When people talk about freemium businesses and business models, most of the time they’re talking about it from a revenue or product perspective. Revenue, because the essence of the model is to give away something for free. And product, because this is the key to freemium success. Only a compelling, unique and engaging experience will keep people coming back regardless of whether they pay you or not.

One often overlooked side of freemium is the cost side. Freemium works if the incremental cost of providing your free product or service is essentially zero. Most of us get this. But if you don’t have limits on the cost side of your free offering, you could be in big trouble.

Let’s look at VoIP for example. Skype charges nothing for in-network callng (calling another skype client). Not coincidentally, its costs are also more or less nothing. Especially given its P2P architecture. If you want to call a phone however, you pay since Skype has real termination costs to pay carriers. Not all VoIP players have so clearly separated their free vs. paid based on costs. I know this from experience, having been CFO at Mobivox, where we were very generous out of the gate with our free calling. The problem was that every new user cost us real money every time they made a call. Mobivox was not alone facing this. We were in an intensely competitive space and we all gave away free minutes.

There are many more examples of freemium businesses that have gotten it right and wrong when it comes to assessing the cost side of their business model. Take the battle of the social networks as the most visible example – Facebook vs. Linkedin. Both networks are successful by any measure, but they have very different cost and capital efficiency profiles.

Facebook, which just passed 200M members has raised $ 516M in total capital (or just over $ 2.50 per user). This amount includes $100M of debt, and they’re off looking for more right now to fund growth.

Linkedin, with over 30M members has raised $ 103M or $ 3.43 per user. However, according to them only $ 28M (93 cents per user) of that (everything before last September’s Bain Capital round) was used to fund growth. The new money is in the bank for acquisitions.

Why the difference in capital required per user? I’m sure there are many reasons, but incremental costs have to a big one. Look at your own Facebook profile for example. How many photos or videos do you have there? Lots, right? And what is the main thing you do on Facebook? If you’re like most people you post photos and videos and look at friend’s photos and videos. There are seemingly no limits to the amount of content you can upload.

Now look at your Linkedin profile: No matter how many connections you have, you’re probably using the same bandwidth. You can have one profile picture. They now have a slideshare app, but you’re limited to four presentations.

The point is: Linkedin has placed very clear limits on the costs to serve an individual user. Facebook has not. As such they have very different cost structures.

We also see this phenomenon with Youtube which is apparently losing $ 1.65M per day .

In your business, if you have real, tangible costs of serving your free users, then you need to better segment your offerings to get these people paying you or leaving the service. If it costs you every time someone uses your product or service, make sure you’re getting paid directly (through subscriptions) or indirectly (through ads).

Category:
Business Model, freemium, Growing Big, Management, Strategy, Web 2.0
  • Mark MacLeod

    Thanks Martin. Read your post. If there is value in the size of your network (like Facebook) then you are way better off not charging and instead building a big audience. If, however, there is no network or audience size value and you just offer a stand alone app, then free should probably be limited to a free trial, since there is strategic value for your free user.

  • Martin Thomas

    Hi.. Your post got me thinking… What is more valuable for a software company (like facebook or flickr). 1,000 paying users or 100,000 non-paying users? What are your thoughts? View my blog post here: ” target=”_blank”>http://www.purlem.com/blog/?p=57

  • Martin Thomas

    Hi.. Your post got me thinking… What is more valuable for a software company (like facebook or flickr). 1,000 paying users or 100,000 non-paying users? What are your thoughts? View my blog post here: ” target=”_blank”>http://www.purlem.com/blog/?p=57

  • Mark MacLeod

    Thanks Ryan!

  • Mark MacLeod

    I don't know if micropayments on that scale are factored into FB's valuation yet. Also, I don't know that FB has the trust and neutrality to be a true payment alternative. Beacon and other missteps would make me very wary of using FB for payments

  • Mark MacLeod

    When you think of the market cap of the biggest carriers, you can easily imagine a scenario where even if Skype's direct revenues were decreasing because everyone was on the network. its value as a takeout target would increase. The carriers would take Skype out if it got that big and became a real threat to their revenues

  • Ryan Yockey

    Mark, I never looked at a break down per user. What you bring up here makes perfect sense and I totally agree. Thanks for the article.

  • AlexFoley

    Now that is an interesting paradox! Each new customer actually creates less potential revenue for the company… but does open them up to quite a relationship with that customer…

  • AlexFoley

    The comparison between FB and LI does discount the fact that FB seems to be angling towards being the micropayment provider to the web. With Facebook Connect leveraging the facebook account system outside of Facebook, the extremely lucrative per-transaction fees that credit card merchants currently earn could be Facebook's bread and butter, with advertising sitting on top. That being said, they definitely need to improve how ads are displayed in context with the site if they ever want to become the premier, targeted-demographic ad network.

  • Mark MacLeod

    Thanks Chris. Would love to see that talk after its done. You have an interesting and very specific business BTW

  • Chris Hopf

    I have been watching the "free frenzy" on the web for sometime now . . . I will also be providing a talk at Seattle Tech Startups in mid August, " Understanding Freemium: How to Create and Grow Paying Customers " . There is definitely a right way and wrong way to take the Freemium approach . . . as well as, the initial understanding of whether or not Freemium is a viable approach for your product, service or solution in the first place. Thanks for the post Mark, appreciate your insight.

  • Shai Berger

    In Skype's case, they've hit against another interesting boundary. As they become more pervasive, the odds that a Skype user needs to use Skype-Out actually decreases. To take it to an (unlikely) extreme — In a world where everyone is on Skype, their revenue model disappears completely! I wrote about the impact of telephony costs on freemium models a couple months ago: ” target=”_blank”>http://www.shaiberger.com/?p=170

  • Shai Berger

    In Skype's case, they've hit against another interesting boundary. As they become more pervasive, the odds that a Skype user needs to use Skype-Out actually decreases. To take it to an (unlikely) extreme — In a world where everyone is on Skype, their revenue model disappears completely! I wrote about the impact of telephony costs on freemium models a couple months ago: ” target=”_blank”>http://www.shaiberger.com/?p=170

  • Mark MacLeod

    I remember that post. Thanks for sharing the link again.

  • Wallen's

    I can only agree with you Mark. Subsidizing costly customers can be ok for a short time but not as an on-going concern. Reminds me of this blog post I wrote few months back: ” target=”_blank”>http://wallen.typepad.com/wallen/2009/01/the-cost…

  • Wallen's

    I can only agree with you Mark. Subsidizing costly customers can be ok for a short time but not as an on-going concern. Reminds me of this blog post I wrote few months back: ” target=”_blank”>http://wallen.typepad.com/wallen/2009/01/the-cost…

  • Mark MacLeod

    For sure, there are even more differences between FB and LI when you compare revenue models. LI has one of the highest CPM rates around on their ads because of the demographic profile of their members. FB on the other hand gets general broad-based CPM rates with low conversion because the ads they serve are not particularly relevant to us and because we don;t do much with them. Look forward to hearing more about your project…

  • Leonid S. Knyshov

    Let's also not forget the premium part of LinkedIn. They derive a non-trivial amount of revenue from their premium users and do so at zero incremental cost as well. I am also taking the bandwidth requirements to heart when developing my upcoming product. Unnecessary fluff, such as images, greatly slows down a site and increases my costs.