Social Media Success Is More Than Intelligent Algorithms

I was reading Mashable over a cup of coffee this evening, and the news about Amazon and Facebook recommendations integration caught my eye.

To cite the specific paragraph that both intrigued and solaced me –

Now Amazon users will be able to link their Facebook (Facebook)  account to their Amazon account. At the outset, this will allow Amazon to show you recommendations based on your Facebook interests and activity.

This integration will also help you figure out what kinds of gifts your friends might like by letting you better search for your Facebook friends’ Amazon Wish Lists and by pulling in information from your friends’ profiles. Of course, it will remind you of Facebook friends’ birthdays, too. It will show you interests you have in common with your friends, and it will show you which items are popular with all the people in your social graph.

Technology fusions always create unexpected, ripple effects. In this case, Amazon seems to be using public profile data from Facebook and using it’s own intelligent algorithms to predict user preferences and behavior. Whether human preferences and behavior be algorithmically predicted, is debatable – and more about it in a while. But this was the reason behind my intrigue.

The solace lies in the fact that –

Amazon explicitly says that it will not share your Amazon account history with Facebook nor will it share your purchase history with Facebook.

A while ago, eBay’s connection with Facebook through the eBay Marketplace Application didn’t fly too well because of privacy concerns. Here’s a quick rundown on what that application did –

eBay Marketplace lets you see what your friends are buying, selling, and watching on eBay. Bonus: Add to your eBay watch list—and share your finds with your friends!

Although users could opt out of the sharing aspect of this application, with a feedback of 1.8/5.0 and a mere 523 active users, it has failed to create an impact. The major concern I had around this algorithm was that I didn’t want to share everything I do on eBay with my Facebook friends.

That get’s me back to my original point. The art of scientifically determining and predicting human preferences and behavior has been around for quite some time now – and has been gradually getting better. Merchandising algorithms routinely do this for up – and cross sells. Advertising networks too are perfecting this art.

I have a fundamental objection to this. No matter how intelligent the algorithms are, there isn’t a way to determine and predict human behavior and/or preferences with absolute accuracy. There always is the human element that makes is possible to fudge this determination. This is true on both sides of the equation – I can very easily create false impressions on an intelligent algorithm about what preferences, behavior and criteria for user segmentation.

In the Amazon Facebook example, I for sure would be a bit uncomfortable picking gifts based on recommendations. I’d rather go with gifts that feel right – rather than what an algorithm thinks are right.

For me, gifting is sacred. I choose my gifts carefully, and one of the criteria they need to pass is that they need to be personal. I’d rather go with a personal gift, rather than a recommended gift. Given an option between choosing a handwritten note and a recommended item, I’ll choose the former. And I’ll happily pass over the algorithmic intelligence to use my own feelings and emotions – and pick up a plant or a artisan music compilation.

The same logic also holds true in my social media world – which is much broader and deeper than just personal friendships. I prefer to slow down – go slow – and keep the human touch in my social world. I like it when my own logic and intelligence are not obfuscated by algorithmic automation.

At the end of the day, success in social media – and success in life – has a lot more human touch to it than these algorithms have us believe.

What do you think?

(Image Credit: theloushe http://www.flickr.com/photos/theloushe/)

4 Responses to Social Media Success Is More Than Intelligent Algorithms
  1. Johnny Laird
    Twitter: johnnylaird
    July 28, 2010 | 6:09 am

    …really nice blog, Kapil.

    You are now in my Google reader!

    J
    Johnny Laird recently posted… The new voice of Afrocyberpunk

    • Kapil Apshankar
      July 29, 2010 | 1:44 pm

      Thanks, Johnny. Good to see you here. And I hope you come back often 🙂

  2. mikemost
    Twitter: mikemost
    July 28, 2010 | 2:14 pm

    well put and you are now in my reader as well 🙂

    • Kapil Apshankar
      July 29, 2010 | 1:44 pm

      Welcome, Mike. Good to have you onboard!