Written by: Mitesh Bulsara
We are a witness to the new wave of the Information revolution. In this age, information is extracted out of unstructured data. And for the first time, the focus is not on data or information per se, but on how a consumer responds to this piece of information.
Given this premise, it’s obvious that our Content Management Systems have to evolve
to match the speed at which information is generated
to find the relevant consumer rather than wait for a consumer to stumble upon it
to support the gamut of devices in today’s digital ecosystem
to adapt to upheavals like economic crises or pandemics
In the space of digital publishing, the volume of content is humongous and the lifespan of a single piece of information is very very short. Therefore the need for a self-aware CMS that understands what to do with such information is imperative.
A new-age CMS is a smart CMS that creates a phenomenal digital experience for content creators, managers, as well as consumers. The table stakes for a digital publisher is publishing content with speed and agility that engages the specific audience it was meant for before the fizz goes flat. Adopting the latest tech advancements in areas of big data, NLP, AI becomes a huge enabler saving significant costs as you go.
We will be talking about the early 2010s if we talk about CMSs that helped you to generate and manage diverse content like text, images, videos etc. We will be talking about the late 2010s if that CMS started helping you in SEO and grammar correction.
This is the era of Web 3.0. AI/ML is pervading every aspect in which we communicate with technology. People are already used to incredibly contextual offerings, whether it is a product or a service.
In the 2020s a CMS shouldn’t just allow you to generate content but make it way easier, so much so, that it should be able to create some valid, relevant content by itself. A smart CMS should let the writer focus on writing great content and do the rest of the work for them to make the article engaging and searchable.
It shouldn't just be helping in SEO, but should also help increase virality. It shouldn't just let you share ideas that you find exciting but also let you know the ideas which your audience would find exciting.
Let’s delve into the details of what potential a smart CMS holds specifically for a publishing company and its customers.
We know about the features of a CMS that lets content creators do their job. It should be intuitive, customizable, version control, search optimization, security, customer support to name a few. However, the features that make a Content Management System doesn’t just get the work done, rather, it awes the content creators.
Smart tagging - tagging your content for more than just SEO:
Natural language Processing (NLP) and image recognition capabilities of AI strengthen metadata tagging for content making it easily searchable over the web in the right contexts. Content creators are freed from their limitation of time and effort when the CMS automatically suggests a label for an image. The NLP engine could be analyzing a piece of content as it is written and suggest an image or a video clip from the repository to go with it.
Ambiguous words in content should not deter apt automatic tagging. Smart tagging allows the same tag to be used across the platform for all content related to, say, “vegan” or “non-dairy” or “plant-based”. This disambiguation results in better search result coverage.
Tagging when done smartly improves SEO, engages the audience, and makes your analytics engine intelligent. And how can we leave behind advertising? A good tagging strategy assists your advertising strategy in turn by either showing relevant ads on a content page or maybe rendering no ads at all for content tagged as sensitive.
Automatic Recommendations - Prioritizing hyper-personalisation strategy:
Adopting a hyper-personalization strategy is a must in a B2C market to attract and retain loyal consumers. An AI-powered engine caters to this strategy at different levels -
Data - predictive analysis
Content offering - content recommendations tailored to individual user profile
UX - User experience that addresses individual users as a valuable customer
Together these form the elements of hyper-personalization. A good recommendations engine considers a lot of parameters like location, user behaviour and history, current affairs, social media trends, etc to come up with relevant content. Also, it constantly measures and keeps optimizing the strategies to keep them aligned with the highest priority of a digital publisher - increased user engagement.
Summarizer - TL;DR (too long;didn’t read) your content:
Content summarization has been a hot ground for research in Data Science applications. You may want to achieve an abstract summary, or maybe just extract the important parts of a document without changing the content. The use cases where content creators may appreciate an AI-powered summarizer are –
Generating newsletters for a series of articles
Taking quick stock of competitors’ views and activities
Internal SEO on archived documents
Social media sharing
These are a few direct use cases. As the research for summarizers evolves to accommodate for different genres and lengths of documents, many more use cases will be identified.
Virality score predictor: measuring and increasing content virality:
A publishing company would benefit a great deal if they could predict with fair accuracy about which article could go viral, in which demography, at what time etc. Knowing how short the lifespan of a news item is, publishers can take the strategic decisions based on this prediction on how and where an article must be broadcasted, what resources should be allocated etc.
AI-driven analytics can provide immense data to match a publisher’s content across different categories (sports, entertainment, news) to its readers. This mapping of readers (and their behaviour) to content category sets precedence to predict the virality of another piece of content. Add to this social media engagement around the subject and you have a score that can predict how many eyeballs an article might get.
This is not all. Apart from predicting virality score, such an engine could also make suggestions for increasing the virality score of an article. By connecting the dots between what your article is about and how it relates with other subjects which are gathering the interest of your audience, you gain the ability to concentrate the readers’ distributed attention at one point - your platform.
Sentiment Analysis - Feel the pulse of your audience:
Digital publishers use feedback in the form of comments, shares, social media activity to receive vital indications about the audience’s disposition towards the subject. The volume of such feedback, however, could be in millions. It is a huge failure on part of a modern-day CMS if it does not use technology to extract meaningful information out of this unstructured, often grammatically incorrect, and jargon loaded feedback.
There are a lot of insights that a CMS can offer by leveraging various NLP models, for example, the BERT model or GPT3. You can gather sentiment analysis data focussing on – polarity (positive, negative, neutral), emotions (happy, sad, hopeful, confusion), and Intention. Having this knowledge, the content creators can decide on their approach towards their content strategy. For example, setting a tone for an article to elicit desired emotion.
Dynamic paywall - Monetize your content intelligently:
There is a shift observed in how digital publishers are aiming to monetize their content moving from ad generated revenue to subscription-based paid content.
The shift is primarily driven by the change in consumer’s behaviour where they are willing to pay for content that is personalized to a very high degree in terms of location, social media trends, personal interest, and online activity. Netflix, Amazon Prime, and the likes are classic examples.
AI-assisted paywall strategy, whether it is a hard paywall, metered, “freemium”, or dynamic paywall, goes a long way in increasing user subscription and retaining loyal consumers. An AI engine performs accurate user profiling based on several parameters - first time, fly-by, or regular user, article consumption rate, the device used, most active time of the day etc. Based on this categorization a dynamic paywall strategy can be applied and kept on being optimized on aspects like -
number of articles to show before showing a paywall
determining what is the right price that maximizes the revenue from subscription
determine the pricing based on user segments, regions, news sections
determining the right offers and promotions; who to show it to and when
A strong AI engine can automate this extremely complex piece of work and maximize revenue without the publisher investing heavily into these activities.
AI is not just a technology buzzword. It is the current reality of how businesses are reaching out to their customers and customers seeking out products that make their life easier. A typical customer is already used to personalized services in almost every aspect of life. And the options are galore.
Many publishers are moving towards AI-based engines to increase engagement and thus revenue. To stay relevant in the competitive market they have to push the right content, to the right people at the right time. While re-evaluating your CMS ensure that it runs on an AI engine and provides easy integration for scalability. is one such headless SaaS AI Engine that easily integrates with your CMS through simple APIs.
All said and done we need to open our eyes to the fact that “customer” and not the “content”, is the king. With all the technology-driven content floating around, if user experience around accessing and interacting with your content is not up to mark, we lose the whole purpose. We are eager to cover this aspect in our future correspondence. Meanwhile, start loading your CMS with smartness.