Pay wall games

Jonathan Stray has posted a tool at the Nieman Labs site to model the financial impact on newspapers that put up pay walls. He uses NY Times data to critique their recently announced decision to start charging Web users for access.

We, too, are in the thundering herd of US newspapers moving toward charging for access. (More on our plans in a later blog.) So I played with the tool a bit to see what it might foretell for us and other papers.

My take:

The tool is useful in that it encourages you to think about different segments of your audience and how they might respond. And it drives home the fact that most newspaper sites almost certainly will see big PV declines when they shift from free to paid. But the tool is too crude to serve as much more than an online parlor game. At least two key factors are missing: The unsold-inventory cushion and the indirect effect on print sales.

The issue of the unsold inventory cushion was raised in several comments to the NL post. It just doesn’t make sense to use an average CPM/page number to calculate the revenue impact of switching to paid. Most newspaper sites have a pretty high percentage of unsold spots. Forty percent would not be unusual, even for a popular leaderboard or rectangle spot. If those are the only page views lost to switching, you could conceivably keep all of your revenue from sold inventory.

Remnant inventory is worth almost nothing. Even with three or four ad spots per page, most of us are lucky to get $2.50 CPM. (The NYT may do better, of course…)

So take a modest example of a small paper with 2 million PVs a month, just 50% of its inventory sold and a remnant rate of $2.50. If they lose half their traffic, they’re down $2,500 or $30,000 a year. That suggests somewhere between 250 and 500 monthly subscriptions (at $5 to $10 per month) would be required to replace the lost revenue. According to API, mature paid sites typically claim online subscribers equal to between 1 and 3 percent of their print circulation. A 20K paper would need to hit the high end of that to break even. Not easy, but not crazy, especially if leaders like the NYT begin to change the expectation of readers.

More importantly, the model doesn’t consider the indirect cost of lost print readership.

This is, of course, a contested issue, and I’ll attack that in a later post (or several). But for now suffice it to say that I have (reluctantly) come to the conclusion that we have crossed a tipping point. With better bandwidth, better computers and browser technology, better sites and the shift of so many of our daily activities online, more people are weighing the free online version against the paid print paper and choosing the former.

There’s some anecdotal evidence to suggest that shifting to paid sites can help slow the decline of print papers. Assume, for the sake of argument, that you stop just 1 percent of the annual decline. For that 20K newspaper above, that’s 200 readers, each of whom might be worth $200 or more a year in circulation revenue. Call it $175 to allow for aggressive discounting of marginal readers. That’s $35,000. So if that 20K paper can hold just 1 percent more print readership with a paid site than with a free site, they’re ahead of the game even without many paid subscribers.

Bottom line for me:

Going paid is risky. It’ll prove to be wise only to the extent that we can slow print sales declines and build online subscriptions over time. That means even better online products and smart/flexible pricing strategies.

2 thoughts on “Pay wall games

  1. This is a very intelligent response to the story. Thanks.

    The “unsold inventory cushion” is indeed an interesting point. I would be curious to know what the size of this cushion is on average.

    As for not modeling the effect of protecting print revenue, also a good point. I don’t think that trying to “protect” a dying medium is a good long-term business plan, but I do understand the need to get the most out of it while it lasts.

    If you l like, I can check with my editor to see if I can open-source the thing and you or your flash programmer friend can add any sliders you like.

    – Jonathan

    • It’d be fun to be able to play with it a bit. Inspired by your work, I started to refine my own spreadsheet model, before I got sidetracked by a couple of minor emergencies like the wind that was ripping the siding of our building…
      I have no idea what the average unsold inventory is. I suspect there’s tremendous variation, especially at smaller papers, where online sales expertise is scarce. There also can be a lot of variation even within a site, to the extent that the site allows targeting. At our site, for example, the obits page is generally sold out in at least one ad size, and we also get a lot of demand (oddly) for home page ads. We’ll certainly be keeping those factors in mind as we tiptoe into paid access.

Leave a Reply

Your email address will not be published. Required fields are marked *