Using artificial intelligence to augment content creation efforts isn’t a road to uniformity. In this post, we take a look at an actual example of how AI can be used to enhance production while retaining creativity.
Content Strategy: Finding the Right Topic for a Thought Leader
People that follow their industry closely aren’t often at a loss for topics. The challenge is to create content that will reinforce their recognition as a subject matter expert.
Sure, you may get lucky and write an article that becomes a viral sensation among your peers. But for most people, leadership marketing requires a consistent effort. You’ll most likely need to publish many blog posts before you obtain that level of recognition.
However, you need not toil in obscurity. With a carefully devised strategy, your thought leadership efforts can significantly contribute to another important goal; recognition by Google as an authority on your subject.
Expertise, authority, and trustworthiness (EAT) are discussed extensively in Google’s Search Quality Evaluator Guidelines (pdf). So, it’s safe to say anything you do to improve that perception is helpful.
No, you don’t need to knock out a viral post that gets numerous accolades from your target audience. But you do need to strategically create a blog populated with clusters of content addressing your topic from every possible angle.
While this strategy won’t appeal to your ego, it will ultimately help your pocketbook through an increase in quality-targeted organic traffic.
Have you ever found a thought-leader, someone with a well-established personal brand, that has a poor-quality blog? It doesn’t happen.
Usually, their site has sufficient authority that it can rank with minimal effort for whatever blog posts they choose to publish. That’s assuming of course that they produce content related to their focus topic.
Aim to build your blog into an indisputable resource at the same time you pursue a thought leadership strategy, and you’re practically guaranteed success. Now that we’ve established a viable strategy for finding the right topic let’s look at a specific example.
MarketMuse Suite is an AI Content Analysis and Optimization Platform that helps optimize existing pages and craft new content that outranks the competition. Our blog exists to reinforce that goal. Its objective is to be the ultimate resource in supporting content strategists and marketers to create content at scale.
Having a focused purpose and a clear understanding of our audience simplifies the decision-making process. Every blog post must help our audience achieve their ambition of producing high-quality content that ranks, at scale.
In this context, the topic “scaling content production” is a good fit for a blog post. Let’s look at the process I used to create the post “Scaling Content Production With Technology” and get to the fourth spot on Google within 45 days.
Deep Learning: Build a Topic Model Using Subject Matter Experts
Regardless of how much I think I may know about a subject, I like to start by first developing a topic model. I use MarketMuse Suite because it’s wonderful at analyzing, categorizing and organizing textual data from the web. It examines thousands of pieces of content to create a robust topic model.
Here’s part of the topic model for the post. The actual model is comprised of 50 semantically related topics plus hundreds of variants. The topics are arranged in the list by order of importance, with most significant appearing near the top.
Each topic typically has a number of possible variants. Instead of using the same word repeatedly in a text, these variations can be used to add variety and interest. Sometimes, I find that variants can be interesting topics in their own right.
As part of that process, I look to see how the competition has used these topics. Remember, the model is generated from data found in the search results. So it makes sense to see how the competition has approached the subject.
Content Creation With the Help of Artificial Intelligence
Finding the Narrative Within the Topic List
Behind every topic model, there’s a narrative. The goal is to determine the story behind that list of semantically related topics. Every related topic in the list includes a suggested distribution, which indicates how frequently a term should be used.
For example, the word “team” is frequently used (10+ times) in articles discussing the scaling of content production. But the phrase “content team,” although important, is infrequently used (1-2 times).
What’s the benefit?
It helps me make an informed decision as to how I’m going to use certain terminology when discussing a subject.
Examining the Competition
With my research complete, I have a general idea of how to approach this topic. Since search traffic is an important factor, I want to see how the top-ranking pages tackle the subject. The ultimate goal is to create the most comprehensive piece of content possible. That’s irrespective of whether it’s considered “thought leadership” or not.
This is where I turn to the Compete app in MarketMuse Suite. The heatmap display provides an information-rich visual representation of how the top 20 ranking articles use the related topics identified in the topic model.
Each column represents a URL (from 1-20) listed across the top. Every row represents a related topic. Squares are color-coded based on how frequently the topic is mentioned in the article (red for 0 mentions, yellow for 1 – 2 mentions, green for 3 – 10 mentions, and blue for 10+).
This provides a good overview of how high-ranking articles approach the topic and more importantly, we know where the competition suffers from content gaps. With the information we have, there should be no problem writing a comprehensive, in-depth article about our topic.
You’ll notice that the topic ‘product content’ is not mentioned anywhere in the top 20 results. MarketMuse identified this as a semantically important topic related to the subject of scaling content production. One option could be to make product content a substantial part of my approach. This could potentially set my content apart from the rest.
Although I could have gone this route, I decided to kick things up a notch. Since this is about thought leadership, let’s take it one step further and go where no one has gone before.
Addressing the Content Gap(s) Everyone Else Has Missed
Keep in mind that the topic model is derived from an analysis of existing pages of all the competitive content on the web. It’s not an abstract concept but is firmly rooted in reality. If a related topic is nowhere to be found, it can’t appear in the topic model.
It was time to think outside of the box. So, I reviewed the top-ranking pages to get a deeper understanding of their approach. As a result, I noticed a content gap that everyone else had missed. All the articles approached scaling content production by either hiring freelancers, expanding an in-house team or using an agency.
I noticed that nobody addressed the role of using technology in this context. 🤔 That wasn’t in the topic model because it couldn’t be. Taking a creative approach, I was able to produce an in-depth article that added a new perspective to the conversation.
AI tools, although sophisticated, are still nothing more than tools. Keep in mind, just because you have a hammer doesn’t mean everything is a nail. Most of use topic models to ensure we properly cover a subject. Alternatively, they can be used to verify whether your idea is truly unique. If it’s not in the topic model, there’s a good chance your approach to the subject is outside the norm.24 28 False
Written by Stephen Jeske