Reality Check: How AI & Machine Learning Change CMS
While driverless vehicles and personal assistants like Alexa are the innovations that steal the spotlight and get the most public attention, there isn’t a day that goes by without news about how artificial intelligence will transform every aspect of our daily lives. Today, you can order a pizza through Echo while connecting it to your lights and thermostat, but the technology is still limited to specific users and use cases. By that, I don’t mean to diminish its value though. After all, who am I not to agree after Andy Rubin, the Google veteran who built Android into the world’s largest mobile operating system, is convinced that artificial intelligence is the next big thing? However, from the business perspective, in reality the advancement of technology gains an actual meaning when it applies to business use cases to create business value.
Therefore, in this article, instead of frothing up the bubble on the hype, we will discuss how to strategically incorporate machine learning and artificial intelligence into our current business processes with a focus on maintaining a future-proof strategy.
Difference Between Machine Learning and Artificial Intelligence
First, I would like to start my reality check with the distinction between Artificial Intelligence and Machine Learning as it really matters. While AI and machine learning are very much related, they are not quite the same thing. Machine learning is the practice of using algorithms that make predictions with some accuracy from data, where those predictions get more sophisticated over time without additional programming as the machine is “trained” using large amounts of data and algorithms to learn highly complex and nonlinear relationships.
Artificial Intelligence, on the other hand, is a broader term used to describe all of the process to get a machine to possess the same characteristics of human intelligence. In other words, machine learning is a subset of or almost a prerequisite to Artificial Intelligence. Basically, the concept of AI includes intelligent machines that can learn for themselves when they are given access to data. I know it is super cliché but to put this in context, it is like from The Terminator; Skynet, a neural network-based conscious group mind and artificial general intelligence system, obtains self-awareness after spreading to millions of computers and attempts a nuclear genocide. The reason why we still keep using these clichés as an example is that AI has not yet fully arrived, but considering the progress made in machine learning represents a great potential of the advent of artificial intelligence and an exciting future.
You can also think of Machine Learning as the vehicle that is driving AI development forward. Therefore, the speed of development of this technology has a direct impact on the AI development. As I stated earlier, the machine learning technology still is limited to specific users and use cases because still, when we are communicating with Siri or Alexa we have to learn how to talk to the machine instead of the machine learning to understand what we are saying as making them work still requires using a certain type of language.
Some potential benefits of the advancements in machine learning and artificial intelligence to content management systems have already been adopted by vendors and users.
One of those areas is the content curation process as machine learning provides so many opportunities for understanding the performance of content. With today’s technology, content creators can analyze how content has come across to readers, which tones resonate for which audience, and so forth. As a content creator and marketer, being able to perform tone analyzing, for instance, is one of the best features powered by machine learning. For instance, the IBM Watson Tone Analyzer service detects emotional, social, and language tones in written text. Some use cases are:
- Social listening and audience monitoring
- Personalized marketing
- Customer engagement monitoring and quality assurance
If you want to see the service in action, here is a quick demo.
To demonstrate business value, let me give you another example. Powered by cognitive computing technologies, Persado’s platform provides an in-depth understanding of emotions within languages and a customer’s response to them within a marketing scenario. The vendor claims that more than a hundred global brands use their platform to optimize marketing campaigns, resulting in a 49.5% average lift in conversions, a 68.4% improvement in click-through rates, and $1 billion in incremental revenues.
Given the mass amount of data at our disposal today, the capability of searching for specific information or files alone will no longer satisfy, especially an increasingly millennial-dominated workforce. When machine learning capabilities are injected into search, it helps users identify patterns and successful outcomes to retrieve better search results from all types of content. Today, enterprise search providers are going beyond conventional metadata searches by making audio, video and graphical imagery files searchable in their platforms and offering nontextual query types, such as the use of an example image or video segment.
While the rising magnitude of content is bringing so many opportunities, especially for marketers, sorting out the information needed is getting overwhelming for both internal and external stakeholders. Getting results that look like a page of links is no longer considered as a service as we all are looking for answers to a question when we use any search capability. Therefore, even the leading analyst firms have changed what they call enterprise search, illustrating that organizations can retrieve much more information through search interactions.
For instance, Gartner now uses the term “Insight Engines” or “Proactive Search” while Forrester calls it “Cognitive Search.” The scope of enterprise search has changed.
Today, services with 32 percent and manufacturing with 30 percent are the largest adopters but the adoption of the enterprise search technology powered by machine learning and artificial intelligence is on the rise. Until today, enterprises have only been really scratching the surface in terms of search’s vast capabilities across all business experiences. In conjunction with the escalating adoption of artificial intelligence and the rising magnitude of content, the AI-powered search is transforming for the better. Tomorrow, businesses that can deliver existing or synthesized information proactively or interactively in the context of employees and customers at timely business moments will be the winners.
Even though analytics have always been the best guide for measuring the performance of content marketing related activities, until recently, the basic analytics have left marketers vulnerable when it comes to measuring the contributions of upper- and mid-funnel marketing and advertising efforts. Even though we were able to generate a number of touchpoints with customers throughout the customer journey, for instance, we were limited to evaluating the contribution of disparate advertising efforts. With advanced tools to support the data-driven attribution model which utilizes machine learning algorithms, we are able to export the analysis of an account's unique conversion patterns, comparing the paths of customers who convert to those who don’t. This capability alone represents a great business value as businesses now have a better insight on, say, every advertising dollar’s effectiveness across different channels and devices.
