Blogonomics: Revenue per page

March 27, 2010
debate we had earlier today, Henry Blodget took to Twitter to explain the numbers behind ad-supported blogs. The most interesting tweets, to me, were these:

" data-share-img="" data-share="twitter,facebook,linkedin,reddit,google,mail" data-share-count="false">

In the wake of the debate we had earlier today, Henry Blodget took to Twitter to explain the numbers behind ad-supported blogs. The most interesting tweets, to me, were these:

Ad revenue for a general news site tend to range from $3-$6 per thousand pages. Ad revs for a business or premium site can run $10-$20… [does a bunch of math] And that’s at $10 per 1000 pages, which is actually a good monetization rate (business sites are higher, thankfully). If you work for a gossip or general news site, the revenue per thousand pages can be far lower, requiring vastly more pages per journo.

If $10 is “a good monetization rate”, then let’s be generous and say that Blodget is making $15 at TBI. And then let’s look at a typical TBI page — say, this one. You’ll probably see different ads than I do, but I see a MetLife banner across the top, a FedEx box in the right-hand column, a big Amex/OPEN ad underneath the post, and a text ad just before the comments start.

Now check out TBI’s rack rates. The MetLife banner is $20, the FedEx box is $25, the Amex/OPEN ad is $30, and the text ad is $3. Or something like that, anyway. Add them all up and it comes to $78, but let’s call it $60.

If Blodget is running $60 worth of ads but getting only $15 of revenue, then either he’s discounting massively, or he’s not even coming close to selling out his inventory, or some combination of both. And the same is true even if he’s making $20 or even $25 per thousand pageviews.

But here’s the problem: when Henry starts talking about the number of pageviews that a journalist needs to generate in order for the site to make money, he’s working on the assumption that every marginal 1,000 pageviews produces $10 (or whatever) in marginal extra revenue. Whereas in reality what tends to happen is that the ad sales team either fails to sell out the site’s inventory, or only does so by discounting so deeply that it’s really only their own fault that the revenues are so low.

This is one area where I think that Henry could take a leaf out of Nick Denton’s book, and refuse to run deeply-discounted ads. Doing that helps to improve the value of the brand among advertisers, and it also creates interesting opportunities for rewarding staff. At such a website, there will always be unsold inventory; at Gawker Media, that inventory is given over to the Gawker Artists program. At TBI, the inventory could be given over to staff journalists, in proportion to their pageviews the previous month, to donate to whichever organization they think could make best use of some free advertising on the site.

But for the time being I think it’s actually quite hard to say that a journalist with underperforming pageview numbers is being uneconomical for the site, especially when the ads you’re running on his pages are nominally worth four times what you’re being paid for them. That looks much more like a problem with the ad-sales team to me. The editorial team, after all, is clearly already producing way more content than the ad-sales team can manage to sell at decent rates. Maybe they should be the ones being fired.


Comments are closed.