Freemiums for SaaS
Posted by Bob Warfield on March 25, 2010
Time for my first Learnings post about my Helpstream Experience.
Freemiums are a fascinating subject. Whole books have been written about the concept of extracting value from something that’s free. The FOS (Free Open Source) movement is all about free, and we benefited tremendously from it at Helpstream. Not one single software license fee was paid in the making of that software.
Helpstream has been my first experience with a Freemium, and it was a contentious one. The original CEO at Helpstream put the Freemium in place for a variety of reasons. First, having worked in pricing for IBM, he wanted to price everywhere along the demand curve so as not to leave a flank exposed at some price point. Interesting idea and strategy. Second, he wanted to use it as a forcing function to make sure the software was built in such a way that the cost to deliver the service was extremely low. Again, interesting. I’ve written in the past about cost of service being critical to SaaS success. We ultimately achieved a cost of 5 cents per registered user per year of our service, which is extremely low, so I would say he accomplished his goal on that one too. Third, I was given to understand that there was a point earlier than my arrival where some thought was given to the possibility of having to make a business off ad revenue.
The contention came from Sales, and to a lesser extent Marketing. Sales was not particularly interested in any of the goals the CEO had set forth. One of the things one learns about Sales (and I came to Helpstream knowing this full well) is that they are carefully selected and trained to be entirely coin operated. If it doesn’t contribute to their commission they are not interested. If they decide it is taking away from their commission, they are actively hostile. In this case, Sales felt that the Freemium was competition: it was taking sales commission out of their pockets. Hence it became almost impossible to talk about the Freemium without polarizing armed camps, which eventually made it a sacred cow. This is particularly difficult when the same executive is responsible for Sales and Marketing (more on that, perhaps, in another Learning Post). The CEO and I could chat about it, but its darned hard without the whole team on board and trying hard. BTW, this is not that uncommon in Executive Staffs, and it is one of the roles of the CEO to manage the discord lest it become dysfunctional.
Our Freemium had some other difficulties that would surface whenever we could get past Sales’ desire to just shut it down immediately. The biggest problem was that it was not attractive to the right kind of customers. Attracting the right kind of customer and repelling unsuitable prospects is a subject about which I will have more to say in a later post because it is critically important to startups. Even the negative aspect of repelling unsuitable customers is important and not something that even occurs to most people to try to do. In this case, our Freemium consisted of full-featured product with an exception–branding was disabled. You could apply a logo, but you did not have fine grained control to present a portal that reflected your brand. This turns out to be important for a Customer Service product with the types of companies we were selling to. They needed that branding capability. A couple of other factors played in as well. Customer Service is by definition Customer-Facing. Very few companies are progressive enough to conduct experiments in Customer Service, and those that are will not do so without a lot of thought and deliberation. This is not a departmental land-and-expand application. This is your fundamental Customer Service Experience. So, prospects would get a free workspace and play with it a bit, but it was sterile and did not give them an ability to see it in the context of their intended use.
Lastly, that point about being sterile, goes to the issue of content. Lots of Enterprise Software and Social Software requires content before it becomes clear how it is used. We had the ability to instantly provision a workspace, but we were not stocking those workspaces with interesting content. Essentially, prospects were getting into a new house that had not yet had finish carpentry done. They could see it had a solid foundation (“good bones” as the house people say), but they couldn’t tell if it was going to be a Mediterranean-style villa or a Western Ranch style house, let alone whether they liked the colors and carpet choices. I had a plan to fix this in the same way builders do. I wanted to create a “Model Homes” section on our web site where people could prowl model homes that were set up with content and so forth based on their industry. We would also put selected customer workspaces in as Model Homes given their willingness to present their portals to non-customers. Alas, we never got the time or budget to make that a reality, but I think it would have helped us out a lot. I haven’t really seen the Model Home concept done anywhere, but I suspect someone does it. ServiceNow has a cool fully functional version of their product available for anyone to play with, and this is pretty close, albeit not specifically targeted at industry verticals. We also did create a Best Practices Community called the HelpExchange that was a much better trial experience than a bare workspace.
The long and the short of it is that we got a ton of Freemium users, but they were mostly not customer prospects. We were not getting as many Customer Service users as we were Internal Collaboration users. Helpstream was great for that too, but it wasn’t our market. In addition, the Freemium was self-selecting customers who were most likely to be willing to pay, well, nothing! This is a big problem for the Open Source world, and makes me wonder how viable Open Source even is as a business model, much as I appreciate all that they have done. We had lots of schools, churches, and other non-profits. We had government organizations, including the US Army, where getting actual budget and project approval is extremely difficult, so the users would simply sign up to a free solution and get on with their work.
Would I recommend a Freemium to other SaaS companies? Absolutely!
Or at the very least a comprehensive trial opportunity. We got a lot of value from our Freemium. It helped us shake down our system. The free users gave us a tremendous amount of feedback and testing. It enabled us to show some momentum in terms of numbers of customers. Interestingly, while we wound up with 200 businesses using the free version and 40 paying customers, the number of seats involved was never large for the Freemium. Very small businesses used us for Customer Service and larger businesses used us as departmental collaboration hubs. Lastly, it made our software better. The things we had to do in terms of making sure it was cheap to deliver the service so the Freemium wouldn’t be too costly, or making sure provisioning could be automated (along with other self-service aspects) were all good things for a SaaS architecture.
What would I change about the Freemium to make it more successful?
I’ve already mentioned the need for Model Homes and more content. An automated guided tour would also be very helpful with a complex SaaS product. In the end of the day, I prefer a trial to a Freemium though for a company like Helpstream. When I took over as CEO of Helpstream, the first thing I did was to limit the Freemium. I looked at the usage statistics to determine who was deriving real value. A hospital using the Freemium to deliver an IT Help Desk to hundreds of people was clearly deriving real value. Based on these statistics, I set limits on the Freemium. A free workspace could have a total of 50 users, including both administrators and end users. Any workspace that exceeded 50 users had 30 days of trial before that workspace would be shut down. Out of 200 free customers, many of them unsuited to becoming paying customers, we actually got about 5 to convert to a paid plan with this change. 2.5% is actually not too bad as a conversion rate, all things considered. In essence, we had a free product for very small organizations and a 30 day trial for larger organizations.
The last thing I would look at doing is lowering the adoption friction in a variety of ways. Letting the free customers access the branding and perhaps the single sign on (the two most common professional services engagements for installation) and making that super easy as self-service would be very helpful. Reducing transaction friction would be helpful too. Here’s a standard order form, enter you credit card here, and you are done. Above a certain deal size, nobody will do business that way, but in the original spirit of pricing along the whole demand curve (a good idea, Tony!), it’s worthwhile to stake out that territory and make it so easy that sales does not even need to be involved.