Since content intelligence platforms integrate semantic technology and information science, businesses also can understand what topics are performing well on their rivals’ sites as the platforms can demonstrate the common topic, differentiators, and topic gaps which gives them a broader view of their positioning against the competitors. Therefore, they can be armed with predictive insights to uncover hidden opportunities and competitive threats instead of trying to battle with them through a biased decision-making process. At the end of the day, the technology provides businesses with a piece of mind by taking the guesswork out of what content to produce and which platform to be on.
Are you familiar with statements like “Frequently bought together” or “People who bought this also bought”? I thought so! It is not a coincidence that Amazon has become a multi-million dollar behemoth in the e-commerce space. Alongside many factors, its integrated recommendation engine is one of the primary reasons behind Amazon’s supremacy. 35 percent of what consumers purchase on Amazon comes from product recommendations based on such algorithms because shopping experiences go beyond a rational analysis of available alternatives, yet they are guided and emotional activities. As Steve Jobs once said: “A lot of times, people don’t know what they want until you show it to them.”
Not only e-commerce but also the content management system space has been taking advantage of this technology. 75 percent of what you watch on Netflix, for instance, is also retrieved from its recommendation engine because there is also a psychological element to it and it comes to consumers in the form of a statement like “People who viewed this also viewed”. The motivation behind taking action is based on this kind of statement is the fear of missing out which is, scientifically, a form of social anxiety and also known as FOMO in popular culture. The study conducted by ShareThis and Paley Centre for Media suggests that content sharing can influence consumers more than price and brand, and motivate people to spend 9.5% more.
In May 2016, Amazon opened up its sophisticated AI technology as a cloud platform by unveiling DSSTNE, an open source artificial intelligence framework that Amazon developed to power its own product recommendation system, in an attempt to extend the promise of deep learning beyond speech and object recognition to other areas such as search and recommendations.
Trends Seen on the Horizon
Understanding the Full Customer Journey
As I stated while explaining machine learning based analytics capabilities, even though we are able to generate a number of touchpoints with customers throughout the customer journey, we still need to fully understand the purchase cycle; from why shoppers have been dissuaded from completing a purchase to what event motivated the visitor to engage with the brand the most. Therefore, Google, for instance, revealed its plans to track down billions of purchase transactions through debit and credit cards, aiming to link offline purchases to online ads. It means that when a user has seen an ad, Google will be able to tell whether they have searched for the product advertised and gone to an offline shop to buy it. The business value is huge for merchandisers. Don’t believe me, then hear from Virgin Holidays as the brand claimed that when it factors in store sales, its search campaigns generate double the profit compared to looking at online KPIs alone. They have also found out that a customer purchasing in-store after clicking on a search ad is also three times more profitable than an online conversion.
Voice-Activated CMS Assistance
Last March, my co-worker Laura Myers, and I had been invited to attend Episerver Ascend 2017 as media, which took place in the glamor of the heart of Las Vegas, with over 600 attendees and more than 70 sessions. During the keynote presentation, the vendor disclosed some projects that they are working on with Microsoft. One of those projects was about giving the CMS platform voice comments to get some tasks done, including content creation, website design, content distribution and so on.
If I can order Tide detergent through Alexa, why wouldn’t I be able to get my content management system to type in this article, for instance, while I am telling the sentences out loud?
During the recent earnings call, Salesforce CEO Marc Benioff said that they are internally testing Einstein Guidance which provides recommendations based on data as a member of the management team. It is not certain yet if the company will roll out this feature but Benioff already seems very happy to have Einstein in his boardroom as he said: “That ability to consult with Einstein has made me a better CEO.” According to him, the most beneficial aspect of having this type of a technology is that the recommendations come without bias, unlike its human counterparts.
Communication between CMSs
We are already able to connect the Echo to our Hue wireless LED lights as well as our NEST learning thermostat. It would be cool to see this type of connection between CMSs when we are trying to tackle responsive user interface for a particular connected device as they work on the issue together, and I do not think we are so far away from that happening.
For a long time, more often than not, the claims and hype have outpaced the reality and practicality. While still there is a long way to go with artificial intelligence and a plenty of uses cases waiting to be unleashed with machine learning, since last year, the technology has made huge progress. From the business software standpoint, for instance, thanks to machine learning and artificial intelligence, now a CRM automatically pulls out the contact details from an emailing platform, an Uber could be ordered automatically based on a calendared flight time on Outlook while a Slack-powered chatbot automatically arranges conference calls.
As the Salesforce CMO said on the London leg of their World Tour: “Welcome to the age of artificial intelligence, the world will never be the same again and it’s going to happen really quickly.” It is almost impossible to cover all the current and future user cases that create business value in one article. Therefore, I would love to hear from you what business use cases your organizations are tapping into and what use cases will be unlocked through AI and ML in the near future. Share your thoughts in the comment section below.