Sunday, November 3, 2019
I’ve written quite a bit about licensing software now, you can start here to follow the whole thread. In The Platform License Problem, I mentioned some free pricing as a hide-the-sausage technique. When there are multiple markets to find product fit for, and the vendor has a software base that tackles those markets, the platform problem applies. But burying the cost of a shared platform isn’t the only reason to give away software, so let’s look at some more ways that can happen.
Freemium has grown very popular in the shadow IT, consumer tech, and open source based tech markets. With a freemium model, consumers can get your product for free without support, but have to pay for “extras”: additional features, related services, and/or support contract. The pwSafe password manager is free, but cloud-based synchronization costs. The Bear text editor is free, but advanced features cost. Splunk is free to use up to 500 megabytes a day, but costs quite a bit for more. One might say these models are equivalent to a free trial on a service, such as getting Apple Music or Spotify for a free trial. However, freemium is different in that there is no time limit; instead of getting the same features when you start paying or losing the service when you don’t pay, with freemium you can use the free version forever and get more features when you pay. It is more like the shareware model, without nag screens.
Shareware, for those who don’t remember it, was a try-before-you-buy model popular in the early Internet. The customer can download the package and use it with full functionality for a limited time, but must buy a key to continue. After timing out, the software might stop working, or continue working with nag screens or watermarks on its output. Freemium packages hosted on app stores have largely wiped this model out because they allow the small developer to outsource the tedious and complex work of handling customer payments.
Even without the app store’s help, freemium models can be great for vendors. They encourage word of mouth networking, as well as advertising from taste-makers and aggregators. Reviewers and users are happier to recommend products that new users can get for free. New users are happier to try products for free. The vendor gets low fidelity signal from the downloads and traffic of a broad pool of potential customers, and can experiment with the potential for features to convert those customers. The vendor can also instrument the free product for higher fidelity feedback. The only downside for a vendor is in mis-calibrating the free/paid version; too much value for free means failure to capture revenue. As long as costs are covered though, those are theoretical losses that can be ignored in the interest of living a peaceful life. There’s always the possibility to add a better paid feature which changes the picture.
Lite editions of premium products are a slightly different take on the same theme. Typically introduced after a successful product has gained traction, the lite edition is used to expand that product’s influence without requiring a full license for every user. AutoCAD, Visio, and Adobe Acrobat all played this game very well. This model is growing less popular as vendors embrace cloud and app-store delivery of software, but the fashions of software are fickle.
The ultimate extension of the lite version is open source: the software is free to use without time limits, and the vendor must build a business around it using services, support contracts, or enterprise features. It can be harder for an open source vendor to calibrate that free/paid value balance as well, particularly when SaaS offerings erode the value of service and support.
These models all have something in common: they allow the vendor to put basically the same software into multiple customer markets. The target market of customers willing to pay for the product is satisfied, while a broader market of potential customers is also satisfied at very low risk. The vendor can easily striate further by offering a “super pro” version for even more money, limiting complex features to the customers who need them.
It’s a great option that costs even less when starting off from a SaaS base. That said, alternate consumption models can also be overly complicated when you’re still trying to find product market fit. That struggle is particularly challenging for the open source vendors who start there on day one, with a default business model of linear-growth services that faces significant erosion risk.