The Freemium Myth – more data

My last post with some thoughts on product pricing has received a ton of traffic, comments and email. Clearly this topic is one that a lot of entrepreneurs care about (and struggle with). A few people pointed me to a great post by Ruben Gamez of Bidsketch on the Software by Rob blog that talks about freemium plans and why, in Ruben’s view, they aren’t always drive the results companies are looking for. It maps well to my thinking (I directly called the freemium model into question in my pricing post). There’s some great data in the post – definitely read the full thing. Here’s a few that caught my eye:

Bidsketch started out with a freemium model. Ruben carefully documents their early success with this (by early, he’s referring to a weeks, not months) and their challenge only a few months after launch of a sub 1% upgrade rate and rapidly increasing support challenges (they had a huge user-base – just not one that was paying). And then he did something “radical” – and completely got rid of the free version. This change led to an 10x increase in paid conversions.

Jason Fried from 37signals had a similar experience. “…the majority of people who are on pay started on pay…” he says. And by correlation, most people who start on free stayed on free.

CrazyEgg doubled their revenue the month they dropped their free plan.

We’ve had similar experiences with companies in our universe that struggled with freemium pricing plans. And while there are clearly companies that have made a success out of offering a free service to a large percentage of their user base and charging the few that are willing to pay (including some very successful ones in our own portfolio) I’m hoping that more companies at least consider that their best pricing plan may not need to include “free”.

  • Interesting thoughts. I wonder what the post free removal acquisition looked like. I could imagine that providing a free option would attract many samplers, some of whom would be willing to pay to stay on if the free option no longer existed. However, if there wasn't a free option (ie after the switch) then there wouldn't be that user acquisition channel (ie they'd just skip trying it, never to convert to paid).

    Or maybe I'm misunderstanding the data.

    Anyhow, I would agree that not enough early stage ventures charge, and they don't generally charge enough.


    • sethlevine

      it’s not clear to me whether in these cases the companies got rid of the free option entirely or if it was just for new sign-ups (i think the latter). relative to the user acquisition channel, i think that’s the whole lesion of the exercise (again – for these companies). by getting rid of their free option they ended up with more paying users. as it turned out, the idea that their free user base would be a good hunting ground for paid users turned out to be false.

  • One can not price for increasing market share and make revenues at the same time from the same product. It is either one or the other. Sun Microsystems learned this lesson the hard way.

  • Andy Blackstone


    Here's a real-life example from the user perspective. I use the free Personal Edition of It works for me because I'm a one-man band, and don't need the features that come with paid-for versions. It works for Salesforce because I have implemented their software for a number of clients and brought at least a hundred new paying users to them. Here's the interesting part – if they didn't have a free version and the single user version was under about $20 per month, I'd still be using it and all of the advantages for both parties would still accrue.

  • @Brendan – Seth was right in that the free plan was no longer available for new accounts. I felt that forcing existing free users to upgrade or lose their data was bad form and didn't want to take that approach — especially when it wouldn't likely amount to many conversions (but maybe some negative word-of-mouth).

    @Seth -Thanks for the kind words. I went back and read the article that you wrote and there's a lot for me to digest there — it's really an excellent article.

    There are a few things you mention that I've been thinking about. For instance "most trial users who will eventually become customers at 30 days will do so at 14 days."

    Interesting point. I have to go back to look at my data but I *think* I saw this behavior when I had my free plan. The way users sign up (with CC info to start) right now doesn't give me an easy way to see when the "conversion" takes place. Though I do notice that most people cancel their trials when they receive my 5-day end-of-trial warning email. Something for me to think about here…

    In any case, it's good to see that you're a tech investor that "gets it." Looking forward to exploring more of what you've written and adding you to my RSS feed 🙂

    • sethlevine

      Ruben – I’m really glad you caught this post (was actually going to track down your email and send you a copy). And thanks for the additional data behind some of the examples you gave. I completely agree that when a company makes this shift it’s a bad idea to cut off your early free users – too much of a PR headache and frankly not the right thing to do anyway. Thanks for taking the time to write in.

    • I found that many people are underestimating the power of momentum. You need to act and close the deal while the person is still very excited about the product. Again this number I found vary wildly. It might be 2 days like 30 days depending on the product.

      You can also play with loss aversion to make them act quickly. Moment + Loss aversion is a great duo.

  • Great post and comments.

    When you're a young and aspiring web service, the most critical thing is to get users (any users) and feedback. Offering the free option opens the gate for a broader swath of folks that will kick the tires, take it for a test drive, and hopefully let you know what you're doing right and wrong. Once your service is established, the free option can be tightened, or converted to a 'free trial'. Original adopters get to keep the free option. New users only get the free trial.

    • sethlevine

      I certainly agree that all web services (all companies for that matter) benefit from getting feedback from early users. But there are different ways to do this (closed beta, open beta, user groups, etc.) that may be better than offering a free version of your service and then tightening it down later.

  • Good post. These are definitely the questions and discussions that we are having right now. Keep 'em coming!

  • You're very welcome. Glad to contribute in any way I can.

  • I do have a thought on this method of gathering feedback. I gathered some great feedback from my closed beta which ran for a few weeks. I built up a mailing list of several hundred users before I went live through a temporary landing page and only let in a few of those users to do a closed beta with a commitment that they'd provide feedback by a specified deadline.

    When I went live with my free plan I started to get a ton of feedback from free users and at the time I didn't know that this feedback was any different from that of paid users. Eventually I found that free users were asking for it to do more (invoicing, CRM, task management, etc.) while my paid users wanted features focused on its core competency and integration options for their business.

    So this is a really long way of saying that one should be careful of the feedback received from free users as it may not necessarily be in the your best interest to implement them.

    • Kaiser Soze

      Wow, awesome thought Ruben. I'd never thought of this difference, thanks for sharing. In hindsight, my experience is somewhat similar too.

  • Good Post. I mentioned it in my reply to Ruben's article.
    I provide some insights on why and how to use Freemium and when not to.
    I agree Freemium is not for everybody but it's now flawed at it core. It's just very easily misused.

  • sethlevine

    Thoughtful response, Diego. While I don’t agree with all of your logic, I do believe that the Freemium model can work for some compaines (as I said in my post – we’ve invested in several companies that have used it successfully). It did make me think about diving into it further – I’ll try to put something up in the next day or two on it (this is such a meaty subject!).

    • Kaiser Soze

      Seth – Is it fair to summarize your thought as follows?
      * B2B apps targeted at "enterprise-type" customers should not use Freemium.

      • sethlevine

        for the most part that’s how i feel Kaiser. there’s the occasional b2b enterprise app that benefits from the scale of its user base but they are few and far between. more likely with an enterprise b2b app you’re better off with a full pay model in my opinion.

  • Kaiser Soze

    Thanks Seth, appreciate your thoughts. At our company, we've been discussing the pros/cons of Freemium model lately. We've never tried Freemium – we just have 3 paid editions, we've been able to build a profitable company with our current pricing model.

    Here's an interesting discussion I saw recently on the subject of freemium for a B2B app (Mktg Automation), with participation from several startup CEOs in that space:

  • Divya Shah

    Seth, What do you think of and Dropbox that are B2B for most parts and seem to be doing well with a freemium model?