Freemium business model case study: AdultFriendFinder ARPU, churn, and conversion rates

A case study for the Freemium business model
There’s been a lot written about the Freemium model over the years, particularly from Fred Wilson at AVC. Here’a couple articles I’ll recommend on the topic:

Anyway, given the great interest in direct monetization due to the economy, I decided to write a post on the topic as well, focusing on AdultFriendFinder, which has recently released a bunch of great data.

AdultFriendFinder files for IPO
Recently, the holding company FriendFinder Networks, Inc., the owner of AdultFriendFinder (and Penthouse), filed to go public. As such, they released a Form S-1 where they go through many of their overall business metrics – you can read the full SEC documents here.

OK, it’s a lot more boring than you might expect from an adult-oriented conglomerate. It’ll put you to sleep.

There’s a ton of data in the S-1, but I’ll just summarize some of the most interesting stuff related to acquisition, monetization, and retention of the customerbase…

FriendFinder’s business metrics
First off, there’s some good definitions for how they think about their business – everyone who is thinking about direct monetization of their customers should be familiar with all the terms below:

  • Visitors. Visitors are users who visit our websites but do not necessarily register. Visitors come to our websites through a number of channels, including by being directed from affiliate websites, keyword searches through standard search engines and by word of mouth.
  • Members. Members are users who complete a free registration form on one of our websites by giving basic identification information and submitting their e-mail address. Members are able to complete their personal profile and access our searchable database of members but do not have the same full access rights as subscribers.
  • Subscribers. Subscribers are members who purchase daily, three-day, weekly, monthly, quarterly, annual or lifetime subscriptions for one or more of our websites. Subscribers have full access to our websites and may access special features including premium content.
  • Paid Users. Paid users are members who purchase products or services on a paid-by-usage basis.
  • Average Monthly Net Revenue per Subscriber. Average revenue per subscriber, or ARPU, is calculated by dividing net revenue for the period by the average number of subscribers in the period.
  • Churn. Churn is calculated by dividing terminations of subscriptions during the period by the total number of subscribers at the beginning of that period.

And in fact, you can think of Visitors, Members, and Subscribers/Paid Users as a funnel that they manage to extract revenue. Then you combine that with the ARPU and Churn to get an understanding of monetization and retention, respectively.

Funnel metrics
And in fact, they list some of the key conversion rates between each of these numbers:

  • Visitors: 59 million uniques worldwide
  • Members: 4 million new member registrations (on 270 million members total)
  • Subscribers*: 900k paying subscribers
  • ARPU: $19.06 per paying subscriber/paid user
  • Revenue per member**: $0.95 per member
  • Churn: 18% month over month

*New subs/month number is not in the S-1, but is derived from the fact they have been about level on total # of subscribers, yet with 20% churn, so to make that up they must be adding a little over 200k to stay even.

**Revenue divided by new members (rather than subscribers)

So let’s try to figure out what their funnel %s look like. One complication is that the FriendFinder S-1 gives total members and new members within a month, but doesn’t give “active members.” At the minimum, it should be 6% since 4 million new members on 59M unique visitors yields 6%, but you might imagine it would be closer to 15%, which is closer to an industry standard.

Here’s a couple ranges, based on the above assumption:

  • Visitors -> Members: 6-15%
  • Members -> Subs: 10-22%
  • Subs -> Renewing Sub: ~80%
  • Revenue per member: $0.48-$0.95

You can compare this to Free-to-play games and Casual MMOGs via these blog posts from Jeremy Liew and Nabeel Hyatt, although you should note that their definition of ARPU is calculated from actives rather than subscribers, so you should use my above “Revenue per member” number of $0.95 rather than the substantially higher $19.06 per paying subscriber. Anyway, you’ll see that it’s actually not much different!

It shouldn’t be surprising that all of these freemium models, after careful optimization, come out at roughly the same numbers.

Acquisition
I also want to highlight a couple interesting points about how FriendFinder acquires their customers:

Marketing Affiliates. Our marketing affiliates are companies that operate websites that market our services on their websites. These affiliates direct visitor traffic to our websites by using our technology to place banners or links on their websites to one or more of our websites. As of September 30, 2008, we had over 110,000 participants in our marketing affiliate program from which we derive a substantial portion of our new members and approximately 44% of our revenue. For the nine months ended September 30, 2008, we made payments to marketing affiliates of over $46.4 million.

[also, in a separate section…]

In addition, for over 10,000 of our affiliates, we maintain private label websites that provide a seamless, turnkey outsourced solution using our technology platform for social networking and live interactive video websites. These websites have the look and feel of the affiliate’s website with the affiliate’s logo and website name but are operated by us. Users who click through the affiliate’s website are tagged with the affiliate’s identifier that tracks the user to calculate the payment due to the affiliate. Private labeling allows our affiliates to preserve their brand while generating revenue for us.

There’s also a block of text on the extent to which they spend money on Search Engine Marketing:

We rely on both algorithmic and purchased search results, as well as advertising on other internet websites, to direct a substantial share of visitors to our websites and to direct traffic to the advertiser customers we serve. If these internet search websites modify or terminate their relationship with us or we are outbid by our competitors for purchased listings, meaning that our competitors pay a higher price to be listed above us in a list of search results, traffic to our websites could decline.

How much do they spend on SEM? Turns out it’s over $50MM per year:

The largest single selling and marketing expense item for Various were “ad buy” expenses which amounted to $54.8 million and $55.3 million for the 2007 period and the year ended December 31, 2006, respectively, the cost of purchasing key word searches from major search engines, together with expenses related to associated personnel.

This means that between their affiliate payments and SEM, they are spending around $100MM buying traffic!

Conclusion
There’s other neat info in the S-1 about their chargeback rate, the split in revenues between their magazine side (Penthouse) versus AdultFriendFinder, and also the breakdown of all the smaller demographics they service. For example, did you know that they also run a bunch BigChurch.com for Christian singles?

Anyway, there’s lots of fun information in there – if you find something fun in there, let me know and I’ll update this blog post to reflect it.

As always, comments and questions are welcome!

Published by

Andrew Chen

Andrew Chen is a general partner at Andreessen Horowitz, investing in startups within consumer and bottoms up SaaS. Previously, he led Rider Growth at Uber, focusing on acquisition, new user experience, churn, and notifications/email. For the past decade, he’s written about metrics, monetization, and growth. He is an advisor/investor for tech startups including AngelList, Barkbox, Boba Guys, Dropbox, Front, Gusto, Product Hunt, Tinder, Workato and others. He holds a B.S. in Applied Mathematics from the University of Washington

Exit mobile version