Facebook and Amazon.com partnered Tuesday in what could be one of the social network’s most important integrations yet. Amazon.com now offers a personalized page, where consumers can see product recommendations influenced by friends and their own tastes. They get notifications on when friend’s birthdays are coming up and suggestions on what to buy for them.
It’s a big deal for a number of reasons.
A deep Amazon.com-Facebook partnership could help corner Google in the e-commerce market. One of Google’s most lucrative use cases is when consumers search to decide which products to buy. Costs-per-click on product keywords like refrigerators, TVs and books are often easily more than $1. But if consumers start looking toward their friends to find out what to buy, they could be able to bypass Google altogether. (This is the bet of a recent batch of startups like Blippy and Swipely that share consumer purchases.)
Such a partnership could also lay the groundwork for Facebook to get key pieces of data, so it can start to quantitatively value how social recommendations translate into sales. (The company says it’s not getting purchase history on specific individuals.)
The longstanding problem with brand advertising is that it is notoriously hard to measure effectiveness. A brand can run a TV campaign, a viral video, or make those innovative Old Spice ads on the fly, but how does a company really know whether these efforts actually boost sales? This is why, as angel investor Chris Dixon argues, ads that generate intent are undervalued relative to ads that harvest intent. Online advertising, as it exists today, rewards the last ad that a consumer clicks on before they buy a product (like Google’s search ads or Amazon’s affiliate links). But it doesn’t reward all the display or engagement ads that might influence a consumer for weeks or months beforehand on content sites or social networks.
Now if Facebook can build out a partnership with Amazon, where it can get key pieces of aggregated data back to study how product sales change in relation to social recommendations or ad campaigns, it can build models to show how financially valuable it is on average when a user “likes” a product or brand.
Facebook has been trying to quantitatively prove for several years that even though users come to the site to socialize and not to look at marketing, its ads pay off. Its partnership and study with the ratings firm Nielsen last fall was part of this. It needed hard data to prove that its social ads improved brand awareness.
An e-commerce partnership could go much farther than that. Facebook might be able to prove that not only do its ads and Pages raise brand awareness, they also lead to better sales too.
Of course, Amazon would get a lot of value out of such a data-sharing arrangement too. It already has a very powerful recommendation engine. And it will only get better if augmented by a consumer’s professed interests on their profile page and the habits of their friends too.
VB's research team is studying web-personalization... Chime in here, and we’ll share the results.