Google Analytics -- How to Blow a First Impression 

Over the weekend, word leaked out that Google was launching a free web analytics package. This was pretty cool news. If you've ever messed around with a web analytics product (WebTrends, ClickTracks, even Site Meter), you know you can get some handy information beyond just "5 people visited my site yesterday." At my last job, I spent a whole lot of my time using ClickTracks to do log analysis to try to determine how effective search campaigns were, if landing pages for marketing materials were working well, where people were falling out of the site when trying to sign up for one of the products they sold.

Now that I've moved on to my new job, I'm looking at web analytics from a different perspective -- trying to figure out how to deliver it to our customers. We've been looking at a couple of alternatives, to move beyond the standard "here are your logs, we'll run Webalizer for you, or you can do something yourself." When I heard about Google Analytics, I made a note to check it on Monday.

Monday rolls around and Google puts out the official announcement of Google Analytics. I get there early Monday morning and the site is swamped and barely responsive. Hmm. How can you underestimate demand for a free web product aimed at web users? I was able to get through and sign up to track some sites, and I threw the little bit of Javascript onto a couple of pages (not here, however; I want to give Google some time to settle down before I try tracking anything more substantive than a single page). It said to expect results in 12 hours.

12 hours later? Nada.

24 hours later? Nada.

36 hours later? Nada.

48 hours later? Holy crap, it works.

2 days to start returning statistics for something that launched on Monday. All the while, not a peep from Google, who have this bizarre affinity for not commenting when things aren't going well. You'd think they'd follow up on the GoogleBlog with a "Hey, we really underestimated demand, so we're adding servers as fast as we can. Expect it to be ready on Wednesday." Then I wouldn't have kept checking in and wondering why I should even consider using it.

Now, of course, why complain about something that's free? Well, because it's not free. Google is giving me a product in exchange for my data, which they use to grow their massive revenue. Gmail is free, and in exchange, they put ads next to my mail. The search engine is free, and in exchange, they put ads next to my search results. My former blog home of Blogger is free, and in exchange, they get a place to put more ads. See where I'm going? None of it is free. Google is making billions from people's data. So it is not too much to expect them to get something right the first time. And why does blowing the first impression matter? Because right now, I'm helping to decide if we're going to point thousands of customers to Google for advanced analytics. Thousands of customers = lots of data for Google to gain more revenue from. Just because they launch these "free" web apps doesn't mean that they don't lose potential revenue when they have such awful launches (see Google Groups, Google Reader, Google Mail).

In the end, it looks like it's a pretty damn solid product. It's got all your normal metrics, in a handy interface (a good use of Flash, for a change). It seems to track everything you'd normally want to track, and the only downside is that you're sending your data out to Google. That could be a bit scary for companies who are doing eCommerce and what not (though they wouldn't have
to put the Javascript snippet on that page). For me, it's a minor haggle to get good web metrics, something I'll probably pay more attention to when I've finally built out the site a bit more.