TechBytes with Peggy Chen, Chief Marketing Officer, SDL
Personalized content provides more value to marketers through highly influential brand messaging and consistency achieved via reusable and scalable content. Multi-channel content that matches customer buying patterns and behavioral information help in establishing a real-time personalized connection with powerful web experience management. To understand how marketers can create a mix of marketing, commerce and product content in a programmatic manner for contextual consistency, we spoke to Peggy Chen, Chief Marketing Officer at SDL plc.
MTS: What inspired the SDL to launch SDL Tridion DX?
For the past decade, the focus of digital experience initiatives has been on sales and marketing. Web content management, personalization and e-commerce solutions, to name a few – all centered on delivering a multi-channel experience in support of customer acquisition.
In parallel, for many years, technical documentation teams have been building product information, which serves customers after the purchase. This content was typically published in print or PDF format and predominantly served a different phase of the customer journey; the post-sale period, enabling customers to use a product or to resolve issues.
However, recent research conducted by SDL paints quite a different need in terms of what customers want.
It turns out that in-depth product information is a critical enabler for sales that actually serves customers in the research phase of their online journey. Product information allows buyers to conduct in-depth product comparisons, arming them with the information they need when they finally decide to get in touch with your sales department or an in-store sales rep. When it remains locked up in the support section of a website, sometimes behind a login, the value of all this great content remains untapped.
When customers look for product information, 61% of respondents chose your company website as the key trusted location, a close second position to search engines at 64%. Call centers dropped in popularity, with only 8% of respondents indicating they would contact a help desk to get the product information they need.
The siloed nature of companies causes an artificial divide between marketing content and product information. Customers, however, want to access all this information in one place. A whopping 94% want a single source of truth. The desire to search for sales, marketing and technical content within one cohesive online experience has never been greater, which is exactly what SDL Tridion DX addresses.
SDL Tridion DX is built to create continuous experiences across the pre-sale, sale and post-sale phases of the customer journey. It is designed to manage the global content lifecycle of both marketing content and in-depth structured DITA product information, while also unlocking content stored in third-party eCommerce, DAM and PIM systems.
This unified approach eliminates the disconnect that characterizes digital experiences today when visitors jump between sites to engage with marketing content, trying to buy products online while searching for support and product information.
In summary, SDL Tridion DX helps businesses create, optimize, personalize and push content across the entire digital landscape at any stage if the customer journey.
MTS: How does an API-first approach allow SDL to serve content better?
Today’s businesses need to reach customers across more and more channels and touchpoints. In the past, a company website was the key online delivery channel. Today, companies also need to serve up information to mobile apps, in-store kiosks, games and a variety of internet-enabled devices, like in-car entertainment/support systems. A well-known US-based manufacturer of electric vehicles, for instance, pushes all car-specific instructions and help information directly to each car using OTA (over the air) technology using SDL Tridion DX technology.
To do this, all the necessary content – that mix of marketing, commerce and product content – needs to be accessible to the content consuming “endpoint” in a programmatic way. This requires an API-first approach – also referred to as headless CMS – that makes sure content can be consumed while also supporting information like metadata, security/access levels for content, context (e.g., what device is requesting the content) and other relevant data to support a personal and fully optimized digital experience.
SDL Tridion DX’s architecture allows it to function as traditional content management system for companies that run myriads of websites, using enterprise capabilities like workflow, security, WYSIWYG authoring and localization management. It is also a core system for organizations to author DITA-based technical documentation for tens, hundreds or thousands of products, enabling simultaneous shipping of products worldwide.
In both these use cases, SDL Tridion DX ships with Digital Experience Accelerators for different use cases including Marketing Websites, E-Commerce Sites and Technical Documentation Sites, to get the system up and running in a matter of days or weeks, not months, supporting all of the required sophistication to manage, translate and distribute multilingual content on a global scale.
