 There is no longer a question whether or not a website owner should be using web analytics. Instead, the important question this year will be, "What can I get out of web analytics?" Getting the most out of your website data begins with asking the right questions.
Based on my own experience and some informal surveys of our top users, here are the top ten questions that online marketers should ask of their web analytics tool.
Which keywords should I invest in?
Whether you buy keywords through paid search or optimize your site for natural search rankings, you need to know which keywords are worth your investment. View your top 100 keywords by visitor count and then take the top 50 in order of highest time-on-site. These keywords and the longer phrases that convert well are great candidates for investment.
Which ad campaigns should I cut?
Whether it's PPC click fraud or just a poorly placed banner ad, it's time to stop allowing non-producing ads to drain your budget. Use a web analytics campaign report to assess time on site, conversion and revenue for each campaign. Adjust or cut campaigns that aren't performing.
Which referrers should I reward?
Often the top producer of converting visitors, good referrers (including bloggers and frequent forum posters) should be rewarded. Short of sending cash or a new television, there are several ways to keep active referrers happy: Consider a link back, free product, membership in an affiliate program, "first to know" status, or just good old fashioned "thank you".
What did visitors want that they didn't get from my site?
If a significant number of your visitors searched for "Purple Fabric" while you only offer "Green Thread", you may be missing an opportunity to meet the expectations of your visitors. Follow the behavior of short visit (or early exit) sessions based on search keywords and in-site search and you'll know the intent of the visitors.
Which pages are the search engine robots missing?
If a page isn't viewed by a search engine robot, it won't show up in the search engines. All too often, our most important content pages are missed by the robots due to a complex URL or accidental blockage in the robots.txt file. A robot report will show which pages of your site are missed by each search engine.
What do my pages look like to the search engine robots?
Search engine robots read your pages very differently than a human on the other side of a web browser. Use a robot simulator to ensure the proper text, H-tags, links, title and meta-description are given to the robots for each page.
What changed that didn't make it to the top 10 list?
External factors cause changes in visitor behavior that can indicate a marketing concern or opportunity. For example, a new blog mention might cause a sudden spike in converting visitors, or a competitor might have outbid your PPC keyword. Although these events can miss the typical top-10 lists, marketers who catch them can take advantage of the opportunity.
What content is most persuasive to my important visitors?
Each page (or page group) has some level of persuasion in moving visitors toward your site goals. Use a conversion funnel to determine which content is most influential to visitors. Segment visitors in the funnel view to determine how different visitors types (i.e., which ads or search engines) respond to each stage in the funnel.
How does one visitor segment behavior compare to another?
There is no such thing as an average visitor. Each type of visitor behaves in a way that that can reveal important characteristics when segmented and viewed distinctly. Compare time on site, for example, among visitors from each search engine. Often these comparisons are the most revealing of all data.
The above questions are generic to most businesses with a web site. Your own questions will be more specific to your business and will measure the performance of specific marketing campaigns and programs such as email, printed advertisements, or even radio and television spots.
Be sure to allocate enough time to react to the answers to these web analytics questions. The effort to change a campaign or site content is significant and often underestimated by the new web analytics user.
Lastly, I recommend a periodic schedule where each important question is checked and addressed. Most marketers look at these each month. High volume sites require weekly or even daily checks. Once your question list and schedule is established, you will be able to measure results and ultimately, get the most out of your web analytics.
 |
Subscribe to this comment's feed
