Robi Ganguly’s Big Ideas

Amazon has come a long way since its early days as “just” an online bookseller. Having been introduced to the company in its infancy (back when it was a plucky Seattle-area Internet business), I’ve followed the company fairly closely over the years and really come to admire it. Whether it was its early forays into personalization or the way it went about executing against its vision regardless of Wall Street’s fickle and unrealistic expectations, the company has convinced me that it has its head on straight and is building for the long run.

Given that I spent several years working on the advertising business of Yahoo!, it’s only natural that I would take some of that experience and apply it to my thoughts about Amazon. In doing so, I’ve developed a point of view on how Amazon could leave their mark on the online advertising space.

First things first, Amazon brings an existing set of strengths built on an understanding of the online customer that should shape any offering that comes from them. As I see it, that implies some core principles for Amazon advertising:

1. Personalization matters: Look across the landscape of online companies and ask yourself this: does any company do it as well as Amazon? If you’re an Amazon customer, you get recommendations from them while surfing the site that are generally quite good. More impressive though, is the fact that the times when they choose to email me a recommendation, they’re often right. This leads to us a second, and related point.
2. Be conservative with your recommendations: Many people see the advertising space as a question of ,“How do we get advertisers to buy against our inventory?” Amazon’s email recommendations come infrequently and, as a result, hold more weight in my inbox. I believe them to be conservative in choosing when to promote a product and if I’m right, it’s one of their great strengths. In approaching the advertising opportunity, they should be thinking about the question, “When are we justified in creating inventory and who gets to advertise against it?”
3. Your customers are your salespeople: Jeff Bezos said in an interview last year with Charlie Rose that, “The Internet is a word-of-mouth accelerator.” EXACTLY.  Amazon Associates is perhaps the web’s most popular and acceptable affiliate marketing program.  Amazon Web Services has grown up around enabling this activity in a very robust way, with a huge variety of endpoints that sell products for Amazon. There is huge power for the future of commercial communications (marketing, advertising, customer service etc.) in this model.

Given those principles and adding in my perspective that just copying existing ad models isn’t that interesting for Amazon, I believe that there are 2 initial areas where the company could focus its time:

1. Open up the recommendations process to paying advertisers, particularly those in the media space
2. Revamp the Amazon Associates program to more easily involve ALL of Amazon’s existing customers AND extend it to appeal to all digital consumers

In short, I believe that Amazon has an opportunity to really change Internet advertising by embracing the idea that personalization is a problem best solved en masse. Creating a marketplace around personalization, such that advertisers can influence it to make recommendations better and consumers can benefit from positive outcomes they’re already driving has the potential to change online commerce.

I’ll add some more detail to these ideas.

The first area of focus:

Open up the recommendations process to paying advertisers, particularly those in the media space.

As I mentioned above, Amazon is already routinely sending out product recommendations, notifications of forthcoming product releases and suggesting products I might like while surfing the site. To my knowledge, these recommendations are currently “pure”, generated algorithmically. I believe that with a very conservative approach to the opportunity, combined with a LOT of testing and the institution of robust feedback mechanisms (a customer should always be able to say NO and tell Amazon when it’s doing a bad job), Amazon could upgrade its accuracy by inviting advertisers to compete to show up in the recommendations.

An example of this: Say that you bought Malcolm Gladwell’s book, Blink: The Power of Thinking Without Thinking a few weeks ago. Imagine that you had previously bought his book The Tipping Point: How Little Things Can Make a Big Difference as well. Now, normally, the recommendations engine at Amazon would most likely notify you in a few weeks that Gladwell has a new book coming out Outliers: The Story of Success.

This is a pretty good process and does a lot for the consumer, but there’s an opportunity to improve the consumer’s knowledge about an author they probably like a good bit while making some money for Amazon. You see, Malcolm Gladwell isn’t just an author, he’s a writer for the New Yorker as well. The New Yorker, being a for-profit publication, is interested in having more subscribers and readers of its content, so it stands to reason that if they could work with Amazon to reach loyal purchasers of Gladwell’s books, they would take advantage of that opportunity. Again, it would be really important for Amazon, in creating this opportunity, to be conservative in its application and highly focused on soliciting feedback from its customers about the recommendations, but done well, it should add a whole new (and welcome) dimension to the Amazon customer experience.

Remember, this is but one example of how this could be applied. It should be noted that you can think about this with regards to other forms of media rather easily (digital music, video and television make perfect sense). From a process perspective, Amazon could experiment with this concept, refine the technology and the user experience and then reap the eventual rewards that would accumulate as the feedback loop on the recommendations makes Amazon’s database much more intelligent than competing systems.

Moving on…

The second area of focus:

Revamp the Amazon Associates program to more easily involve ALL of Amazon’s existing customers AND then extend it to appeal to all digital consumers

In the last year, Amazon has introduced some enhancements to its Associates program that makes the process of sharing links and recommendations far more easy and powerful. However, the opportunity to expand this program to every single one of Amazon’s customers strikes me as a no-brainer that’s not being fulfilled. Right now, in order to be an Amazon Associate, you have to be aware of the program, navigate to its location on the site and go through a sign-up process that is relatively intimidating for the average customer. The goal should be to make every Amazon customer an Associate at the outset.

Every time a customer shares a product, they should be given a unique link that attributes any sales to their account, at the very least giving them future purchase credit. Instead of focusing massive amounts of attention on just the developers, as Google, Microsoft, Yahoo! and others are doing, Amazon has the opportunity to expand past the developers to the end consumer. Of course the company should be enabling the developers and the Amazon Web Services offerings are exceptional.

BUT, the opportunity to enable others is much more larger than just  the development community:

In our emails, IMs, Twitters, Facebook messages, blog posts, comments and all other forms of digital communications, we are influencing others to take action and make purchases. Amazon’s Associates infrastructure could hold the key to really unlocking the value in these small moments of influence. In so doing, they stand a strong chance of mobilizing the world’s largest distributed sales force: the consumers. Implementing this idea certainly carries a great number of risks around fraud, spam and general abuse, but because Amazon has dealt with many of these issues already in releasing and maintaining their products, they have exceptional opportunity. This idea really deserves much more discussion on its own, but I think that you can ask yourself a few questions here to really think about Amazon’s opportunity here:

• Who else can successfully run the experiments in this arena that connect consumers and commercial action?
• Who else offers the infrastructure and support to developers that could quickly extend the compensation model for consumers to to other sites?
• Who else can work on this area without risking cannibalization of their existing lines of business?

This post has run pretty long already, so I’m going to draw it to a close so you all can digest. What I’d love to hear in the comments is how you react to these ideas and what you think needs to be fleshed out more for discussion. I’ve got a lot more to write about on this topic, so let’s get a discussion going.

Popularity: 97% [?]