Content Marketing in 2020 and Beyond
The biggest challenge content marketers are facing is dealing with the vast amount of data involved in all the stages I just described. There’s far too much information to process this manually. Those days are over and content marketers that fail to realize it will fall by the wayside.
Hi Jeff! Can you tell us a little about yourself?
I am the Co-founder and Chief Product Officer for MarketMuse (opens in a new tab), and I frequently speak at conferences like Content Marketing World, ContentTECH, and MAICON.
I am also an active member of the Content Strategy Collective Slack Community (opens in a new tab) and the Rhodium Community. I host a twice-monthly Content Strategy Webinar Series with MarketMuse and have an open calendar for anyone to book time for anything at any time (accessible from my CSC community profile).
I strive to be an empathetic marketer that is at the center of the Venn Diagram that consists of Editorial, SEO, and Conversion Rate Optimization. I want to be the force that brings those people together. I live on an island in GA, brew beer, and have an amazing wife and two boys under 3.
Can you explain in a few sentences what MarketMuse does?
MarketMuse is an AI content intelligence and strategy platform that helps you research, plan, and craft predictably better content faster. It analyzes your site’s content and performance to determine the topics for which you are most likely to succeed. Turn those topics into high-performing articles using our AI-driven content briefs.
They provide everything a writer needs to quickly create expert-level content including questions and subject matter to address, links, word count, and KPIs to meet in their writing.
Why do you think most content out there is shit?
Where do I start?
Most people are focused on creating keyword-based content without understanding the intent behind the search term. If you don’t get that right at the start, there’s nothing you can do to overcome that error.
Even if they do get that right, expert-level content demands proper topical coverage, something most content creators frequently fail to execute. The primary reason for this is that few content teams take advantage of content briefs.
What’s the secret sauce to being successful at Content Marketing?
To perform well at the highest level requires precise execution of the following stages: research, planning, briefing, writing, editing, publishing, optimization, and reporting.
What’s the biggest challenge content marketers are facing in 2020?
The biggest challenge by far is dealing with the vast amount of data involved in all the stages I just described. There’s far too much information to process this manually. Those days are over and content marketers that fail to realize it will fall by the wayside.
How good is natural language generation these days?
It depends on what model you’re using to generate content. We recently published a comparison of popular natural language generation (NLG) models (opens in a new tab), GPT-2, GROVER, MarketMuse First Draft, and XLNet using real-world metrics that a content marketer would use to evaluate a submission.
It turns out that MarketMuse First Draft (our natural language offering) is the only model that can consistently produce content over 1,000 words in length. The other models struggle to produce content anywhere near that length. Coherence is another challenge with which these models have to deal, as is comprehensiveness. Sometimes the models use many words, yet say very little.
In my opinion, these models approach NLG the wrong way. They rely on a phrase or short piece of content as an input and use it to generate further content.
That’s like giving a human writing a keyword phrase and telling them to run with it. The results are going to be hit and miss. It’s no different when using natural language generation.
Would you be able to rank reasonably well if you were to just publish 100 AI-generated pages without editing them?
That’s wishful thinking. It’s been tried before, and while the effort may have received lots of publicity, the results have been dismal.
Our approach to natural language generation is that it augments the human content creation process. It doesn’t replace it. Artificial intelligence can perform certain tasks with extreme efficiency, yet there are other endeavors better suited to humans.
MarketMuse First Draft accelerates content production and develops optimized drafts to your standard. Still, there are a few items that can be added to enrich your content, such as imagery, pull quotes, video, additional links, and step-by-step instructions.
How do you see the future of Content Marketing evolve?
No doubt there will be a greater emphasis on artificial intelligence to help with all aspects of content marketing. Analyzing vast amounts of data is best approached from a machine perspective. That’s where they excel so it makes sense to take this perspective.
Last question: what’s your number one tip to level up in content marketing?
Improve your research capability. Better research leads to better outcomes. All the later stages of content marketing depend on the accuracy of this stage. Set yourself up for success by getting it right the first time.