If you want sustained meaningful Web metrics that will truly help your business, here are six Web analytics implementation do’s and don’ts to help prevent false starts, confusion and disillusionment.
- Don’t: Buy tools and start tracking before you’ve determined why you need to track.
Do: Before you start to track anything, understand clearly how you expect Web site analytics to improve your organization’s bottom line. Define your improvement goals with specific targets.
- Don’t: Track anything and everything.
Do: Make sure you count what matters. Ensure your Web analytics process measures Key Performance Indicators (KPIs) that will jolt people into action when they aren’t on target. You know you’ve got the right KPIs when being off-target would definitely be detrimental to success.
- Don’t: Take short-cuts selecting your Web analytics tool. Do: Ensure you select Web analytic tools that help you deliver clear, timely and relevant KPI reports without lots of secondary data processing, so that your Web analysts can recommend confidently and quickly what you should continue to do and what you need to change.
- Don’t: Expect ‘silver bullet’ performance from Web analytic tools, expecting insight without any human interpretation.
Do: Staff adequately and know what skills to look for in a Web analyst. Automated tools and reports alone do not generate insight. Automation simplifies analysis. Humans glean insights.
- Don’t: Keep learnings to yourself.
Do: Frequently share learnings and results with a wider and wider audience within your organization. Web analytics is relatively new and unfamiliar. So evangelize freely, excitedly and often about how continuous testing, observing and tweaking enables your organization to hone its competitive edge.
- Don’t: Treat Web analytics, particularly training, as a one-time investment.
Do: Invest in expanding and deepening Web analytics knowledge within your organization. Web analytics is a field of great depth and continues to evolve. Encourage learning on-site specifically for your organization, at vendor learning centers, or more generally through Web metrics courses such as those offered as a 4-course online Web analytics training program at the University of British Columbia or a 1-day Web Analytics course at the University of Toronto.
* Jupiter Research report, companies in the United States, March 2005