Commentary
Why Rankings Still Matter In SEO Reporting
- by Janet Driscoll Miller , March 15, 2011
Yesterday at South by Southwest, Danny Sullivan held a Q&A session with the search engines, with Matt Cutts representing Google and Duane Forrester representing Bing. While I followed the session on Twitter, one comment especially caught my eye: there seemed to be some general consensus that rankings are no longer a valuable metric in SEO reporting. So is it time to abandon rankings as an SEO success metric?
The Case Against Including Rankings as a Marketing Metric
The main reason that members of the panel gave for discouraging using rankings as a measurement of success is that rankings shift so greatly from searcher to searcher because of a variety of personalization factors, including location, social circle, and other factors. Forrester even compared ranking measurement to a "sugar pill," emphasizing that a far better measurement of success is conversion.
The Problem with Measuring Conversion in SEO
While I agree with Forrester that conversion should be the main metric of really any type of marketing effort, measuring conversion from SEO isn't always a measurement of whether or not your efforts drove targeted traffic to your website. As with pay-per-click advertising, you can optimize for the most targeted keyword terms and drive traffic, but if the page the searcher lands on doesn't convey your message effectively, searchers will abandon the site. That's why SEO practitioners may not want to be held accountable for conversion, as SEOs are often not themselves in control of the messaging or design of the website.
advertisement
advertisement
Enter Conversion Rate Optimization
Conversion rate optimization (CRO) should be applied to more than just landing pages for pay-per-click campaigns. Why not use A|B testing tools like Google Website Optimizer (GWO) on your website's pages to find out what changes might improve organic conversion? Unfortunately, not many marketers are embracing this philosophy yet. In fact, last week on a panel at Pubcon South, I was asked if GWO could be used on website pages as well as landing pages. I commented that I'm still surprised by how many website designers and agencies don't use tools like GWO or Adobe Test and Target to measure success before scrapping an entire website, essentially "throwing the baby out with the bath water."
So should you focus on conversion measurement? Absolutely. But conversions are the measurement of many factors, not least of which are the methods that drive targeted traffic to the website -- including, but not limited to, SEO.
The Case In Favor of Including Rankings as a Marketing Metric
Or should you measure rankings, too? Rankings serve as a benchmark, as do other valuable metrics, like search traffic. And while you can't always account for all of the personalization variables, you can get a good idea at least of how your rankings are improving by using benchmarks.
But how can you get the cleanest, least personalized rankings measurements? There are two ways I recommend: Advanced Web Ranking and Google Webmaster Tools' Search Queries.
Advanced Web Ranking and Proxies
Advanced Web Ranking is a rankings measurement tool that allows you to collect "clean" rankings using proxy servers -- spreading the ranking requests over several servers at once, which speeds up reports and provides more accurate ranking results.
Google Webmaster Tools' Search Queries
A few weeks ago, I covered some of the benefits of using Google Webmaster Tools (GWT) -- one of my favorites. The Search Queries report shows the top keyword queries in Google that your website ranked for in the past 30 days. The report also shows valuable information such as impressions, average ranking, and click-through rate. However, like AWR, which is essentially a snapshot in time of the rankings at the moment you run the report, the Search Queries report in GWT shows the information spanning the last 30 days and does not archive data prior to that time.
Which Way to Go
So, in the end, should you include rankings or not? I do. But rankings can't be your only measurement; they are one of many metrics search marketers need to consider when determining search success. While I agree with Forrester and others about the shortcomings of relying solely on rankings for SEO success, rankings are a baseline to work from, and if you use the right tools, you can get great insight about your movement.
What should you measure? I recommend using a combination of rankings and analytics data (including ecommerce and/or goal data). But whatever you do, make sure your measurements clearly gauge the area you're trying to improve.
Great post Janet. I couldn't agree more. Clients tend to get a little too obsessed with rankings, and don't focus enough on conversions from search engines; because conversions are really the most important metric. But I always include rankings in reporting. I also use rankings as a sales tool, because this gives a great snap shot of present day performance vs competitors, and proves that the website is in desperate need of SEO in order to increase qualified traffic, rankings and hopefully, conversions.
Well said Janet. I agree rankings should not be the lone measurement of how well a site is doing especially since the keyword selection process can be subjective. There are long tail keywords with very low search volume which are fairly easy to rank for and there are high volume head terms which take significantly more effort to rank. A SEO campaign can be wildly successful garnering all #1 rankings but not be driving any traffic or conversions.
My preferred approach is the following holy trifecta. #1 Measure Traffic SEO Traffic (Including any cookied typed/bookmarked that was orginally first touch SEO) #2 Measure Conversions, (Including any cookied typed/bookmarked that was originally first touch SEO)& #3 Measure the number of keywords driving traffic and several buckets of rankings data for top keywords. These buckets can be, "TOP converting HIGH CPC SEM KEYWORDS ", "TOP CONVERTING SEO KEYWORDS", "TOP CUSTOMER LOYALTY KEYWORDS (such as keywords with high site engagement, such as page views per UU) , and so on. I have several buckets for measure rankings in the head body and tail, of keyword distribution and have considering bucketing users by phases of customer lifecycle and landing page types. That said, my final word of advice.. Dont rely on GWT, if you have more than 1000 visits a month, the query data from GWT is sampled and averaged data. I have found it pretty useless compared to other tools out there. And would highly recommend a combination of Omniture, and BrightEdge or Conductor. Thanks great post!!
http://seo-cubed.com
I believe that rankings play an important role in contributing to the overall climate of your search/SEO success, but as you point out, have many inherent pitfalls. I recommend that folks focus in on one a very small handful of their top keywords to monitor over time, but the list should be less than a dozen. CRO is by far and away the best method, especially for businesses that have a well established online lead funnel, so that everything can be tightly integrated from a reporting and analysis side. There are issues of attribution and lift from other advertising sources above and beyond SEO efforts, but the bottom line is that with proper analytics, CRO can provide you a successful route when it comes to identifying the success of your SEO. [more at www.dansoschin.com]
Excellent topic for discussion. While localization and personalization can definitely improve search results, I haven't seen any research on how often personalization would impact search, especially B2B search. By its nature, search implies a new area of inquiry, so why would it be affected by the user's history? Another research question: how many searchers are actually logged in to iGoogle when they start their search? These are metrics that really matter for SEO practitioners. What we know and GWT doesn't is what search terms our sites are optimized for...their sampling on a small site gives meaningless results.
Thanks, y'all, for the excellent comments and observations. If anyone is aware of relevant research, please let me know: jrasco (at) refreshweb.com
Janet - great write up on the seo rankings discussion from the sxsw conference. I agree SEO rankings should not be the sole source to measuring SEO success. Marketers need to concentrate not solely on ranking data but also the ROI of their search rankings.
Marketers need to take it a step further and understand what keyword phrases that people are actually searching on and if they are converting. This can be done by mapping your web analytics data to your keyword ranking data.
At gShift, we plot keyword phrase rankings over time and tie SEO and web presence actions along a time line. We also like to trend competitor web site rankings.
Conclusion, include rankings in your monthly seo metric reports, but include other seo metric data points (such as crawl dates) that can be related to actual rankings because it will show trends and insights.
Thanks, everyone for the great comments and discussion points! All good stuff. :)