• by Erik Sass , Staff Writer, January 11, 2011

Although I have expressed skepticism about the threat posed by Facebook to Google in the past, J.P. Morgan analyst Imran Khan says the threat is quite real in a note to investors circulated on January 3rd. Khan bases this analysis on the proportion of traffic to large commercial sites coming from the two sources, Facebook and Google.

Looking at the Internet's biggest online retailer, Amazon.com, Khan notes that the proportion of traffic arriving from Google searches dipped slightly, from 20% in 2009 to 19.6% in 2010. Meanwhile the amount of traffic arriving at Amazon.com from Facebook more than quadrupled from 1.8% to 7.7%. When I looked at the current stats from Alexa, roughly 4% of Amazon visitors had visited Facebook just before, while about 24% had visited Google. What's more, the overall percentage of traffic coming to Amazon.com from search sites in general seems to have declined from about 22.5% in the first quarter of 2009 to 20% presently. And Khan found similar trends at other big online retailers, like eBay, and non-retail content publishers like the New York Times Web site.

These stats are pretty convincing: Google does seem to face a threat from Facebook in terms of Web traffic to third-party sites, and therefore in advertising revenues derived from this traffic. I was prompted to wonder, however, how much overall traffic to Amazon.com has grown over the last couple years: it's possible that traffic has expanded so much that Google's actual numbers are even or growing, even as its share in percentage terms declined.

Indeed, it's worth noting that the share of total world Internet users who visit Amazon.com has more than doubled from 2% in the first quarter of 2009 to about 4.5% presently, also according to Alexa -- even as the world Internet population has grown by tens of millions of individuals over the same period. Similarly, Amazon.com's share of total page-views increased from 0.2% in the first quarter of 2009 to about 0.4% today. And the proportion of "bounces" (single page views followed by the visitor leaving the site) decreased from 38% to 30% over this period as well.

Summing up, Amazon.com seems to have roughly doubled its share of the world's Internet traffic, as measured both by number of users and page-views, and is also becoming "stickier" or more engaging for people who arrive there. All of this is good news for Amazon, and good news for Facebook, which is playing a larger role in sending traffic there -- but it may also temper the seemingly bad news for Google.

Facebook's share of Amazon traffic is increasing hand in hand with Amazon's traffic growth overall, which seems to suggest that Facebook, rather than competing with Google in a zero-sum game (where one player's loss is another gain), is actually helping generate new traffic and contributing to Amazon's long-term organic growth. This would make sense, as Facebook is -- among many other things -- a new way of discovering content. And because Facebook's function is essentially different from Google, it can play a growing role in online content discovery and e-commerce -- without necessarily threatening Google's core search business.

In short, maybe Facebook and Google can play complementary, rather than opposed, roles in content discovery and sales?