This post in Banknet360, Customer Service: Striking a Delicate Balance (registration required) was pretty interesting, so here is my take on this.

Check these excerpts from this post

By year end, most major U.S. banks are expected to roll out basic live-chat services, according to one vendor. The conversations are meant to clarify product offerings and cross-sell additional loans to customers.

The basic assumption is that people visit web sites in search of some specific tidbit of information. If the bank can provide the answer and eliminate the need for — often fruitless — web surfing, customers will be happier. For example, users checking out information about home equity products at Bank of America’s web site may receive a pop-up message asking them if they have any questions about the bank’s products.

Now, here is my take on this idea. Lets make some rough calculations here.

Lets suppose that the number of sales guys available online at any point is 10. You need 3 shifts anyway, so that makes it 30 guys on hire, and suppose that each guy makes around 50K (not sure how much these guys make !!), so it adds up to $1.5 million.Add to this the cost of training, recruiting, office space etc. The total sums up to an approximate $2 million a year. And mind you, these are recurring costs.

Now, I have another alternative here. Check this out !

A bank offers loans, credit cards, investment options and other products. So, you want a visitor to your site to be able to ind the info that he wants, and you also want to be able to suggest products (that he is looking for) to him , and also cross sell in the process!

Think of a rule based question answer Web 2.0 Ajax driven wizard.

1. This wizard will guide the visitor through a series of carefully chosen context specific questions.
2. The answers to the questions posed till a point will determine the next set of questions.
3. The sequence of questions will be driven by a set of rules that are formulatedand written by experts
4. The wizard will complete by suggesting to the user the products that is most appropriate for him.
5. The Business Rules are captured and modeled using a BRMS

The success of any customer service initiative is in delivering the kind of service that the customer is looking for, and not what the service agent thinks is right. Now, the live chat model is greatly dependent on how qualified the agent is is.

If you ask me, it makes sense in automating the product suggest service, and put as much intelligence as is possible into the rules that drive the service, rather than investing into costly resources.

The rules driven product suggest service will also work out to be far less cheaper, and work in a far more predictable fashion, while delivering consistent service to the highest standards possible.

Moreover, an automated product suggest service can be tested thoroughly, and changes/improvements to the service will work out far cheaper in the than the agent driven live chat idea!

Additionally, the reasoning behind every suggest can be configured to leave behind an Audit trail, that can be used to improve the system over a period of time.

Now, You think about it!