Real-Time Online Profiling

Imagine you’re a salesperson in a retail store. A customer walks in and starts to shop. How do you interact with this prospective buyer in the most efficient and personalized manner, having never seen this person before? There are several ways to go about it.

You can watch where he goes first. If he heads to the sales rack as soon as he enters, you know he’s probably cost-conscious. If he starts browsing the high-end items, price probably won’t be an issue. As a salesperson, you’ll likely act differently toward these two types of customers. You’ll cross- and up-sell differently, too.

Often, you’ll see what kind of shopping bags the customer comes in with. If he was previously shopping at a high-end store, you may have a sense where to lead him once he’s in your establishment. As this may be the only clue you have to go on, you work with what you have. You must market in real time.

The online world can be quite different. Acting in real time may not seem as effortless to achieve. To this end, serving customers with a personal touch may not be simple, either. It’s more difficult to walk away from a live salesperson than leave a Web site, a frequent occurrence in the online marketplace.

Many online browsers are first-time visitors. They may have no previous history you can utilize to better serve them. They may never have completed a purchase on your site. You may not know which customer segment they belong to.

So how do you market to them in real time? How do you customize and personalize the online shopping experience for customers with no history? It starts with online profiling in real time.

There actually are similarities in the on- and offline worlds that help answer these questions. As with looking at the shopping bag in the offline store, see where shoppers arrive from online. What are the referral sources? Did shoppers arrive at your site from specific ad banners? What can you infer about them based on the content or the placement of those ad banners? Did they come from search engines? What keywords did they search?

This can easily define segment and specific interest. Then customize your site accordingly. Customers should be immediately differentiated based on these criteria.

Where do customers go once they’re on your site? If they go right to the high-end items, you’ll market to them differently than if they go to discounted items. Record all products they look at. Leverage this information on the fly. Modify what they see on your site appropriately. Offer different ad banners or special offers based on behavior and product selection.

If you’re able to personalize the experience further by offering an opportunity to speak with your offline sales team, arm your sales reps with this initial data. With this relevant information, sales reps will not only better service customers but better up- and cross-sell, too.

Hopefully, customers will return to your site after their initial visits. Use the information gathered to tailor their next online experiences. Keep in mind items they previously selected or viewed, and customize your site and email communications. Once you have information on a shopper, act in real time from then on.

Change visitors’ classifications as they move through your site, make purchases, or interact with your agents, even when they abandon an online order. By leveraging each action, you can predict what they may do next or what their interests might be. Then display applicable online offers, personalize ad banners, and send relevant email.

All these efforts should help you to build a rich user profile. When building that profile, keep the following in mind:

• Record user information over time and sessions to continually enrich the profile.
• Know referral sources: where the users came from and how they arrived at your site.
• Track behavior: frequency of online visits, sections/product categories visited, products viewed, and exit points.
• Request personal information: name, contact info, and demographics.
• Maintain full online purchase history. Try to combine it with offline information.
• Know responsiveness to online and email offers.

Profiling starts when a customer enters your site, not once a purchase is made. Shoppers may come to your site several times before acting on an offer or buying. Make the experience relevant as soon as possible to convert them into customers.