An article in CEO Magazine says that the latest trend in companies is to respond to their market and not predict.

Of course it is a wise idea to always keep your finger on the pulse of your target market, competitor and the economy. You have to know when they shift and change, and you have to adapt or die.

But it's also wise to use this data to be able to predict future behavior. I believe it is better to be proactive than completely reactive.

One of the best ways to do both - respond and predict - your markets behavior is analyze the metrics from your web site. A significant majority of our population is online.

You can get a wealth of data from your log files. It will tell you where they are coming from, what they are searching on, which word or phrase is getting the best traffic, how long they stay on the site and what they are doing there. When last did they visit and how often. Who buys and who doesn't.

Analysis of your web stats can help you understand your market. You can respond to the changes, make sound decisions and yes, even take a stab at predicting the future behavior of your customers.

I thought one of the best presentations at EMetrics was from Jared Spoole of UIE

He reported on a UIE study done with committed buyers - people who knew what they wanted to buy, where they wanted to buy it and were really interested in acquiring the item.

UIE gave them the money to make the purchase and, although under these circumstances you would assume a 100% conversion, only 32% succeeded.

How is this possible? The websites presented so many problems that 68% of these really committed buyers with money in hand (in this case not even their own money!) were unable to complete the purchase.

The point? Your analytics are often skewed by usability problems.

When they can't buy - they wont buy! You need to find out what content your users need in order to be able to complete a purchase.

Can they find all the relevant info they need before they buy?
One young man attempting to buy a birthday present for his fiancee could not figure how a size 6 translated to XS S M L sizing.

Is the site easy to navigate?

Can they find what they want to buy?

Does the buying process go easily and smoothly?

Your content - or lack thereof - can seriously affect your conversion rate and revenue.

Use analytics to show you where a potential problem lies - but do usability testing to find out what is actually going on.

I'm afraid life really caught up with me on my return form the summit!

Here is more info from EMetrics

Day 2: We got some excellent feedback from the front lines. We heard from HP and Amazon.com on their personalization and how it's working for them.

We also heard how Dirt Devil uses analytics and some interesting info on how hotel sites are battling with lookers vs bookers. There was a good article and study on this recently.

That afternoon we had a productive round table discussion and came up with a list of tools that make up a full complement in the web analytics toolkit.

1) Clickstream Analysis with reporting. Vendors : Omniture, Webtrends, Coremetrics, ClickTracks, WebSide Story.

2) Web Site Performance & Navigational Analysis. Vendors: Keynote, Maxamine.

3) Search Engine Optimization/Marketing (driving visitors to the site). Vendors: BidRank, Inceptor, EC Next.

5) Email Campaign Reporting & Management. Vendors: CheetahMail, Responsys, Constant Contact

6) Surveys & Satisfaction. Vendors: Forsee Results, Opinion Lab, Vividence

7) Clickstream and Off-Line Reporting Integration. (Key is an open API to get clickstream data into your CRM system) Many vendore have this capability

8) Usability. Vendors Include: Netraker, Usability Sciences, a lot of boutique firms.

9) Competitive Intelligence/Benchmarking. Vendors: comScore, Hitwise

10) Ad Management Reporting. Vendors: DoubleClick.

11) Brand Protection/Monitoring (RSS Feeds, Multimedia). Vendors Cyveillance, BzzAgent.

12) Forecasting/Predictive Modeling. Vendors: SPSS, Intelligent Results.

13) Enterprise Search Analytics (analyzing the effectiveness of search on the site). Vendors: Endeca, iPhrase, Mondosoft, Verity.

EMetrics Summit
Santa Barbara 2004

Day One:

Jim Sterne was in fine form today. His keynote on 'Darren meets Dilbert' was a brief history of web analytics and how it evolved from a pure IT (Dilbert) play into a Marketing (Darren) essential and is now being seen as a mission critical business tool.

The friction between IT and Marketing, and how to get and interpret statistics off the website, really hit home. It's always easier to see one's shortcomings when it's amusing - and the truth resonates. There were many heads nodding in agreement as Jim spoke.

The drawback as Jim sees it, is that IT thinks Marketing will figure it all out by themselves and Marketing is expecting IT to not only supply numbers, but tell them what it all means too.

So the missing link is an analyst who gets both sides. This point came up over and over again throughout the Summit. There are just not enough analysts.

Dilbert and Darren have to work together and find a way to get the numbers and then figure out what they mean, so Marketing can act on the data to cut costs and raise revenue and customer satisfaction.

Stats are pretty useless unless you act on the information.