There are some changes coming to paid search at Google that sound exciting on the surface, but may leave many guessing how exactly those changes might manifest themselves. Over at the Inside Google Adwords blog, we were greeted with a blog post titled New matching behavior for phrase and exact match keywords on April 17th, that tells us that Google will be returning a few more results for paid advertisements that are phrase and exact match keywords. The post tells us to expect to see this start in mid-May.
While I don’t offer paid search as a service, I do often use the Google Keywords Suggestion Tool, and it left me wondering if the search volumes reported by that tool would change in response to the broader match in Google Adwords. Will it continue to show me only “exact” match volumes for keywords that I enter into the tool, or will it start reporting matches for keywords that are broader? Coincidentally, Google was granted a couple of patents this week involving search advertisements, including one on ways that the search engine might modify or expand the range of terms and phrases that advertisements may be shown for.
The first one that caught my eye was the following, which lists Ramananthan V. Guha as one of the inventors behind the patent. He’s known for a few things, including early work building the first version of RSS, as well as being a major force behind Google Custom Search Engines. He also developed Google’s version of trust rank, as an annotation system from “trusted sources” that could make search results more relevant for certain terms and phrases.
What we see in this patent is that Google might conduct search log file analysis to understand when certain queries might evidence a very close matching intent, or a matching “dominant intent. For example, on a query for “train cake pan,” a look through query logs might find a number of terms that may tend to show up during the same query session, and tend to do so frequently. Some commonly associated terms might include “train cake pans” “train pan,” cake pans,” “baking pans,” and “decorative cakes.”
An analysis of how closely related the intents behind those queries might be could determine whether or not advertisements triggered by those other query terms might appear for the initial query of “train cake pan.” The patent suggests that some of this analysis could involve looking at the landing page as well, and determining how relevant it might be for some of those terms.
Given the timing of the Adwords blog announcement and the granting of this patent yesterday, and the fact that both point towards a broader matching of advertisement terms where the intent seems to be close matches. In fact, here’s a snippet from the blog post that echoes that:
We know users are happier when they get search results that reflect their intent and help them achieve their desired action, even if it’s not a precise match for what they’ve typed. So we’re extending this behavior to ads.
The patent may not the impetus behind this change, or it might only be so in part, but if you’re interested in how Google might find ways to broaden advertising results for those queries, this patent holds a number of possibilities:
Query identification and association
Invented by Ramananthan V. Guha, Shivakumar Venkataraman, Vineet Gupta, Gokay Baris Gultekin, Pradnya Karbhari, and Abhinav Jalan Assigned to Google US Patent 8,171,021 Grnted May 1, 2012 Filed June 17, 2009
Abstract
Apparatus, systems and methods for predictive query identification for advertisements are disclosed. Candidate query are identified from queries stored in a query log. Relevancy scores for a plurality of web documents are generated, each relevancy score associated with a corresponding web document and being a measure of the relevance of the candidate query to the web document.
A web document having an associated relevancy score that exceeds a relevancy threshold is selected. The selected web document is associated with the candidate query.
This second patent describes how certain ads might be seen to be more relevant to the pages they point to if the page itself is considered relevant for the keywords being used in the advertisement. Does this mean that better SEO for landing pages will be more important for paid search?
One thing that struck me right away with this patent is that the three named inventors are amongst the most well known search engineers at Google, with Georges Harik as a major force behind Adwords, Adsense, and advertisemens in GMail.
Methods and apparatus for serving relevant advertisements Invented by Jeffrey A. Dean, Georges R. Harik, and Paul Bucheit Assigned to Google US Patent 8,171,034 Granted May 1, 2012 Filed March 22, 2010
Abstract
The relevance of advertisements to a user’s interests is improved. In one implementation, the content of a web page is analyzed to determine a list of one or more topics associated with that web page.
An advertisement is considered to be relevant to that web page if it is associated with keywords belonging to the list of one or more topics. One or more of these relevant advertisements may be provided for rendering in conjunction with the web page or related web pages.
Takeaways
I should probably look more closely at search patents involving paid search more frequently.
I’m concerned that the new matching behavior of queries in advertisements might influence Google to change the search volume that it reports for those queries in the Google Keyword Suggestion Tool.
The two patents both tell us that Google might be paying more attention to the content found on landing pages pointed at by advertisements in the future. That could play a role in the display of those ads, and when they might appear for which queries.
Google recently updated their page on how they calculate quality scores for advertisements, on April 24th, and it now provides a greater level of transparency regarding scores for things like “Expected click-through rate”, Ad Relevance” and “Landing page Experience”.
Under these recently granted patents, it’s possible that an ad might be triggered to appear in search results for a term closely matched (in searcher intent) to terms chosen by advertisers might alo be influenced by how well of a match a landing page might be to those other terms.
Is Google following the processes described in these ads, or something close to them? That’s probably worth exploring.
If so, it might be worth doing a little more analysis of the on-page SEO of those landing pages, and the terms and topics they could or should be optimized for.