Google Believes Online Ads Could Be Worth More Than Print Ads by 2012

When you look through the delightful PowerPoint presentation prepared at Google on the newspaper industry, a few things stand out.

Now, this graph, while entertaining, isn’t a useful comparison of anything. (You could compare construction employment, or Medicare spending, to GDP, and it would look meaningful, but I’m not sure it really would be!) All this means is “newspaper ad revenue isn’t a component of GDP growth,” which, well yes.

This seemingly nothing little slide actually reveals huge truths about the newspaper industry. Which is that subscription numbers-while they’re the backbone of what newspapers can sell against!-bring in basically negligible income. It also shows the number one mistake of newspapers. If editorial costs are a mere 14% of revenue, then the hugest mistake of the last 5 years has been the brutal hollowing out of editorial staff, which results in less content to sell and, indirectly, in less good content to sell against.

This presentation runs with a piece recapping long talks with Google by James Fallows.

The piece is increasingly vague as it goes, with an alarming lack of examples, but that’s Google for you-their method is that they build stuff and see what works and what sticks, and that model totally works in the long term. But one thing that is clear is that everyone at Google seems to believe in massive growth in the role, frequency and even cost of display advertising online.

And then there’s this startling sentence from Fallows, summing up what he heard from everyone at Google: “Sooner or later-maybe in two years, certainly in 10-display ads will, per eyeball, be worth more online than they were in print.”

The immense gulf between the cost of print ads per eyeball and the value per user of ads online is usually viewed these days as something unresolvable; and so it follows that sites that are attached to newspapers believe they just can’t support themselves even on huge traffic. That Google believes otherwise is striking.