Porous paywalls

Don’t dismiss porous paywalls just yet. This is a reaction to today’s Pando Daily article on how porous paywalls don’t work, which is a reflection on their experience with freemium.

A “porous paywall” is just another term for the freemium model that online publishers are experimenting with, where the paygate is not based on the type of content a user wants to read but on the amount of content they can read for free before having to pay.

Freemium is about adoption. The only reason a business gives out something for free is to widen their funnel and increase adoption that will fuel a chain reaction to eventually increase customers and revenue. Washington Post, New York Times, and other newspapers cited in that article can afford to have a strict paywall because they already have enough adoption. For now.

Freemium is easy to screw up. The good part about that PandoDaily article is where it links to this essay: You’re Doing Freemium Wrong. A restaurant cannot give out free food in the hopes that it will convince its regulars to pay (and subsidize the rest). It can, however, let people walk into its doors to use the restroom, or get a glass of water, or browse the menu for free – in the hopes that some of those people will grab a table and buy food. If you set expectations that something is free, forget about being able to charge for it ever again. If you can’t charge for what you are selling, sell something else. Freemium is not about power users. It is about widening the funnel with goods or services that (a) have broader adoption than your core product, (b) are not cost prohibitive to give away for free at scale, and (c) making sure there is a natural progression from the top of the funnel to the bottom of the funnel (where revenue exists).

Pando Daily’s experiment was incomplete. Their article mentions their freemium experiment where “the vast majority of articles” was given away for free. The rate of new subscriptions plummeted. So what this experiment has effectively proved is making the majority of your primary offering free does not lead to greater revenue. It is one data point along the graph of all possible tuning for a freemium experiment. NSFWCorp just hasn’t explored enough to find the optimal operating point.

More about incorrect conclusions: If you read that article to the end, you will read about NSFWCorp selling lifetime subscriptions priced at a one time amount of $3 and increasing all the way to about $1500. They generated a one time revenue of $350,000. It is a membership to a club of sorts (your caricature on their site, some event invitations, subscription, etc) so it is hard to comment on the value. Assuming they hope to be around for 10 years, that is $3/month x 12 months x 10 years = $360 / subscriber in 2013 dollars. Giving that away for low prices is foolish, but there were only a handful of cheap “rooms” and the publicity may be worth it. Btw, most of the memberships sold so far are actually less than $300. But here are the INCORRECT takeaways from this: (a) users are buying “rooms” in their “conflict tower” (as they’re calling it) to support independent journalism, and (b) a lifetime subscription at some fraction (1/20th, 1/10th, 1/5th, whatever) of the total cost is better than a freemium model. There is no way to prove these with this experiment.

What it does prove, however, is that they have been able to create a “bundle” that is valuable to their existing users at those prices. That is, NSFWCorp is able to upsell to their users. Sort of supports the freemium theory.