Search engine optimization, or SEO, has changed significantly in the past few years. Thanks to the power of machine learning and artificial intelligence, the only way to build a sustainable, long-term SEO strategy is to create content people want to share. To combat these new trends, we need to employ our own machine learning technology to find what works and perform modern SEO at scale.
In this post, we’ll look at step 7: construction.
AI-Powered Content Construction
While we are not at the point where artificial intelligence creates coherent, well-written, contextually sensible content, we can use it to help understand what great content is. Based on our scans of our own content and the content of those top-performing pages, what might we learn to augment and improve our content?
Sentiment and Tone
How does the best content in our field sound, from a sentiment and tone perspective? Using natural language recognition like IBM Watson Personality Insights, we build an understanding of the tone and voice of top content. Is it authoritative in tone? Analytical? Confident? Emotional?
Knowing how top-performing content sounds helps us improve our content by sounding like our audience expects an authority in our space to sound – and every industry/niche is different.
Parts of Speech
What specific words do top-performing content pieces use most? One of the most interesting takeaways from Tom Webster’s Brand Election speech was that during the 2016 elections, data science teams identified how audiences in each location spoke, what specific words they used to describe their problems. Free, open-source speech taggers like Spacy for Python and R let us have access to the same capabilities for free today.
Using a parts of speech tagger, we unpack language to find the specific words and phrases our audiences use. Once we understand the specifics, we tune and upgrade our content to match. We might also use more than one data source; instead of just top-ranking articles, what could we glean from conversations about our chosen topic?
If top-ranking content isn’t textual, how might we understand it? With services like Microsoft Cognitive and Amazon Rekognition, we process image and video data at scale, then use speech-to-text transcription to turn rich context into digestible text.
With this data, we can now understand even photos, video, or audio that ranks highly, then use this insight to guide our content creation or modification.
If An AI Falls In the Forest
We’ve spent a great deal of time preparing our content for search purposes. However, the best content isn’t just high authority, it’s also high-ranking. High search rank comes from external factors such as inbound links. In the next post in this series, we’ll look at some ways of identifying the key opinion leaders we need to earn links from. Stay tuned!
The AI-Powered SEO Process
- Step 1: Inventory – understand what we have and how it performs.
- Step 2: Enrich – building off what we have, see what else is out there
- Step 3: Validate and Select – check that we’re working towards high-volume topics, make choices from the topic buffet of what to pursue
- Step 4: Extract – determine what else makes the top performing pages work so well from a content and topic perspective
- Step 5: Predict – understand the seasonality of our chosen topic
- Step 6: Construct – improve or create new content for our chosen topic
- Step 7: Associate – explore Key Opinion Leaders (KOL) who are influential in our topic’s space
- Step 8: Distill & Reach Out – understand what sites our KOLs share and their search authority, pitch our new content to our KOLs and their media sites of choice
- Step 9: Wrap Up/Conclusion – review where we’ve been
You might also enjoy:
- What Is The Difference Between Analysis and Insight?
- How To Set Your Consulting Billing Rates and Fees
- Marketing Data Science: Introduction to Data Blending
- How I Think About NFTs
- B2B Email Marketers: Stop Blocking Personal Emails
Want to read more like this from Christopher Penn? Get updates here:
Get your copy of AI For Marketers