iEntry 10th Anniversary
RSS
Contact
 |
|
High-Level Search Engine Reporting 
By: Gary Angel 2007-07-25 Most web measurement efforts begin with a high-level view and then descend into increasingly detailed cuts of the data. Search Engine Marketing is no exception. The best place to start a Search Analysis is with a report on total Paid Search Traffic, total Organic Search Traffic and total Site Traffic - preferably trended over a good chunk of time. Web measurement tools will provide this data easily. At such a high-level, you can't expect dramatic learnings. But there are some important take-aways. I've talked before about how reporting is really the gradual building of context around a business issue. For Search Engine Marketing, part of the context is how significant the overall effort is in terms of total site traffic and total prospect traffic. If Search is driving 50% of your site traffic, then optimizing SEM is going to be a major concern - but you might also assume that the bulk of the Search opportunity has already been captured. So instead of looking for growth opportunities, you're more likely to be focused on efficiency opportunities. Conversely, if 5% of your traffic is from Search, then optimization may be less important but growth opportunities are almost sure to exist. As we get into Search Analysis, you'll see that most types of analysis address one of these two concerns - growth or efficiency - but not both. So having some perspective about where your site is can help determine which analysis paths to pursue. Similar reasoning also follows when comparing the percentage of Organic to Paid traffic. When one or the other dominates, the likelihood is that you'll be looking for tuning opportunities in the dominant channel and growth opportunities in the secondary channel. This may seem basic and obvious, but it's surprising how often both analysts and marketers don't (at least consciously) understand how to use big-picture numbers like % of traffic. The next drill-down is to the Engine-Level. I actually like to extend this a bit by including drill-downs to campaign-type. So I think of an Engine-Level report as having rows like:
2. Organic-Yahoo 3. Paid-Google-Search 4. Paid-Yahoo-Search 5. Paid-Google-ContentMatch The basic Traffic Metrics used in web analytics are Page Views, Visits, Visitors and Time on Site. All are interesting. But with Search Engine Marketing, when you look at traffic numbers to understand share, you're primarily interested in the Visit count. When you start looking at how successful Search, individual Engines, individual Campaigns and individual Keywords are, then this focus will change. For understanding share of traffic, however, visit should be your measure of choice. When sites service a mix of Customers and Prospects, the tendency is to compare Search Traffic to the Prospect Universe - especially for Paid Search. This is good in theory but tricky in practice. The fact is that Customers will often source from Search - even Paid Search. So counting ALL Search Traffic as a share of the Prospect traffic will overstate Search's true influence. Many, many basic assumptions about Search turn out to be incorrect, and one of the more damaging fallacies is thinking that Search traffic is always Prospect traffic. What can you learn from this Engine-level share reporting? Quite a bit. As with Search in general, comparing the sourcing rates by Engine can help identify growth opportunities. Most sites should expect their numbers by Engine to track reasonably to the wider world. Where site numbers diverge from this expectation, it often signals a potential for growth. If Yahoo only sources 10% of your organic traffic, then you almost certainly have growth opportunities in Yahoo SEO. In addition, these overall sourcing rates are handy to have when you start drilling down into the site. If Yahoo sources 20% of your organic Search traffic, and you find that it sources 50% of your organic traffic for a particular topic or area, you may have found either a Google optimization opportunity or a great prototype for Yahoo optimization. As with so many high-level numbers, these share numbers form the context within which more detailed measurements can be compared. Other Posts in this Series: Introduction, Searchnomics Issues, and SEM Data vs. Web Analytics Data -Don't forget to register early for X Change - Like "Top Gun" for Web Analysts! The Early Registration Discount ends July 31st! Use Coupon Code XCForum at http://www.semphonic.com/conf to reserve a spot!- Comments Tag: SEM, Search Engine, Marketing, Web Analytics Add to  Del.icio.us |  Digg |  Reddit |  Furl Have a bookmark! -  About the Author: Gary Angel is the author of the "SEMAngel blog - Web Analytics and Search Engine Marketing practices and perspectives from a 10-year experienced guru. |
|
||||||||||||||||||||||||||||
| SearchNewz
is an iEntry, Inc. ® publication
©
1998-2009
All Rights Reserved. Privacy Policy - Legal - Sitemap - Contact Us - RSS Feeds - Newsletter Signup |
