Posted by
The In Search SEO Podcast
Don’t forget, you can follow the In Search SEO Podcast by subscribing on iTunes or by following the podcast on SoundCloud!
The In Search SEO Podcast Poll Question of the Week!
Let us know if you analyze keyword intent as part of your keyword research process! If you do, we’d **** to hear what you do to determine how Google views intent for a keyword. Let us know so that we can feature you on the next episode of In Search!
Summary of Episode 10: The In Search SEO Podcast
This week on the podcast our hosts discuss:
- Determining Google keyword intent: 3 easy steps to getting a solid intent understanding
- Google or users? Who’s responsible when it comes to confirmation bias?
Determining Google Keyword Intent [2:00 – 21:00]
Our hosts went deeper into Mordy’s latest blog, “Cracking the Code: Determining How Google Interprets Intent”. Understanding how Google interprets intent is a crucial SEO issue. How Google views query intent can either limit or create ranking opportunities for your site. As an example, the keyword “Airplane” can be interpreted as either a flying vehicle or the film, Airplane!. Currently, Google shows results for both the film and mode of transport. However, imagine if Google decided that users only want information on the film. What would happen to all of the results on the SERP that deal with the actual aircraft? Presumably, they would all fall off the face of the SERP! With that said, it’s important to know how Google interprets a keyword from the lens of user intent.
So how do you interpret intent? Mordy breaks it down into 3 steps:
Step 1: Run a Query & Categorize the Sites on the SERP
First, find a keyword that is representative of your whole campaign. Meaning, it represents your overall keyword set (i.e., it’s not an “outlier”). What you then want to do is run a search for that keyword (in incognito mode preferably) so that you can analyze and categorize the different types of sites Google shows on the SERP. Why? Well, because Google shows different types of sites to meet different intents. When you run this keyword, you’ll see them!
We experimented with the keyword ‘camping gear’. Here, Google shows nine commerce sites and one informational site (offering tips on camping gear). Quite evidently, Google does see the user wanting a bit of information about which gear to buy. That said, you could argue that from a purely intent perspective it’s not worth a commerce site’s while to create informative content targeting the keyword (though it’s a good idea for a slew of other reasons).
Having seen that Google does include a second intent for the keyword we need to know if that breakdown is pervasive. Does Google offer informational and commerce sites for similar keywords? Does Google perhaps show more informational pages for similar keywords? On to step two!
Step 2: Determine If the Intent Pattern Is Pervasive
Now that you see what the intent breakdown looks like for your keyword, you need to start looking at similar keywords in order to find a pattern. To keep to the above example, you might want to search for keywords like ‘best camping gear’, ‘buy camping gear’, ‘camping gear near me’, etc. As you analyze the SERPs for this keyword record the ratio between informational and commerce sites (or whatever site types appear on the SERP for the keywords you run).
You’re doing this because even though the keywords all relate to similar topics, different variations can produce different site type ratios. You might find that more complex products come with more informational sites on the SERP. For example, when Mordy ran his keyword set, software product SERPs contained many more informational sites than simple sort products. Here, in our case of ‘camping gear’, if you search ‘best camping gear’ you will find the opposite of what we saw for ‘camping gear’. For this keyword, we get nine
What you’re trying to do here in step two is to find a pattern of what intent looks like at the sub-category level. What keyword variations within the same keyword set produce what ratio of site types on the SERP.
Step 3: Isolating Subcategorical Intent
Knowing that Google treats sub-sets of keywords differently we arrive at step three. While we may know the ratio of site types for various keyword sub-sets we need to look at site type sub-categories. That is, it’s not enough to know the ratio of
You might find that ‘hiking boots’ and ‘camping gear’ both show a lot of informational sites, but ‘hiking boots’ produces more review sites whereas ‘camping gear’ more tip sites. So as you undergo step two, also breakdown the site sub-categories you see (you could also run your list of keywords, see that intent patterns you find, and then go back and analyze those SERPs with a finer comb)!
To recap, start with a keyword, see how Google parses intent, run more and similar keywords to see if there are any unique patterns or subcategories of keywords and branch out. Then, or concurrently, break down the various types of sites within a site category (i.e., sites offering tips as opposed to reviews, etc..
