Once you’ve factored in the collection methods, kinds of reports, and testing capabilities you’ll need for your analytics, it’s time to choose a platform.
Eric T. Peterson (http://blog.webanalyticsdemystified.com/weblog/2007/07/the-problem-with-free-analytics.html) concluded that if you’ve deployed a free analytics solution, you’re probably:
Only casually making use of web analytics
Understaffed in the analytics department
Lacking experience with web analytics in general
Far from criticizing free tools themselves, Peterson makes the point that you’re much more likely to take analytics seriously if you’ve paid for it. What’s more, you get what you pay for—free tools provide information, but require much more manual effort on the part of marketers to make and test their changes. In a small organization, analytics often becomes a lower priority than selling or design.
When you pay for something, you’ll have access to support, and you may even be able to impact the road map of the product, depending on how common your request is or how much clout you have with the company providing the service.
On the other hand, Brian Eisenberg, CEO at FutureNow, Inc., says:
My philosophy has always been to “get good at free then pay.” There’s no sense paying for something until you really operationalize its use. With today’s free tools offering 65–85% of the functionality of high-end tools, I am not sure free is only for the causally involved. About 30% of paid implementations also have Google Analytics or Yahoo! analytics installed.
At the entry level, analytics tools can be divided into two broad groups. Real-time tools, like those from Clicky (Figure 5-56), show you what’s going on right now, and are the closest you can get to spying on individual users, giving you their locations, IP addresses, and so on.
Real-time tools concentrate less on goals, outcomes, and conversions. They are useful for root-cause analysis and quick answers (who tweeted that message, who blogged about me first, and so on), but are not designed for long-term decision making.
Goal-oriented analytics tools, on the other hand, have less detailed information. They tend to be tied to backend business models (for example, Yahoo!, Google, and Microsoft’s tools automatically integrate with their search and paid search keyword systems). This is consistent with their business models. Services like Google Analytics want you to convert more people through AdWords so that you and they can make more money. Therefore, companies who create these tools want to help you optimize conversions without getting down to individual users.
Some solutions may offer both real-time and long-term perspectives.
If you need to keep data to yourself for legal or ethical reasons, or if your visitors’ browsers don’t connect to the public Internet, you may have no choice but to run your own analytics platform.
If you can use a third-party solution, we suggest you do so—you’ll get faster development of features and useful functions, like comparative reports that show how you’re faring against others. The only exception to this rule is if you have a business that’s tied tightly to some custom analytics.
If you can’t get a copy of your data from your analytics provider, you can’t leave. Having the ability to bring your data in-house—or better yet, to import it into an alternate analytics provider—means you can negotiate better pricing and keep your service providers honest. This alone is a good reason to care about data portability.
There’s an important second reason that’s central to the theme of this book. As we’ll see in the closing chapters, you’ll want to combine analytics data with other information in order to make better decisions about your business. Your analytics solution should either be able to import third-party data for analysis, or export analytics data to a data warehouse you run so that you can analyze it yourself.