Creating the Best Freemium Model

Many subscription services attempt to build their business by offering a Freemium Model, which means offering a free service to basic users while hoping a good share of users will upgrade to a premium plan and start paying for the service. The keys to success for this model are to maximize the conversion rate of customers who will pay, while keeping the free version attractive enough for customers to sample the service.

Companies experience a wide range of conversion rates, from Spotify’s amazing 27% to a more typical range of 1% to 4%. The conversion rate is a big driver of profitability, and finding a way to move from 1% to 2% is doubling the success.

The key is to design the model into the product rather than making it an afterthought. Think hard about how you will give your free users a great experience while educating and tantalizing them about what more they could get by paying the premium.

Some of the less-effective models I have experienced on either the customer or the company side, thus the wrong approach, have included:

  • Send free users emails and offer webinars to describe how much more they would get with the premium version of the product. This does not require any specific product design, so is easy to implement. In my experience this approach is ineffective, as it appears to be more like a stand-alone sale rather than an add-on to a sale you have already made. It does not show up in the context of daily use of the free product, so it is difficult to get the user’s attention.

  • Load up the free version with paid advertisements, promising to block those if the user opts for premium. Sometimes this involves outside advertisers, and in others the service is adding their own marketing messages to some customer interactions. Excessive promotion can be annoying, and can detract from the free user experience. Some users are price sensitive and willing to tolerate the advertising, so will not convert anyway.

  • Provide the premium model for free for a limited time period, then turn it off if the customer does not buy it. This is an easy product to build, and is more of a billing system design. One could hope that users will get hooked on premium features and then want to pay to continue them. Many times I have found that users barely have time to explore the product within the trial period, so never get to see what they are missing. Very often your best target users are busy people with many other distractions. And without some plan for long-term use the free users are unlikely to load their own content into the service, to get value from it and make a commitment. Some companies will extend the premium trial period, but that may not be enough to get the user hooked.

  • Build pop-up messages into the free version to highlight features that could be available if the user pays for premium. Sometimes these features appear as greyed-out options in the free version. This approach requires some product design to create the pop-ups or to limit access to some features for some users. The challenge with this approach is that it is difficult to effectively deliver a feature value proposition in a simple pop-up box. In addition, no single feature might be compelling enough to justify the upgrade. In one recent experience, a single feature I thought would be nice to have required an upgrade from free to $50/month.

  • Restrict collaboration platform to one user in the free model. I have actually seen this done. It is simple to administer, but it fails to demonstrate the benefits that come from having multiple people sharing. It would be far better to restrict the team size to three or four people to demonstrate the power of collaboration.

What many of the successful freemium companies have discovered is the best approach is to give free users full functionality, but to hobble it in some way to limit the value. This enables the users to experience the additional functionality in the context of their regular use, and then justify for themselves the value of upgrading.

Some examples of good application of this “hobbling” approach include:

  • Slack, with a purported 10% conversion rate, provides teams full functionality for the free product, but limits the archive of messages to 10,000 messages. Larger teams hit this limit faster. To the extent the team makes Slack their core communications, many do not need any additional selling to see the value of having access to all their communications history. By that time they are hooked.

  • Evernote, with a 4% conversion rate, offers several benefits for moving to their premium model, including more storage space. They offer just about all their functionality in the free plan. Another feature they add with premium is the ability to sync data across more than one device. So users can get most of the value in Evernote in one device (desktop or mobile), then want to extend the value to another device, so pay the premium.

  • Spotify, with its awesome conversion rate, allows free users to create playlists, but then injects their own related playlist into the mix. When a free user tries to skip ahead more than 8 times, skip an advertisement, or load their own choice into “up next”, Spotify suggests they upgrade to premium. Once again the free users get to experience the convenience and the wide range of music, but then can choose to upgrade to get more control over the music.  

  • When reporting tools are part of the premium package, I have seen companies successfully offer several free reports per month in the free version, but with the next request tell the user he or she must upgrade to get the report again. Several others have had success in making the full range of reports available, but running them against only a subset of the data (clearly marked with a warning label), perhaps a portion of the customers or for, say, only Tuesday usage. In both cases the sampling is free and the user gets hooked. In the first instance the user sees value with full reporting on the data, and in the second the user can experiment freely with multiple reports, but not rely on the results.

In each of these cases the users get a free taste of what they could get more of with the premium model, while they are experiencing the free version. At the same time, each of these companies required some foresight in the product planning phase to build the sampling functionality. Most times I have this conversation with a freemium-model founder who is frustrated with the low conversion rate, I hear “that would be nice to do but it is too late for us”.

All these models require constant trial, measurement, and refinement once out in the market. Build in the A/B testing tools early and try out different variations until you find the combination that works best. Once again, that benefits from pre-planning in the application design stage.

Plan your business model ahead of time and build your product to support it. Do not wait until it is “too late” for you.