So in my last post, I re-introduced the concept of my intelligent shopping agent. The foundation for the agent could be built by querying the consumer and analyzing past buying behavior. But that’s just the start. The focus of this post will be indicators (beacons?) which provide a deeper lever of understanding into a consumer’s tastes, and thus can start to intelligently make recommendations. That’s part of the point of the agent. It should always be on the lookout for potential products, deals, offers, sales, new trends, and more which might interest you.
Indicators:
-Search. This is a pretty easy one to understand. If you search for ‘red cashmere sweater’ or ‘cuisinart blender’ it makes sense that you might be interesting in acquiring those products. The intelligent agent will start to learn about brands or products that you’re interested in by these simple queries. But the value of search for the agent doesn’t end there. Search queries can make up a very complete picture of your life. We don’t typically think about this on a daily basis, but privacy advocates obviously do. Through my searches, the agent can figure out my socioeconomic status (Wealth. Am I searching for high end goods or coupons? Education. Am I searching for worldly issues or more pedestrian facts and figures? Occupation. Am I searching for computer courses or fundraising tactics?) It can determine my life stage (Am I searching for mortgages? Am I searching for a wedding planner? Am I searching for a SAT prep course? Am I searching for the AARP? Am I searching for income comparisons between two different cities?). There are some generalizations in there, but the idea that your searches can tell a lot about you and who you are today or who you’ll become tomorrow can be pretty powerful, especially when combined with other indicators. Interesting to note that Facebook doesn’t have complete search information.
-Display Ads/Re-targeting. Through the wonderful (old is new again) world of display advertising, we’re all carrying around a lot of cookies. These cookies are great indicators of interest. Each site you visit says a little something about you. The intelligent agent will pick up on your interests in the same way that re-targeting technology from Criteo or Remarketing technology from Google follows you around the web in order to provide ‘interest based‘ advertising. As Google explains, “Remarketing allows you to communicate with people who’ve previously visited key pages on your website, giving you a powerful new way to match the right people with the right message.” Google allows you to work remarketing into your AdWords campaign…think about the potential lift that marketers should experience with proper re-targeting or remarketing campaigns. Tying it back into the agent, all the sites you visit are indicators of something. Tied together with other indicators, your browsing history will become a cornerstone of the intelligent agent.
-Your Social Graph. In particular, I’m talking about Facebook. I could probably write a paper on the indicators which can come from your social graph, but I think everyone groks the value here, so I’m going to go a bit light. Don’t get me wrong, though, as Facebook integrates more and more with the web in general and the time on Facebook (or Facebook powered sites) increases, our social graphs could potentially have the strongest set of indicators about us (one of the reasons that Google is putting over 1000 engineers to work on social). Which is why when I first started to re-visit my intelligent shopping agent a couple months ago, a number of people stated that Facebook is the intelligent agent I’m talking about. I disagree. It’s part of the equation and Facebook probably desires to be the intelligent agent, but without search, it’s just not good enough. We’re not going to stop searching. The best intelligent agent needs multiple indicators, so search without social is just as bad as social without search. Expect Google and Facebook to be not so comfortable frenemies for a long time.
-Conversations on Facebook and Twitter. A big part of Facebook’s advertising opportunity which hasn’t been realized is conversational marketing. Right now you can target Facebook Ads by a multitude of filters (30-45yr old female who works at Cisco and likes Steve Madden), but you can’t just target ads by the keywords in conversations as Google does with AdWords on the side of Gmail. This will be a huge opportunity for Facebook. Twitter is just starting to explore this type of conversational marketing with their promoted tweets. Basically, if I post on my Facebook wall or tweet that I’m looking for running shoes and friends start making recommendations, my intelligent agent should be all over that (Display ad, notification that there’s a nike store around the corner – assuming I’m ready to purchase, price range I can expect to pay, note about an upcoming 10K that my friend registered for, etc. Too many indicators to talk about when it comes to social.
-Location! 4 years ago we had our cell phones, but we weren’t doing as much with them, and we definitely weren’t indicating what we were doing. We were more passive. Now we’re tweeting about what’s happening around us all day AND checking into various locations. More indicators. So as opposed to talking about the typical shopping experience providing what you want, when you want it, our intelligent agent (though Foursquare’s API, for example) can now add the qualifier: where you want it. If your agent knows you check into a restaurant for a sandwhich a couple times a week for lunch, it should be smart enough to tell you that the store around the corner has the new Donna Karan shoes you’ve been looking for (searching for online? conversation about with a friend? 8 months since your last Donna Karan shoe purchase?) and entice you to take a couple minutes to walk to the store and check it out.
Read Part 3.
[...] Read Intelligent Shopping Agent – Part 2 [...]
[...] Part 1 and Part 2 before reading this [...]