Why Web Analytics Tools are Just Not Enough

We're all using them, some of us even drill down to their deepest functionality but for many critical goals in web marketing and especially in B2B marketing and sales Web analytics are just not enough.

Why is that?
Well, while Web Analytics tools are obviously important, they lack in some areas that are critical for effective sales and marketing over the web. They are very good at showing the big picture but lack the ability to pinpoint relevant, high potential visitors and more importantly then that, they lack the ability to do something about the data they supply.

I always used to end-up with a good feeling about all the new visitors and referrals to my company web site but could not really monetize on the traffic or the data I saw. Think about it this way, Web Analytics tools let you "see" your visitors but don't let you "touch" or "interact" with them and since your visitors are usually anonymous - thats where the story ends - nice statistics but few results.

Here's what I mean:
I guess were all using some kind of web analytics tool, whether its Google Analytics, PIWIK , Omniture, webtrends or any other web analytics tool, they are generally similar in their approach and the values that they offer. They are also quite similar in what they do not offer. So you can get a lot of info, trends, graphs, pie charts etc. You can get good information on your stronger traffic sources, content and visitor locations and track your online campaigns. But if you want to use your web site to get new leads and turn them into customers all of the above mentioned data can be quite frustrating since it gives you a "sense" of whats going on in general, but not much information about specific visitors or any direct means to convert them into leads.

To summarize, Web Analytics tools are a great first step to turn your web site to an effective marketing machine as they give you a sense of what comes in, which campaigns worked better etc. BUT, if you really want your web site to convert visitors into customers you have to think Reactive.

Online Financial Applications, Top 10 Use Cases for Real-time Reactions

I am often asked what are some common use cases for the ReactiveWeb real-time behavioral reaction service in the financial sector so here they are starting with the .
The basic idea behind all of these scenarios is the same - React to online behavior immediatly and according to its business relevance and value.
Here we go...

1. Sales: Customer Up Sell & Cross Sell.

The Behavior > An existing customer examines his current investment portfolio. This customer's porfolio has a positive yield and he has an avarage order history of over 3 stock orders per month.

The Reaction > Offer this customer a VIP brokerage service proposal online (a free consultancy meeting or 3 month free service may also do the trick...)

2. Sales: Potential Lead generation.

The behavior > An existing customer enters his online banking application, after reviewing his balance he visits the loans sector of the site, visiting more then 4 pages of long-term loans.

The reaction > Send business event to CRM systems, "John Smith, interested in long term savings, event date: 10/08/2008"

3. Sales: New Customer Hooking.

The behavior > An anounimous customer visits the banks commercial website, he enters the student's loans sector visiting more then 5 pages for longer then 5 minuts.

The reaction > Interact with online user by 1. offering a specific loan and 2. Offering a local branch consultant to contact upon contact details submition

4. Customer Service: Improve Customer Retention

The Behavior > An existing customer checks the cash-out options on his savings account or insurance policy. He does this for the first time in atleast several months and goes into the details page of the terms of termination/cash-out.

The Reaction > Offer him an immediate online loan or connect to local branch for financial planning. Save the customer before leaving and sell another product.

5. Customer Service: Assist in Transaction Completion

The Behavior >A customer tries to buy stock through the banks online application, after filling out all the relevant details in the 3 step form he submits the order and recieves an error as a response. The order amount is higher then $20,000. He tries twice again and recieves the same error.

The Reaction > 1. If this is a VIP customer offer immediate online human assistance, via audio or video 2. If the error is relatedto missing information, pop up a window highlighting the error cause (only after the second time) and offer the solution.

6. Customer Service: Improve Online Self-Service

The Behavior > A VIP customer is trying to retrieve a specific report or any other action and repeats it more then 3 times with no success.

The Reaction > Check service representative's availabilty and offer immediate audio/video call to resolve issue

7. Customer Service: Customer Experience Management

The Behavior > Detect slow serving pages or user session from specific geographical regions with slow avarage response rates across their sessions

The Reaction > Alert system applications and personnel with specific information

8. Security: Provide Detailed Audit Trail

The Behavior > Any monitary related action performed by the customer or even lower levels of behavior can be audited non-intrusively according to predefined rules

The Reaction > Log detailed user session to any logging or SIEM application and offer different levels of auditing according to regulation or customer requirements.

9. Security: Detect Fraudeulent Behavior

The Behavior > Detect specific fraudulent behaviors not only in the geo/ip/browser level but in the actual application behavior

The Reaction > 1. Feed existing fraud detection applications in real time 2. Invoke scoring processes 3. Redirect/Block user sessions

10. In-house Advertising

The Behavior > A customer with a large positive balance enters the web site

The Reaction > Display investment related banners

These example use-cases are obviously a small portion of the various relevant use cases but I hope they provide a guideline to the high potential in real-time reactions to online behavior