by actively engaging with the environment, anticipating changes and preparing for the same. In these experiments to understand the organization's environment - the company creates various business models which that could not just help the company identify its customers but also serve them better. We could think of these business models then as complementary to each other and assisting the business generate a higher revenue than through individual business model.
Returning to our discussion on the "Freemium" business model, we could understand the model as one where  the company gives away a chunk of the offering (product/service - I do not comments on quality) to the users for free and attempt making the larger portion of its revenue through the paid higher end of the services. For example: In a professional networking website like Linkedin, the basic business networking service offered is  for free. Linked in also offers its users are a free access of the - "Linkedin Premium" service for a short duration of time after which the user could return to the free version or cotinue the premium being a paid user.
It is not surprising that the conversion from free users to paid premium services is small, but that is one of the means the company uses to generate revenues. I have only highlighted the premium offering from Linkedin, however this not the only business model that Linkedin uses - there are revenue generation opportunities like the advertisements or analytics or query results etc that could be of significant monetization impact for the company.

Reiterating the point I made earlier - the widespread usage of the freemium model by web-service companies gets one to think of freemium being a necessity in the web enabled world than one of the options. After all, users would love to try out the website before they begin paying for the same; isn't it? and, its is but logicall to expect to expect convesion to the paid users category - even if a small portion converts, it would be a significant revenue!
The almost zero cost of reproduction of a web service helps the cause.  
It is this attractiveness of the model that Ash Maurya cautions his readers - the "lean start up" enthusiasts, against. To understand the caution - lets return to analyzing the Linkedin example in greater detail.
Linkedin is a professional networking site - where people connect up with other with the intention of growing in their profession and also engaging professionally others having a similar intent. This professional network has millions of users and without information - finding the connects you might like to engage with would be a herculean task if there is no information about the user is available.  The solution - create your basic profile when you sigh up for the website. The basic information provides to data - that could make networking exciting and useful. To realize the importance of the data - imagine a the website users not update their profile pages! You would not get any "relevant" search results that you could use - If this prolongs for a long duration - the newer sign ups on the website would leave saying - "There is nothing in here that would be of interest to me" and move one...possibly never to come back!
Linkedin leverages the free component of its offering to create a valuable data for itself - which move beyond holding the interest of the group together and create a foundation of sorts for the premium services! The large user base acts as the fuel for growth. The psychological attraction to the "FREE" offering is really creating a wealth of data that could prove to be valuable information for the company's operation.
Observe here that it is always possible that there are always more unintended and inactive people (lets call these "Noise") to the active people and further a even small proportion of premium service users (lets call them "Signal"). The signal is what we intend to have, while noise is what we need to decrease - there by improving the signal-to-noise ratio.
The Lean Startup - pushes the entrepreneur to reduce "waste" by eliminating distractions - and getting a niche that the company could eventually build on. The signal-to-noise ratio, thus has to be very high when starting off and the freemium model could be damaging in this regard. Ash Maurya in his book -  "Running Lean" raises few more apprehensions when using this model as start ups [2]:
  • Low or No Converstion
  • Low Validation cycle
  • Focus shifts to the wrong metrics
  • Low signal-to-noise ratio
  • Free users aren't "free"

The crux for the lean methodology of starting your venture is - "move from plan A to a plan that works", before running out of resources.

The widespread utility of the model would possibily indicating something beyond the risks highlighted by Ash Maurya. I use the Effectuation theory to potentially explain the large scale adoption in the web-services industry. By this, I don't intend to claim that the drawbacks arent true - it is another way of explaining the phenomenon. 

Effectuation has been found to be a good answer to manage the Knightian uncertainty that one faces. We have discussed this earlier in here, for more on the topic, you could join the online community at 

Most web-service founder have an expertise in technology, and less often ones with a market experience - which makes the product development aspect of their business relatively flexible and agile to experimentation. Their technical strength is a possible afforbale loss in terms of the effort, this is leveraged pretty heavily by these founders. Note, I am not considering the time and monetory affordable losses at this moment - the entrepreneurs might discount this if they are running these in parallel. Also note that the assumption I have made is one of a purely technical co-founder - if a market oriented founder starts the venture, it might not be same scenario. 

Remember that for a successful business - the technology and market are both mandatory; however the technical founder might trade off the technical provess and patience to gain the market experience here. The othe benefit of the affordable loss approach is that it gets the team going on their venture and not get stuck into the analysis paralysis. They are fighting the real world and picking up the neccessary experience and learning instead of just slogging it out with ideas in the fictious mental space where the assuptions identifications activity become difficult. 

For example, we could think of the FaceMash as one such experiment (I am not commenting on the intention of the developer in any way - I interpret this as an experiment).  The outcome of the expriment was for sure something that even the founders didn't anticipate! 

The early experimentation - help not just identify the underlying current of the business, but also open up enormous opportunities for future growth of the company [Akin to the Corridor Principle]. Recollect here another principle of efffectuation: when life throws lemons at you, make your lemonade.
We see indications that there are a lot of factors that emerged as a consequence of the experimentation mindset. This may be a the identification of the latent need by someone else, or the lack of exclusivity - which could aid the "stickiness" of the website, or the learning of what territories not to venture into. 

The learning from the experiment doesn't limit itself to takeways on features but the interaction with the various elements - which allows identification of new opportunities. Recollect also that the interaction that happen subsequently are equally important - though difficult to estimate apriori the impat of these interactions - they could really mould the direction of the start up. 

All is not hunky-dowry in the journey there are surprises - call them sour lemons if you like - but the entrepreneur and his attitude help take charge and act more akin to the "pilot-in-the-plane" (the effectual terminology), or a driver handling the streering wheel; rather  than the scientist how tries controlling the rocket sitting remotely. You need to realize - you can smash the lemons, extract the juice to make your lemonade! 

What I havent highlighted is the entreprenuerial mindset - the essential components of this being reselience and learning. Its not easy to just let go of something you started off because it isn't functioning as expected. The key is it fix the mind to an affordable loss - instead of optimization of effort (may not work in extreme uncertainty). One has to continuously strive towards sensing the results of an experiment early - fail quickly and learn the most from these experiments. 

Freemium then could be considered one such experiment with an affordable loss and set of metrics that helps the entrepreneur measure the progress with the expectation.   While the expectation correction  and restrategization is a key for entrepreneurs to work on their idea, it is the variety of the experiments that could be beneficial too - one could over a period of time expect to develop an entpreneurial experitise. 

To conclude - the use of freemium business model is definitely cautioned against by Ash Maurya, but its utility might be better understood if one applies the afforable loss lense to see the same experiment.  The key is the mindset one has to develop as an entreprenuer who continuously senses, thinks, and acts.
  1. The wikipidea definition: (Accessed 10-07-2014)
  2. Running Lean - Ash Maurya - Second Edition - Pg No 181 - 183


  1. Nice attempt Sachitanand! Being an uninitiated reader, I had to call Google mata a couple of times. Otherwise, I think it is a very well thought out analysis of the subject. Secondly, I liked the suggestions you gave for a new entrepreneur - like resilience and learning, experiment early. Finally, I felt LinkedIn is a very good example for this analysis.

  2. Interesting analysis Sachi. It is an interesting exchange of gains & losses for consumers and companies. Can we potentially apply game theory here? I believe the freemium model would be a success only when there is a win-win situation between the consumer & the company.