Ebay Knows What You Did (Bought) Last Summer and Is Not Afraid To Use It…

Running successful banner ad campaigns or “Display” campaigns are a really big challenge for many marketers. For the most part, this ad inventory is used for branding advertising which doesn’t show a direct response measurement and is therefore too cost prohibitive for your average small advertiser. I get asked all the time how to run successful banner ads. The secret lies in big data and programmatic advertising.

Big data and Programmatic marketing is the big buzz words for the last year and things are starting to get crazy. The simple explanation is every interaction you have with a website is a datapoint that can help a marketer target you with more relevant offers. The most classic example everyone is aware of is site retargeting, where you buy display advertising in Real Time targeting visitors who recently visited your website.

What many people don’t realize is that you can target datapoints and interactions that happen off your site, on other websites. For example, if someone searches in Google for “Digital camera reviews” and clicks on a link to ConsumerReports.com, you can retarget that person through a company like Chango. This is called Search Display Retargeting and works almost as well as paid search, but not quite.

This idea of using datapoints and interactions that happen on other websites to target visitors with banner ads is the basis of Programmatic Advertising. These opportunities became available through what’s called Real Time Bidding platforms which let you bid to advertise to visitors in real time. In other words, when a visitor visits consumer reports, Chango drops a cookie on their browser, and sees that the visitor came to the site from google searching for digital camera reviews. So, now a Camera Retailer can retarget that visitor across the internet in real time. Every time that visitor goes to a site that shows these types of ads, it quickly checks for the cookie and if its there, it says, “show this ad”. I am obviously oversimplifying this process, as it is a live auction model where various ads are competing for that spot, but the idea is simple. You can learn about a customer from their behavior on other websites, and use that information to target them with more relevant offers.

Currently lots of websites use their customer data for this type of targeting, but there is another type that is very powerful as well and that is customer data big retailers like Amazon own. Since amazon knows what you buy on their site, they can let you advertise with that data off their site targeting similar shoppers.

Yesterday, Ebay joined the party and announced that they are now offering a media buying service that will let advertisers target people in ads off of ebay using ebay purchase data. For example, if you bought a cell phone on ebay 2 years ago, your contract is likely up and verizon might want to target you with an ad for a new cell phone. If you bought shoes on Ebay, perhaps Zappos would want to target you, etc.. While ebay wont share you customer data, they can pool this information into targeting buckets, and create a “profile” to let you target people who likely bought shoes or likely will buy shoes, etc…

Since there is so much available advertising inventory when it comes to display advertising, any little trick to find a positive ROI can mean a huge windfall for a company and perhaps its for this reason that Ebay requires you to go through them to target against their data. I would love to see this in action and will try to find out how to get setup with them.