SEO News & Analysis [21:00 – 26:57]
Test Your Google Rich Results: Google recently announced that the Google Rich Results Test will now allow you to make edits to the code that you can run and test on the fly. SEOs are sure to find this very helpful. Thanks, Google!
Google Assistant Updates: A whole slew of updates for Google Assistant were announced at CES. Users can check-in to United Airlines flights using the Assistant; you can also book a hotel room with several hotel chains.
While some of these updates are great, like the live translator which is perfect for voice search, others like booking a hotel will require multiple steps and only begins on voice search which can be annoying. Why undertake a multi-step process where you might not even have access to what your hotel room will look like when you can just book online. This relates to what we saw last week where the data showed voice device users prefer voice search because it’s faster (as in a multi-step and multi-device process is not faster).
Google Manufacturing Center Gets New Features: Google recently announced that its Google Manufacturing Center feature will now offer deeper analytics, rich content and expand its availability to more countries.
Local Listing Match Score: Google recently tested a new feature that utilizes machine learning to show you a personalized restaurant match score inside the Local Pack. The score, for example, will say you have a 94% chance of liking a restaurant listed in the Local Pack.
While Jacqueline thinks this will be a great feature bringing value to users, Mordy is skeptical about how Google knows what he likes and doesn’t like. You may have browsed certain restaurants in the past, but it’s not a true representation of what you like. At the same time, and again in Mordy’s opinion, it also feels a little “unsettling” that Google knows you so well that they feel they can calculate what your taste preferences are. Mordy is also curious how will this affect rankings. If, for example, you search ‘best restaurant near me’, and there is a five-star restaurant one block from you, but according to your match score Google thinks it’s not a great fit, will it rank lower than an equally rated restaurant that might be further away but has a greater match score?
Search Console Features Being Discontinued: Google recently announced that they will be discontinuing some old search console features some of which include the crawl errors section.
A lot of SEOs aren’t happy with this one as it could make dealing with site migrations a bit harder. Also, these sort of changes
Content of the Week [26:57 – 33:22]
Our content of the week comes from Alexis Sanders who wrote a piece for Search Engine Land entitled 2019 in search: Find Your Seamlessness. According to Mordy, this is the single best piece of content he has ever read on Search Engine Land (which is saying a lot since he’s a big fan of SEL)!
Alexis’ article thoughtfully takes up a
Google Confirmation Bias: Alexis took a hard look at Google’s confirmation bias and subtly got into who may be responsible. Quoting the article, “This is exacerbated by the general public’s limited critical evaluation of the information they digest (leading to a garbage-in, garbage-out problem). Ultimately, the public has shifted from individuals being responsible for validating source, to search engines
The reason we chose this piece of content is that it was so thoughtful and dealt with some of the bigger themes related to search in 2019. So while Mordy may not have agreed with Alexis on the specific topic of confirmation bias, it is with the utmost respect! When it comes to confirmation bias Mordy believes the onus is on Google and he explains it via a ‘Socratic parable’:
Say you go to an open market to buy apples and the apple cart has Grade A, Grade B, and Grade C apples, with Grade A being the best and Grade C is the worst. In your quest for apples, you ask the apple cart owner to pick out some “good” apples for you, ones he/she knows you’ll enjoy. The apple vendor thus proceeds to give you Grade B apples knowing you can’t differentiate between Grades A and B. Was it wrong that the seller gave you the wrong apple? Vis-a-vis apples does it really matter? You don’t know the difference and are perfectly content with your middle of the line apples!
Now say you find out…. Would you be mad? Of course! The vendor lied by omission… no one likes not having all of the facts! Confirmation bias is not about search results according to Mordy. The commodity is not a set of results. Rather, the commodity is
The In-Search Podcast’s Fun SEO Send-Off [33:23 – 36:30]
Jacqueline once again takes a crack at the Fund SEO Send-Off this week with:
If Google were a high school student what “group/click” would they be in?
Jacqueline thought Google will be a mathlete or on the debate team, but at any time will drop out of high school because it knows everything already.
Mordy believes Google would have created its own high school but plays to every cliche by saying they would only do as such after spending a year in the AV club.
About The Author
