Step 3 - Checkout - Page 10

Finally, we'll address the issue of dropped shopping carts. After we've determined that our checkout mechanism is as streamlined as possible, it's time to start personalizing.

We need to build a rules-based system. We decide to construct an XML file containing our rules, so it is easily parsed by our application. This file contains rules and consequences — for example, if the item in the cart is a John Deere model lawnmower, then display an ad for John Deere lawnmower replacement blades. You can place many of these items on the checkout page — and in fact, we might argue that the more items you place on the checkout page, the better chance you have of getting the user to buy one (remember though that there is a limit before the customer becomes exasperated at having to wade through other items to find the submit button).

More enticing to customers would be a rules-based sale system. "Buy x John Deere lawnmower and y blades, and get 10% off." This is a very common practice used in brick and mortar stores, but seldom used in online shops.

One type of this method has recently generated some flak in the online shopping community, but it is an interesting concept nonetheless. You can combine data from the user's activity to provide them with discounts and options. For instance, if this customer spent a lot of time "browsing" it may mean he's more likely to abandon his shopping cart, so present him with a greater discount. Or if this customer has bought fertilizer in the past, prompt him if he's run out of fertilizer and offer him some more at reduced prices.

The problem is that customers get confused about what the actual price is, and when multiple visitors team up to find out that they're all being offered different prices, you may be in a heap of trouble. However, done sparingly, this can be a very helpful mechanism.

Finally, another good method is to allow customers to compare your prices versus other retailers directly on the checkout page. "Our John Deere model x costs $300, but Home Depot has it for $315." Customers often know that they can get better prices online, but having a confirmation provides customers with a feeling that they are making the correct decisions, and gives them more bravado at buying things. Don't forget to mention other benefits as well.

Another simple trick is to prompt the user. Many users simply spend a lot of time looking at the checkout page, with their mouse pointer over the submit button. If you see that a user is spending a lot of time doing nothing, you can prompt them (via a dialog box, perhaps) that buying here is a good idea and safe, with price benefits.

ACME likes this mechanism, because the higher-ups already know which items need to be pushed or which are going on sale, so it is very easy to incorporate this information straight from the top brass. Thus, rules-based personalization is implemented.

Step 1 - Making Customers Happy - Page 8
Everything you Need to Know About Personalization: Part 2 - Page 6
Conclusion - Page 11