What are Content Curation and Content Relevance?Before we dive into how companies are using AI right now to create and deliver relevant content, let’s get our definitions straight. Content relevance: Relevant content means getting the right content, to the right person, at the right time. It can correspond with SEO and search queries, or with engagements within a social network, or a user’s data graph. You want to make sure you are able to train an AI model to recognize what content is relevant for your customers so that you only serve content they find helpful or entertaining. Content curation: Curated content is content that has been selected for a specific purpose or person. Good content curation means the content achieves the engagement or usefulness goal you have assigned it, and is part of a bigger stream or journey. Content curation is an important method for content relevance where teams can add a layer of AI to personalize content at scale. By programmatically curating your content, you make each piece relevant to a specific customer at a specific time.
Real-World Examples of AI Powering Content CurationAI-powered content curation is already relatively common in the real world, and you have been consuming AI curated content without even realizing it. Great AI content curation can feel like science fiction comes to real life. AI-powered content curation is happening on Netflix, most social media platforms, Spotify, and in your favorite search engines. All of these companies use AI-powered algorithms to serve up the right content to the right users at the right time. While most keep their algorithms private, we do have a little insight into how they work. Netflix is the perfect case study as most people have seen it in action. Netflix uses a machine-powered algorithm to:
- Tailor movie and TV show recommendations
- Optimize streaming quality
- Personalize artwork and thumbnails
- Editing oversight in movie post-production
What You Need to Know to Create Relevant ContentWhile not every company can or needs to build a Netflix-worthy recommender system, there are a number of strategies you can adopt to create relevant content for your customers. There are some effective AI algorithms out there to help you get that content in front of the right eyeballs at the right time, you just need the right training data to get them started. Here’s how you can ensure you’re making relevant content for your customers.
Continuously Update Content for Historical OptimizationHubSpot — one of the go-tos for content marketing knowledge — has recently started editing and reusing old content, which they have dubbed historical optimization. And, what they’ve found is that by optimizing and updating old content, they’re saving time and money over what it would have cost to create new content. And — it’s working for customers. Through historical optimization, HubSpot has doubled its number of monthly leads generated by those old posts and has seen increased organic search views by an average of 106 percent for those old posts. While new content is a great way to serve up relevant content to your readers, don’t let old content languish and die. With a little sprucing up, some new images, some editing, and updated statistics, the old piece of content becomes brand new.
Continuously Update Search AlgorithmsWhen you’re creating content to address your customer’s pain points, it’s critical to have a good understanding of how effective your search algorithms are and keep them up to date as behaviors change, new products get added or new content gets published. Optimizing your search algorithm on a regular basis is a great way to make sure it works and customers can find your content when it’s relevant. It’s good practice to add more training data to prevent model drift and keep your AI delivering high quality results.
Understand the Customer JourneyThe customer journey is the route your customers follow from discovering your product to buying your product. To have the most relevant and curated content, it’s critical to have content that serves the needs of all your customers, no matter where they are on the customer journey. HubSpot has a fantastic explainer on the different types of content you should create for the different stages of the customer’s journey. This is also a great place to apply AI, when your customer base is in the millions or even billions of people.
Create Content For Specific CustomersTo have the most relevant content, you not only need to understand the customer journey but also recognize that you have different customers. You need to create content that will be relevant and useful to different groups of your customers, instead of just making general content. Many companies create customer personas or profiles of imaginary customers based on customer data to inform their content curation. Different types of customers need different types of content.
Use Customized and Personalized MessagingCustomers no longer want just an experience, they want a customized experience that’s designed just for them. Customers want to feel a sense of connection with brands and the best way to build that relationship is through personalized messaging. By the way, what we’re talking about here is way more than personalized email greetings, but that’s not a bad place to start. Personalized experiences and messaging is where AI is really useful to organizations delivering content at scale. AI allows them to automate timing, type of content, and to align marketing strategy with the type of customer. Through data analysis and machine learning, AI can get the right content, to the right customer, at the right time to create a customized user experience.
Use Technology to Get Content to the Right Customer On TimeAI takes content relevancy to the next level, getting the right content to the customer just when they need it. Some AI tools that might be helpful for your content relevance products include:
- AWS Machine Learning services lets you create ML models without having to dabble with the algorithm
- Microsoft Azure’s Machine Learning is suitable for both beginner users and pros and businesses of all sizes, with over 100 methods that help with regression, classification, recommendation, text analysis, and anomaly detection
- You can also upload your datasets, train models, and deploy them with Google Cloud AutoML’s graphical interface
- Advanced users can leverage NVIDIA’s Transfer Learning Toolkit, a python-based AI training toolkit that allows developers to train faster and accurate neural networks on the popular deep learning architectures
- IBM Watson brings a fully automated ML service to the table, which reduces the learning curve and takes no prior training to use. This makes the ML Studio a good fit for beginners and experienced individuals for building, training, and deploying models.
- Appen, which can help you kickstart and improve content relevance algorithms through high-quality training data.