The secret to attracting visits in organic search today isn't about matching keywords but matching the intent of your audiences. This helps you to develop a clearer understanding of what your readers want to know and demonstrating the relevance of your posts in organic search rankings.
In classical Information retrieval, what many of us might be familiar with from our university library terminals in accessing books, journals and other documents, information needs defined by a query would match several objects in a database with different degrees of relevancy. There were several models used to determine relevancy and, despite some disadvantages, it was a lot easier that navigating library reference cards (let alone maintaining them!).
The basis for early search engines was based on this and had the advantage of adding additional factors to enable a more robust weighting agent to determine relevancy, such as the number of incoming links a URL had.
While this worked well in the early days of the web, there was something fundamental different about the web and the systems that information retrieval were based on in the first place.
As it turns out, not all search queries are information needs. In surveys of billions of Alta Vista queries in the late 1990's and early 2000's, three different intentions we're noticed amongst search queries:
Have a look at Andrei Broder's A taxonomy of Web Search to dig a bit deeper into this (he works for Google now, by the way).
We might take this for granted but recognizing primary distinctions such as these are fundamental in understanding how organic search engines have evolved to provide a more relevant and personalized experience.
Whereas the classical way of doing SEO was mostly about sprinkling relevant (high volume) keywords strategically on your content (URL, title, meta descriptions, headers, etc), the way to win today is acknowledge the following:
In some ways, it's actually liberating to consider this. For one, it helps you dig deeper into understanding what people are really looking for. An easy example for doing this is in Google search itself.
Let's the term Dark Data which has a few thousand queries a month: what about it? For one thing, at the top of the results I get a definition in case I don't know what this emerging term is about. Google has assumed that me searching for this is an information request for a definition:
Then, right there at the bottom of the search results, Google is providing me with Searches related to dark data: These are based on what has been calculated as relevant.
Some of these are navigational or transactional, such as getting to an IBM page on the topic or perhaps downloading a paper on the topic from Gartner. Meanwhile, most are informational and provide insights into what questions people have about the topic. In fact, clicking on dark data gartner takes me to a new set of results with an even more sophisticated set of ideas based on People also ask:
So in addition to expansions for the keyword Dark Data, I'm presented with a group of correlating keywords relevant to what I started with: unstructured data, data rot, data lake and so on. The results above will vary user to user but you have a way to access Google's understanding of a topic that is based on past user behavior that can be useful in helping you write content that answers these.
Google's Keyword Planner and Trends are also good sources for digging into topics.
Fundamentally, reaching your audiences in 2017 means recognizing that every query carries with it a different intention. You'll have more success in optimizing your content to answer questions rather than to match keywords.
In classical Information retrieval, what many of us might be familiar with from our university library terminals in accessing books, journals and other documents, information needs defined by a query would match several objects in a database with different degrees of relevancy. There were several models used to determine relevancy and, despite some disadvantages, it was a lot easier that navigating library reference cards (let alone maintaining them!).
The basis for early search engines was based on this and had the advantage of adding additional factors to enable a more robust weighting agent to determine relevancy, such as the number of incoming links a URL had.
I Want My Query to Do More
While this worked well in the early days of the web, there was something fundamental different about the web and the systems that information retrieval were based on in the first place.
As it turns out, not all search queries are information needs. In surveys of billions of Alta Vista queries in the late 1990's and early 2000's, three different intentions we're noticed amongst search queries:
- The classic Information Need was apparent approximately 50% of the time. These include basics such as What is CRM and How tall is Tom Cruise.
- 30% of queries were transactional: Show me sites were I can do something, like look at a map, order dinner, watch something or download something
- The remaining 20% are navigational: Show me a site I think or know exists.
Have a look at Andrei Broder's A taxonomy of Web Search to dig a bit deeper into this (he works for Google now, by the way).
We might take this for granted but recognizing primary distinctions such as these are fundamental in understanding how organic search engines have evolved to provide a more relevant and personalized experience.
Queries Are Just Questions After All
Whereas the classical way of doing SEO was mostly about sprinkling relevant (high volume) keywords strategically on your content (URL, title, meta descriptions, headers, etc), the way to win today is acknowledge the following:
- Search queries aren't just a string of keywords but are a question, that are often inferred. That is, the search engine analyzes the query in context.
- Search engines are able to understand what the content is about and isn't dependent on keywords to achieve this.
In some ways, it's actually liberating to consider this. For one, it helps you dig deeper into understanding what people are really looking for. An easy example for doing this is in Google search itself.
Let's the term Dark Data which has a few thousand queries a month: what about it? For one thing, at the top of the results I get a definition in case I don't know what this emerging term is about. Google has assumed that me searching for this is an information request for a definition:
Then, right there at the bottom of the search results, Google is providing me with Searches related to dark data: These are based on what has been calculated as relevant.
Some of these are navigational or transactional, such as getting to an IBM page on the topic or perhaps downloading a paper on the topic from Gartner. Meanwhile, most are informational and provide insights into what questions people have about the topic. In fact, clicking on dark data gartner takes me to a new set of results with an even more sophisticated set of ideas based on People also ask:
So in addition to expansions for the keyword Dark Data, I'm presented with a group of correlating keywords relevant to what I started with: unstructured data, data rot, data lake and so on. The results above will vary user to user but you have a way to access Google's understanding of a topic that is based on past user behavior that can be useful in helping you write content that answers these.
Google's Keyword Planner and Trends are also good sources for digging into topics.
Fundamentally, reaching your audiences in 2017 means recognizing that every query carries with it a different intention. You'll have more success in optimizing your content to answer questions rather than to match keywords.
- SAP Managed Tags:
3 Comments
