eBay Enterprise Bids on Personalization
John Sheldon, eBay
When most people think of eBay Inc. they think of its consumer marketplace. But it also has a robust B2B presence in eBay Enterprise, which runs Web stores for more than 1,000 brands and retailers across all of its solutions, including Aeropostale, Ralph Lauren, and Toys “R” Us.
EBay Enterprise leverages multichannel personalization for its clients, notes John Sheldon, its head of growth programs. For example, eBay Enterprise can create customized banner ads that greet consumers when they arrive from its affiliates and see conversion rates increase by 10%.
Also, the company is currently beta testing the ability to leverage first-party data from across eBay's three divisions—the marketplace, eBay Enterprise, and PayPal—to give its retail clients more information about consumers before they even land on a specific retailer's site, such as what type of shoppers they are, how often they shop, what categories they prefer, did they purchase, and how they behave.
“What we're trying to do is model people and aggregate them based on the data points that we know,” Sheldon says. “More data is better personalization. One data point does not make good personalization.”
Sheldon spoke with Direct Marketing News and broke down the basics of personalization and what marketers need to know.
What's your advice to marketers just starting their personalization efforts?
I would start with some basics. First is welcoming them based on where they arrived from…. It lets them know that you understand a little bit more about them. It's particularly important when you're dealing with people coming from an affiliate.... Just being able to differentiate between repeat visitors and first timers is another great place to start. It's easy to do technically, and there are some obvious things you can do to reintroduce a new customer to your brand, [such as] how your site works.
Is one type of data more telling than others when it comes to personalization?
There's nothing more telling than what the person has previously bought from you as to who they are and what they're looking for. We all, of course, have the counter- example—you make one purchase on Amazon or you go to Netflix and put on one movie for your kids and the rest of your life you're getting recommendations for them and not you. Ignoring that exception for the moment, knowing what people have bought previously is by far the most telling and insightful way that you can personalize experiences.
How do you strike that balance between being relevant without pestering?
First and foremost, performance dictates that. A promotion or a banner ad that doesn't perform is the first indicator that you're not connecting with the consumer. I'm an old-school direct marketer and I think [marketers should] try a lot of things and follow up on the things that work best. To some extent, the consumer tells you what works or what doesn't work for them. You'll learn over time—based on geography, age, or other things— that certain things work better than others. There are plenty of studies out there that show that there's an enormous difference in creepiness factor [between] 40- and 50-somethings versus 20-somethings and teens. [Younger consumers are] just much more comfortable with that data being available and being used.
What are some misconceptions marketers have about personalization?
I would hate for [marketers] to think that it's easy because it's not. It takes a lot of muscle and rigor to make intelligent personalization work. The second is that everybody wants it all the time. I'm not sure that that's really true. Consumers on a hard-target search don't want to be distracted by your beliefs about something else that they're interested in. Combining those things together, you need to very thoughtfully wade your way into creating personalized experiences.