SERP Out – Brand & query refinements

Walk with me…. If there is one area that is anything but a science, it’s keyword research. It’s not always self evident as to which terms or strategy is best employed for a given SEO program. There is one way that you may not have considered before, so today we’re going to look at just that. First, a simple concept for you;

SERP Out – this is a a mindset for the search optimizer of getting into the habit of actually looking at the SERP you’re targeting. It may sound obvious, but given the many data sets and tools we have on the market today, many times SEOs forget to actually LOOK at the SERP itself.

And to get us started getting a SERP-out mindset, we’re going to look at query refinements (and Google).

What is a Refinement?

These are actually something very familiar to most of us. Traditionally we’ve seen them at the bottom of the page in the form of related searches that Google might suggest to you. More recently, we’ve seen them at the top with brand, location and other types of query suggestions.

This is done through query analysis and some form of recommendation engine. You know those right? When you’re at a site such as Amazon and it ‘suggests’ other products to you? Right. That’s the one. This is about the same thing that a search engine does when it makes a suggestion.

Query refinements

The query data is often culled from other users and refinements they’ve made in the past. If a lot of users that searched [coffee makers] ended up finishing the task with a query refinement to [coffee maker reviews] then it makes sense to add this for that particular query. There is also the potential for personalization, but it’s not likely as strongly in play in this area.

How query refinements can define your strategy

The first type we’ll look at is the traditional one, pictured above; query refinements. As noted already they are generally produced from looking at past searches done by others and subsequent refinements they may have made towards finishing the search task successfully.

  • [coffee maker reviews] – informational modifier
  • [commercial coffee makers] – transactional modifier
  • [cuisinart coffee makers] – navigational/brand modifier
  • [thermal coffee makers] – subcategory informational (possible transactional)
  • keurig coffee makers – navigational/brand modifier
  • pod coffee makers – subcategory informational (possible transactional)
  • vacuum coffee makers – subcategory informational (possible transactional)
  • bunn coffee makers – navigational/brand modifier

Now if we assume to some degree that these are being generated statistically, there is going to be a fair amount of value in also potentially targeting some of these query spaces as well. The interesting part is that not only are we seeing what users are searching for, but the query funnel of their search session towards completion.

Which ones, will obviously depend on the type of site you have. Are you running a straight up commercial site? Then following the path of the more transactional terms would be more fitting a strategy. If you have informational elements to your site, then we can look in that direction. The obvious strategic concept that immediately emerges; kill them all!

So, next let’s expand on one of them; [thermal coffee makers] – when we do that, we end up with a new collection of refinements to look at;

Query refinements

Although this time, the breakdown is somewhat noisier;

  • white thermal coffee makers
  • commercial thermal coffee makers
  • thermal cup
  • thermal coffee makers reviews
  • cuisinart coffee makers
  • krups coffee makers
  • automatic coffee makers
  • report coffee makers

The main thing though is that we can mine these refinements for new ideas and directions to take with the SEO program. Given the degradation I’d not go more than two queries deep, but you get the idea. There is the potential for some lateral thinking.

The new breed; Brands, Stores & Categorization

The next area we want to look at are the more recent query refinements that are appearing at the top of the SERPs, once deemed a ‘brand bias’ by grumpy SEOs. It’s not though, it’s ‘user friendly’.

more refinements

Now, once again, when this first came out is was just brands and a lot of SEOs wrote about how Google was giving preferential treatment to them. These are query refinements. Meaning that X percentage of those searching for [coffee makers] bounced back and refined their query to [Cuisinart coffee maker] or a related search. The search engine is offering up a ‘best guess’ to move the user along to their end goal faster.

So let’s have a look at Insights;

Google Insights

So, what do we make of it? Well obviously if you’re carrying these brands, you are going to want to drill down a refinement layer and look for opportunities (to rank on those queries as well). As with the lower refinements, we can look at these for hints at directions we may want to take.

The same goes for the store (entity) and the types (categorizations). Now, obviously you aren’t likely to be one of the stores listed, but that doesn’t mean you can’t click that and drill down into those SERPs as well. Look for opportunities. They aren’t always there, but sometimes they are.

What about personalization?

One of the problems that comes to mind is of course, personalization. Meaning; are there risks of noise when we use data such as this?

Of course there is. That’s why we also want to look at other data sets such as PPC, Insights, Trends and so forth. Any data being used in strategy development should always be cross referenced. But, what is interesting here, is that personalization isn’t the person-to-person application that many people believe. In truth, Google is actually grouping statistically and socially related user types. What does that mean?

It means that that this type of feature, query refinements, isn’t all that far off computationally. It is a grouping of data sets. Thus, even if there is some minor personalization at play, it’s still a relevant statistical data point.

Putting it all together

Once you’ve done all the due diligence and collected the refinements, establish how in-line it is with your current strategy and where you can implement the targeting. Some considerations that spring to mind are;

  • Run standard KW research / valuation processes.
  • Develop concept streams – informational/transactional etc..
  • Site architecture – play into the funnel
  • Internal Links – page mapping
  • Landing page strategy – adapt to logical path

What we’re doing with this exercise is looking at data beneath the SERP. Google has recently said, “…don’t chase algorithms, chase what we’re chasing (good user experience)”. Or something like that. Well hey, let’s do both shall we? By better understanding the algorithms we can look at the results and start to glean some insight into what users want. And be there.

And now for the punch line….

No, seriously, there is one… I just made up everything in this post. I haven’t been actively using this approach (so far). Yes, there is a ton of sense in it. But that’s not the point. This is what I see when I look at a search result. After many thousands of hours learning about search engines, there is more than meets the eye. That is what I wish most to impart to you my friend.

The goal (of this series) is that you start to really LOOK at a SERP and get a better feel for what’s there and what it might mean to your SEO campaign, (because it’s all situational right?). Data from tools is great, but it’s akin to driving along looking at the dashboard, not the beautiful nature around you.

Next time; SERP Out – Universal Strategy

Facebooktwittergoogle_plusmail

2 Comments

  1. Great insights, David. A lot of people don’t take query refinements into consideration so I’m glad you have pointed that out.

  2. Great insight Dave – isn’t it amazing how much we learn when we look at SERPs?

    I often use the ‘searches related to’ as a sign of intent, and select information or transactional keywords as appropriate.

    Good stuff.

Leave a Reply

Your email address will not be published. Required fields are marked *