2. The shopping cart is too hard to find.
Clearly identify your shopping cart linkâ€"your customers will reward you by filling it up.
Whatever you do, make sure your site's shopping cart is a clearly identified link placed in the top right corner of the page, traditionally a high-click area. Moreover, make it available on every page in the same place. In our labs, users became frustrated when they couldn't find a site's shopping cart right away. Also try bringing your site's shopping cart area to life. Sports retailer The Sports Authority does that with its dynamic shopping cart. The cart displays the latest item added, allowing the customer to keep a running tab more easily. The company also made sure customers could click on the cart icon to start shopping. In contrast, Drugstore.com offers a live cart halfway down a left-side tool bar but ignores the top right convention entirely.
Major don'ts
Don't expect people to instantly notice items added to a dynamic cart. When shoppers add items, route them to a cart summary page each timeâ€"even if it seems redundant. This page indicates confirmation and invites the shopper to check out.
Also, if you use a dynamic cart it's best to list only the price of the last item added and the number of items in the cart. No one wants to be reminded that they're racking up a US$600 tab while casually shopping. When they get to the cart summary page, customers then have a chance to remove items they no longer want.
You have missed the most important point. Users abandon shopping trolleys - because they can! It's no problem. There's no comeback. They aren't real trolleys. No one gets hurt.
Why consider it as a problem? It's no different from window shopping. More people can do it and more will walk away because they had no real intention of buying in the first place.
You have lots of good advice about how to encourage potentiall buyers to purchase, all of it valid. But where's the problem? In a different paradigm, people will behave differently. Just because they put stuff in an internet shopping cart, ther is no implication that they ever intended to buy. Many will be just testing the system to see what happens, to try to understand how it works.
It may be that with the most perfect system, most visitors will never buy, so expending effort to improve the percentage is not cost effective.
My point is, understand the motivation of users and fix problems that really exist.