The Demand for Extreme Differentiation in Your Content Strategy

We've all heard it said many times: "Content is king." In general, this is true, but the statement needs some qualification.

We'll begin by digging quite deeply into the reasoning behind query deserves diversity (QDD) before looking at the impact of QDD on your content strategy, to illustrate the importance of extreme differentiation.

The Role of Query Deserves Diversity

QDD is the notion that a search engine may intentionally diversify results to create variety in the search results even though normal ranking signals would suggest something different.

Let's look at a search result for a naturally ambiguous search query: "Jaguar":

serp-jaguar

Notice that web pages for the car and the animal are mixed together across the first five results. We can't get access to the raw calculations that Google uses to determine search ranking order, but what if these results are influenced by QDD? What if the normal ranking signals would suggest ranking that looks more like this?

serp-altered-jaguar

Here's what the demand looks like in the AdWords Keyword Tool:

keywords-jaguar

The results are sorted by local monthly results in descending order. The first result, [jaguar], is ambiguous, and all but the seventh result, [jaguar animal], pertain to the car. I dug into this a bit further, and downloaded the results table, subtracted out the ambiguous results, and calculated what percentage of the remainder were queries related to the animal. That number was 9.1 percent.

What if the link profile of the corresponding web pages more or less mirrored the search demand? That would suggest a rather different ordering of the results. Here's what Open Site Explorer says about the page authority for a few of the pages:

domain-authority-jaguar-sites

This certainly suggests some re-ordering of the results if we look at just the page authority. Note that the 5th result is the international page for the Jaguar site and used just for navigating to the right regional site for the brand.

So we need to dig a bit deeper into how this might work. Let's start by outlining the types of factors and intents that may be present for this very generic query:

  1. The first major factor is that I am sitting in Massachusetts. Chances that I want the Jaguar Brazil site are relatively slim. This is also why chances are lower that I want the regional navigator page at jaguar.com.
  2. Looking at the search query data from the AdWords Keyword Tool, more than 90 percent of the queries (after ambiguous results) relate to the car.
  3. OK, result number 1 explained. Without any use of link analysis or any other factors, we have determined that the number 1 result should be the home page of the U.S. version of the Jaguar site.
  4. Next question is, what percentage of queries are satisfied by this result? I have no way to know that for sure, but we know that 90+ percent of the audience wants cars, and this is a branded search query, and the information on all the various car models is readily accessible from the home page of the U.S. site. This study of CTR by SEOmoz includes information on CTR that suggests that 40+ percent of people click on the first result, but then they look for other results on the brand, perhaps by independent sources.
    However, the SEOmoz information doesn't cover the case of an ambiguous query like ours. But, as you can see, the third result in our SERPs is a Wikipedia page, which is a page related to the brand query, but that is not under control of the brand owner.
  5. This leads me to the need to speculate that results 1 and 3 cover 70 to 90 percent of the various search intents. The high end number of 90 percent I take from the AdWords Keyword Tool data we examined. The low end number acknowledges that there may be other intents such as looking for international version of the site, a specific review article, and other variants of the search phrase where information on the car is the basic user goal.
  6. With that in mind, what is the next most likely intent (i.e., what should go into the second slot)? Well, 9.1 percent of searches appear to have an intent related to the animal. If between 70 and 90 percent of the queries are satisfied by results 1 and 3, 9.1 percent of all queries is a very large portion of the remainder. Finding the most authoritative page for that intent to put in the second position seems like a really good idea!
  7. What of the Jaguar OS or the Jaguar guitar? Neither of those even made a dent in the top 100 search queries, so they don't make the first page. Not enough people have that intent when they are thinking about jaguar, so if that is your intent you need to formulate a more specific query.

Notice how the above discussion more or less explains the ordering of the results and there isn't mention of link analysis or social mentions in that entire discussion? The domain authority analysis didn't really explain it.

What we found is that there is a good chance that the positioning of the Wikipedia page for the animal is easily justified by analysis of query data (let alone the comprehensive CTR data that Google has) without any link analysis at all. Link analysis may explain why the Wikipedia page is the first page on the animal.

I admittedly did a fair amount of extrapolation of the available data in the above discussion, but my goal was to illustrate what QDD is all about. The reason animal results are needed on the first page of results, even though they are only a small percentage of the total query intent is that they are a large part of the query intent not satisfied by the other results.

How QDD Impacts Your Content Strategy

The key segment of the discussion was the part on search intents. Our entire focus was on the types of information that people might be looking for, and how we diversify results to make sure that the different SERPs satisfy different possible results.

In a recent interview with Matt Cutts, we discussed how sites that offer the exact same information in the results, even if they aren't duplicates of one another, aren't likely to rank highly.

Here is a key part of the discussion with Cutts, starting with my comments after showing mockups of three non-duplicate different web pages that had the same basic information:

Eric Enge: The resulting page isn’t a duplicate of the first, but the information provided is the same. So they go back and click on the third result and get yet another non-duplicate page that still does not have what they want. At this point, they’re very frustrated. It turns out the information they’re looking for is what frogs eat (this was not on the example pages shown), and they’re not finding the information they’re looking for.

The reason I use this example is I am trying to show clients that being non-duplicate is not enough, and they need to do more to expect to rank in the search results.

Matt Cutts: That’s absolutely right. Those other sites are not bringing additional value. While they’re not duplicates they bring nothing new to the table. It’s not that there’s anything wrong with what these people have done, but they should not expect this type of content to rank.

The discussion with Cutts wasn't about article spinning. It was about multiple different sites producing high quality content, but all of which had the same information. Taking QDD into account, all the pages were targeted at the same intent.

To make this concrete, if you create a brand new site focused on providing advice to consumers buying homes, and you make your revenue from advertising, you must figure out how to differentiate it. You may want to write an article on mortgage tips, but it has been done a few times before:

serp-mortgage-tips

The same is true for home inspections:

serp-home-inspection-tips

This reality goes deeper than Google's link algorithm as well. Potential linkers, or people on social media sites that share content related to buying homes, aren't going to share your new articles on mortgage tips or home inspections, unless the articles say something very new, different, and compelling. You may require these articles on your site to make it complete, but don't expect to build an SEO or social media strategy around it.

What is your strategy for extreme differentiation? Don't have one yet? Now is a great day to get started!