Practical Ecommerce

SEO: Rankings Don’t Indicate Performance

Editor’s Note: Meet Jill Kocher at Ignite 2015, our conference on Sept. 16 and 17 in Dallas, where she’ll present two sessions: “The Essential SEO Checklist for Ecommerce Companies, for 2015” and “Keyword Research for SEO Success: How Do Consumers Search for Your Brand and Products?”

Owners and managers are obsessed with rankings. They search on Google for something, they see an immediate result, and they make assumptions on their search engine optimization program’s performance.

Here’s how to dispel the myth that rankings indicate performance.

Competition: What Else Is Ranking?

Competition in search results goes far beyond your traditional organic search competitors. Yes, you want your site to outrank your competitors’ sites for the best chance to be seen, win the click, and make a sale. But many ecommerce search queries trigger elements in the search results that draw attention away from organic listings and push those results down the page.

The example below shows how a Google search result can push organic listings so far down the page that there are none near the top. The searcher would have to scroll to see the number one ranking site.

A personalized Google search result for “picture frames.”

A personalized Google search result for “picture frames.”

In the example above, the combination of four paid ads, product listing ads, and the three local search results obscures Bed Bath & Beyond’s number one ranking in the search results The ads and local results in first view are visually engaging and could snag more clicks from searchers than paid ads normally would when organic ads are available.

Your C.E.O. may see some of these ads and elements in her search results, or she may not. Your ranking tool may be able to report these elements or not, as well, leaving your visibility lacking into your site’s ranking. This is one of the biggest reasons one can’t rely on rankings as a performance indicator.

Personalization: Whose Rankings?

Rankings are highly personalized, to the point that there are no unpersonalized rankings for most searchers. The rankings your C.E.O. sees on her iPhone will vary from the rankings you see on your computer, or your customer sees across the country on his tablet.

There is no longer a way to say definitively, “We rank number one in Google!”

Rankings are personalized based on a number of factors.

  • Location. The most obvious of all: Search engines determine the location of the searcher using the device’s IP address or cell signal and deliver localized search results. Locale can impact both the map-based local search results and the traditional web search results.
  • Device usage. Rankings will also vary between smartphone, tablet, and desktop devices. Mobile devices will get links into apps and sites with mobile-friendly experiences rank higher. So if your C.E.O. is searching on her iPhone, the results may look very different from your desktop results.
  • Past search behavior. When you’re logged in to your search engine’s account, your search engine has access to all of your past search behavior. This data is used to modify the search results you see. For example, if you’re logged in to Gmail on your computer, you’re logged in to your Google account and Google search.

Google searches on an Android device are also automatically personalized in this way. So when your C.E.O. searches for her name or the company’s trophy term repeatedly without clicking on your site, she’s skewing her own search results even further.

  • Social behavior. Google has full access to social behavior on Google+ and Twitter. Social content from friends and brands you’re connected with could be displayed in your search results, but not displayed in other searchers’ results.
  • Demographics. Search engines algorithmically place each searcher into demographic categories based on their behavior, to deliver more targeted ads. Until they learn otherwise from a profile you create, search engines may presume you’re an 18 to 22 year-old male, as they did with me thanks to my obsession with technology.

For fun, see who Google thinks you are in the Google Ads Settings. Search engines use this demographic data, and related data they glean from other sources, to personalize results as well.

These are just a few of the ways that search results are personalized. The combination of these and other factors can result in some interesting differences in rankings.

Most ranking-reporting tools will run return data that is unpersonalized. But how close to reality is that data if all of your customers are seeing personalized search results? Perhaps your ranking tool reports a result of number one for your most valuable keyword.

But Sally in Detroit, with her bulging wallet and a penchant for buying home decor online, sees your site ranked in the third position — if she sees your ranking at all. Is she going to buy from you or the site that really ranks number one on her screen?

If a ranking report can’t connect to what really matters to your business — revenue — then it can’t be relied on as a key performance indicator.

Representation: Which Keywords?

Tracking rankings typically means choosing a set of keywords and phrases that you’d like to rank for, and plotting your site’s performance in ranking for them. The words and phrases you choose can dramatically skew the appearance of performance.

It’s easy to choose keywords and phrases that you’re already ranking for, that include your brand, or that have low demand and competition. Doing so would make it appear, perhaps incorrectly, that SEO performance is stellar. Companies that work with search marketing agencies should be alert for this tactic.

It’s easy to choose keywords and phrases that you’re already ranking for, that include your brand, or that have low demand and competition.

Individual pages can rank for thousands of keyword phrases, most with very little demand. This is the long tail of search — the concept that the many phrases that represent 1, 2, 10, or 50 searches per month all add up to drive more traffic than the trophy keywords that you would track in your ranking report.

Tracking the trophy terms alone leaves you blind to the performance of the long tail, which drives perhaps the majority of your site’s organic search performance. Since a ranking report doesn’t represent the breadth of your site’s performance, it shouldn’t be considered an indicator of SEO success.

Temporality: When Did It Rank?

Rankings can change at any moment. Your C.E.O. may search for her pet phrase and become incensed that it doesn’t rank number one anymore, but it may rank number one again the next time she searches for it.

For SEO professionals reporting program performance, rankings are similarly tricky because the problem is multiplied across hundreds of keywords in multiple search engines on multiple devices. A ranking report is a snapshot in time. Its relation to the real state of your site’s and your competitors’ rankings is transitory.

While I agree that rankings are a necessary metric in managing SEO performance, they shouldn’t be considered key indicators. If you’re obsessed with rankings, hopefully this article will help.

In addition, each of my recent articles listed below will give you more insight into choosing key performance indicators for SEO, and using metrics like rankings to support program growth.

Jill Kocher

Jill Kocher

Bio   •   RSS Feed


Sign up for our email newsletter

Get the Practical Ecommerce RSS feed

Comments ( 4 )

  1. Sam Partland August 31, 2015 Reply

    Why wouldn’t you use the average rank data from Google Webmaster Tools to track performance then?

    That will give you rank data based on where you actually appeared for the given impressions, thus being rather accurate. It would take into account personalisation, along with location, then average it all out.

    • Carlos Rivera September 1, 2015 Reply

      Sam, great idea!

  2. Carlos Rivera September 1, 2015 Reply

    Jill, you predicted this many years ago when Google first implemented their ‘auto-complete’ feature as default on their search results.

    …things were so simple back then.

  3. Hariharan September 23, 2015 Reply

    In google webmaster, any way to analyze ranking of a site based on Gender,device type and geography?