"Making 'Freemium' Work" by Vineet Kumar in Harvard Business Review (5/14), nicely sheds light on companies' real world experience with freemium, highlights the challenges of making it work, and suggests how companies can tune it to get good results.
Here, I expand on my online comment on that article that refers to the HBR Blog post I did with Marco Bertini, introducing FairPay to HBR readers. Excerpting most of my comment on Kumar's article:
...a new strategy beyond freemium that addresses the same business needs, and exploits the attraction of "free," in a way that promises to be more powerful and flexible. We motivate this as moving the exchange between seller and buyer from the transactional to the relational, based on three pillars of relationship: empowerment, dialog, and experience/reputation. The six questions raised here are important to this new strategy as well.
Freemium has offered a good start to dealing with new economics of digital offerings, but this new strategy, called FairPay, takes the driving objective of freemium - exploiting "free" to move toward a profitable relationship over time - and makes that the driver of a variable boundary between free and paid tiers. FairPay moves the strategic question of what to exchange at what price from a pre-set seller decision to an emergent, dynamic process that balances the interests of the seller and different individual buyers. That provides more nuance and flexibility than freemium's gross segmentation into just two (or a very few) set tiers (free or paid).
We suggest the more individually adaptive techniques of FairPay can be applied to get better market reach and profit, and to build deeper and more profitable long-term relationships. The six questions identified in this article remain important: FairPay provides a systematic method for (1) adapting (and softening) the boundary between free and paid, focusing on (2) customer understanding and (3) conversion rates, (4) evolving over adoption life-cycles, (5) encouraging referral incentives and communications, and (6) guiding ongoing innovation. With both freemium and FairPay, we see an opportunity to move beyond the debate over free vs. fee, to focus on empowering and communicating with customers and finding ways to reward those who opt to pay.How FairPay does this is more fully explained elsewhere in this blog and related Web site (see sidebar), but here are some comments specific to Kumar's six questions.
- What should be free? With freemium, this is a very visible static parameter that is hard to guess right, and hard to change. With FairPay, the boundary is soft and dynamically individualized. The buyer sets it for each transaction, always getting a price he considers fair, but the seller controls future offers, thus providing a balancing force to drive the buyer to price in a way that both sides consider fair, over the course of a series of transactions that build toward a profitable relationship. The seller need not shoot in the dark to set a rigid boundary that is inevitably too high or too low for many buyers. Instead of upwards of 95% of customers paying nothing, a majority can be driven to pay some workable amount -- and some to pay very well. (But the buyer is always free to pay nothing, at any time he feels that the product is not satisfactory.)
- Do customers fully understand the premium offer? FairPay is built around a structured dialog about offers and value received, and lets customers in good standing try both regular and premium offers whenever they want, and then determine the value they see in it. The seller highlights the value, based on their individual usage patterns, and if a buyer does not value it enough to satisfy the seller, that buyer's trial of a premium service can end (but it can be extended long emough and often enough to be a proper test -- at any stage of the relationship). The basic process is structured, but lightweight, informal, dynamic, and intuitive.
- What is your target conversion rate? Freemium centers on a single all-or-nothing boundary between free and paid that make it costly to guess wrong, and either miss much of the market or leave money on the table (by undercharging good customers). (See my "Long Tail of Prices" post for more on this.) With FairPay, "free" users are permitted to pay 0% of suggested price, while paying users can pay 10%, 80%, 120%, 200% or whatever -- the conversion process is one of nudging customers up the pricing curve, and getting them to try (and pay for) more valuable features. This multivariate dynamic optimization process is more complex, but even simple heurisics can offer far more nuance and flexibility than the hard boundary of freemium. (Adding additional premium tiers to freemium can add a bit more nuance there, but still in a static way that is not easily changed.)
- Are you prepared for the conversion life cycle? As Kumar observes, early adopters are less price-sensitive than others, and are often people for whom the value proposition is unusually compelling. Freemium has no way to adapt to such variations over time, except to move the boundary for everyone. The core process of FairPay is driven by ongoing dynamic adaptation to different price-sensitivities and value perceptions, so identifying and dealing with such individual behavior is baked in to the process.
- Are users becoming evangelists? Free users can have value as evangelists (as Kumar notes) and also as a target for advertising (a key revenue source for many services), and viral marketing can be very important. FairPay can accommodate whatever value factors the seller and buyer choose to consider: Free users can claim credit for evangelism and receiving and acting on ads, and viral offers can draw on the same process to offer FairPay "trials" that suggest pricing with a trial discount that encourages those new users who do see value to start paying something immediately or very soon (while still letting those who try it but do not acknowledge any value pay nothing).
- Are you committed to ongoing innovation? Freemium is very focused on customer acquisition, but FairPay is designed to do its adaptive work throughout the life cycle, as usage and understanding of the product matures and changes over time. Because it is based on an adaptive value discovery engine that always sets prices in accord with current perceived value, it works throughout product and customer life-cycles, and continually drives the seller to make more desirable offers, based on detailed, real-world customer preference data. FairPay offers the potential to not only serve as a pricing engine, but also to serve as an engine for partly automating product innovation, as well. Detailed value perception data can be used to drive offer bundling and product development. So, once you get FairPayworking adaptively to set near optimal individualized prices, why would you drop it for a less adaptive alternative? If things are stable it works near optimally, and if things are changing, it is a nealry ideal tool to identify and adapt to the changes.
