How to Predict Results from SEO - and How Not To Wil Reynolds , Seer Interactive - Jun 28, 2011

Should you spend the time and money to optimize your site to get higher rankings for certain keywords? How much extra sales will you achieve? What's your Return On Investment (ROI).

When you're trying to predict results you'll get from organic search engine optimization -- especially ROI -- you find all sorts of complications. Let me outline problems you'll encounter in click-through rate data, personalized search, as well as changes in the search engine results pages: image results, local results, shopping results, and news results. Then, I'll explain where to get the best data to predict your ROI.

Industry Reports of Click-Through Rates

Click-through rate data that you find in "industry reports" can be extremely misleading. For example, a study by Danny Goodwin ("Top Google Result Gets 36.4% of Clicks," Search Engine Watch, 21 April 2011) compared the relative number of clicks from the top 3 or 4 positions. But if you try to correlate such ranking data to click-through predictions from Google AdWords Tool, you're not going to get a very accurate results.

Here's why. Different types of queries have different kinds of click-through rates.  For instance, people often type a URL in Google's search box rather than in the URL bar, sometimes yielding a very different result.

Consider two other types of queries.

1. If you're searching for something inexpensive like socks, you might click on the first 2 or 3 listings. The first one to look reputable and have good pricing is probably going to get your purchase.
2. But if you're searching for some expensive service like "Applicant Tracking Systems" for your company, you might you go to page 2 to do your complete due diligence before recommending a solution to your boss? I would.

So, using click-through data as reported in industries studies is misleading if you are trying to predict how many clicks you can get from a specific ranking.

Complications of Personalized Search

Personalized search, that is, Google deciding what results to show you based on past searches, introduces another complication. Browsing behavior impacts personalized rankings, making it difficult to predict rankings from one person to another.

Complications of Ever-Changing Search Results Pages

Click-through rates are impacted by the various blended search result results now being offered by Search Engines. Here are a few examples. It's hard to predict how the images and different types of results surrounding your company's entry will affect click-through rates.

Definitions

Here's a definition of the disease from National Institutes of Health shown on top when you search on "diabetes."

Scores

A search for "philadelphia phillies" shows a baseball image with the latest score.

In these cases, Google gives you a result that may answer your question. You may not even have to click any further to get your answer.

Images

A search on a celebrity may bring the top set of results as images, which will obviously impact your click-through rate as well and affect eye-scanning patterns on the results page.

Local Businesses

Local businesses are increasingly found in search results, since Google can tell the approximate location of the searcher and show nearby businesses. As a result, the click-through rate of your number 1, 2, or 3 rankings go way down.

Here's a search on "photography" made from my office in Philadelphia that resulted in local Philadelphia photographers.

News and Shopping

News results and shopping results show up regularly in the search engine results pages. They, too, disrupt the regular click-through rate flow that is reported in studies expected click-throughs, like the one cited at the beginning of this article. Observe both news and shopping results that show up on a search for the keyphrase "bike shop."

Broad Phrase Clarifications

And finally, for very broad phrases, especially one word phrases like coffee, paint, or furniture, Google will ask you to clarify your query to encourage more specific searches.

You can readily see how difficult it can be to turn "industry studies" into some kind of a model that allows you to predict your Return on Investment for SEO. You're just not going to get meaningful results. So where can you get reliable numbers?

Understanding SEO from Google Webmaster Tools

Here's how to proceed. If you haven't already, set up a Google Webmaster Tools account and give it time to accumulate some data. Then click on your site on the web > search queries option and you'll see a graph something like this.

Now, before jumping into data, take some time to look around at your options by clicking on the filters button in the top left.

Here are some instances when you might want to use these filtering options:

Queries containing/excluding. This is a simple field that lets you narrow down your results. Remember that your site gets traffic on tons of queries, so if you have to sift through 4,000 different queries, it can be daunting. This option will let you focus in on specific query types.

Search. This drop down gives you options for different types of search: image search, mobile search, and video search. You might want to look at mobile, since it is an emerging platform with a lot of focus from Google right now. It may be important to see how click-through rates are impacted by searches on mobile devices.

Locations. This option simply shows your Impressions / CTR / ranking position by country. Depending on your goals, this might not make much difference, but if an international audience is important to your business, click-through rate may vary from country to country as people in different places search differently and see different search results formats.

Traffic. With the traffic filter, for example, you can exclude queries under 10 searches. Again, due to the prevalence of the long tail, you may end up with a lot of queries with little traffic, making it harder for you to sort through the data and get what you need.

Observing Different Click-Through Rate Patterns

Let me show you two examples of how click-through rate is impacted by position. In the top example -- for a free service -- you can see wild swings in click-through rate (CTR) when ranking moves between positions 1, 2, and 3.

In the second example -- for an e-commerce product -- CTR is quite steady throughout those same positions.

This flies in the face of the reports we see over and over again about the relationship between position and click. Use your own data and stop relying on industry reports.

The difference between the two examples is that the top one is for a keyword that revolves around a free service. So if the user picks the wrong freebie, there is nothing lost but time.

The second example is for an e-commerce product with a keyword that results in a sale of $100 to $500. Picking the wrong item might be a big deal. So we we see people not just clicking on the first item they see, but doing their due diligence. They are clicking throughout the search results at a very different pace.

The Bottom Line

The point is, when trying to understand the Return On Investment from organic search engine optimization, respect the fact that search results pages are in a constant state of flux. Results vary based on your browsing behavior. Therefore, study the high traffic keywords report in Google Webmaster Tools to indicate what click-through-rates you can expect when you attain higher rankings for some of your target keywords.