Google's Avinash Kaushik offered a series of tips -- squeezed in between an impressive series of one-liners -- for how Web analytics professionals and those driving Web businesses must evolve in a Web 2.0 world. Kaushik, who describes himself as an author, blogger, and analytics evangelist -- delivered his insights at a Web 2.0 expo New York session.Here's a rundown:

1. The critical importance of competitive intelligence and the Web's role in helping companies assemble it. "The Web is the best competitive intelligence tool in the world," Kaushik says. He likens the failure to use such data to driving a car at 90 miles an hour with the windshield painted black, then scraping off the paint and realizing "you're going 90 but everyone else is going 220 and you're going to die." He urges Web managers to leverage some of the wide array of tools on the market, most of which are free. That was a not-very-subtle plug for Google's own analytics tools, probably the only downside to a very engaging presentation.

2. Use English, not data dumps, when trying to get top-level backing for Web initiatives and in reporting on whether they worked or they didn't. A clearer way to describe the "bounce rate" at your site? "I came, I puked, I left." Reports can best be understood -- and leave the least room for interpretation -- when individual items are coded in green (good) and red (sucks). Yes, Kaushik likes the words "sucks" and "puke."

3. Move away from page-view oriented analyses of performance to event logging. A few examples: analyze the top 20 pages on your site and their bounce rate (a.k.a., something people are puking on instantly). Look at keywords that have delivered 20% higher traffic over the last 7 days, as well as keywords that have delivered 20% lower traffic over the last 7 days. Focus on business goals, not data points. His key message: most top 10 lists rarely change, and offer little real insights into the "events" that are actually driving your business.

4. Listen to the voice of the customer. Kaushik recommends simple three-question surveys that ask: Why are you here? Were you able to complete your desired task? And if you weren't able to complete your task, why not? Respondents' write-in answers can be incredibly helpful. 5. Any Web business that expects to be successful must be disciplined and effective at testing and experimentation. "Most Web sites suck because HIPPOs (highest paid person's opinions) create them," he says, adding a HIPPO's view too often overrides common sense and doesn't factor in the reader's experience. The only way to overcome that obstacle, he says, is to "prove them wrong fast" with data that demonstrates when and how their Web mandates don't work.

Is your company taking full advantage of Web data to see how you're performing against competitors? Are you using event data to understand whether you're giving customers what they really need?