Having lots of data isn't enough, though. A good leader is going to make decisions based on his or her personal beliefs. The metrics matter only in as much as they validate a leader's religion or inform the overall vision. They become the historical record that educate decisions, not define them.
Without that vision it's very easy to become a slave to the numbers. I got stuck in that trap at InfoWorld while we battled daily to find new visitors and identify ways to get current visitors to return more often and click on more stuff. I became locked into the world of unique visitors, page views, pages per visit and visits per visitor. Those numbers reinforced the story of the page view model.
It was either the occasional debate with Jon Udell that rarely ended well or the constant fight to move those dials that finally got to me. I realized that squeezing margins from page views was a poor excuse for defining success on the Internet.
The CPM model isn't broken, actually. It works well enough, just as the CPC and CPA models work. But all those models have ceilings when limited to your own domain. Once you reach a certain audience saturation level, the success of your business is dependent on your ability to keep getting just a little bit more out of each asset. It won't be long before you're spending as much to drive your metrics as you're earning on them. Then your only way to profitability is cost-cutting. Ugh.
The Internet is amazing because it makes it possible to connect things. If you operate your business like an island, that's what you'll be. That realization has been reinforced coming to Yahoo! where the explosion of publicly accessible APIs and content available via RSS have opened up a whole new world of opportunities.
I can't say I understand all the metrics that inform success in the openness strategy, yet. But there's no metric that will convince me the page view model is going to continue to dominate.
Tags:
strategy,
management
Comments:
Re: Balancing data and vision to make decisions
by
Anonymous
on Mon 13 Mar 2006 07:40 PM EST
If you were unsuccessful in finding proof points from scouring the data, would you have questioned what you were trying to prove?