But over and above these traditional use cases, it serves as a touchpoint-agnostic “single source of truth” for any digital channel, so organizations can blend information that spans the entire customer journey using SDL Tridion DX’s API-first approach.
MTS: How do you see AI/ML technologies impacting the growth of content management systems by 2020?
More content is being created than ever before. Yet, the reality is that for enterprises to deliver truly personalized experiences to their end customers, even more content needs to be created, managed and intelligently delivered to the customers.
Content creation is a bottleneck today for most organizations. Companies suffer from the “multiplier” effect. A single piece of content needs to be served across touchpoints, requiring multiple variations. On top of that, each piece of content needs to be delivered in multiple languages, multiplying the number of pieces of content by the number of languages. And the need for personalized content adds yet another N personalized variations of each piece of information.
AI/ML technologies can help out solve this exponential content demand in three key ways:
- By creating derivative content, based on source content. Using the language models developed for machine translation, AI can create derivative content such as summaries, keywords and alternative versions in a faction of a second. This is perfect for publishing a core message out across web, print, social and other channels, each of which has specific needs, tone of voice and writing guidelines.
- By creating brand new content. Especially with structured information at the basis, computers can generate content using Natural Language Generation technology, directly in the target language. Great examples are weather forecasts and financial reports, but also website descriptions for hotel rooms, retail products and other items that are underpinned by a set of structured data points/attributes are viable candidates.
- By translating content. Machine translation is rapidly improving thanks to recent developments in neural machine translation technologies. The ability to translate content at scale at acceptable costs is now within reach for many more types of information and a wider variety of business applications.
Achieving true personalization requires a lot more content, a lot more data, the ability to analyze it, and the means to combine it. Could humans figure it out? Conceptually – yes. But with marketing teams under more pressure than ever before to deliver more with lower budgets, AI and ML can help.
Traditional personalization relies on humans creating personalization rules that tell the delivery systems to choose a specific piece of content based on specified criteria. For example, if a visitor is from area X, and visits the site using device Y at time of day Z, show a video of topic T. While these simple rules may suffice for basic use cases, for more advanced personalization scenarios the number of rules explodes.
This is where adaptive personalization has a bright future. It relies on self-learning algorithms that take into account multiple data points including current behavior, past behavior (e.g., purchases), contextual data points (e.g., browsing device, time of day), and any other variables. The algorithms determine what the best next piece of information is, in real time. Adaptive personalization looks at the current visitor and also compares it with people with similar characteristics, behavior, trending content and related information to predict the best possible content to present.
Clearly, there is a bright future for AI/ML within content management systems as it will impact the future role and effectiveness of content management systems. Companies cannot ignore the added value that AI/ML brings to reduce the content authoring, translation and delivery workload while improving content effectiveness.
SDL is in a privileged position in its portfolio to play a pivotal role in this space, based on its Natural Language Processing track record and with the enterprise-class SDL Tridion DX offering.
MTS: How should CMOs plan their investments into AI-for-Content to improve their analytic on omni-channel customer journeys?
The digital experience is driven by content – it is the currency that creates and defines the value of the modern customer journey. From the comfort of your armchair, you can ask a wireless device to order almost anything you want and have it delivered within minutes (for pizza and taxis) to hours or a day for most other physical products.
Companies now need to cater for dozens of digital devices and deliver personal experiences. Doing all of this is tough, and that’s where AI can help. Rather than look at how to use AI for specific channels, it’s important to apply its value across all aspects of the customer journey.
Over the past two years, humans have generated more content than in the total history of humankind put together. There are simply not enough people capable of creating the amount of content to achieve one-to-one personalization and to deliver that content in each language.
So it is essential to harness a supply chain that uses technology to create content and rely on machine translation to translate it for your web and mobile properties, emails, e-discovery, forums, etc. For comparison’s sake, SDL translates 20 billion words a month using machine translation, while our human translators can only do 100 million.
Does that mean humans no longer have value in the digital experience space? No. What it means, however, is that humans play an increasingly strategic role, while AI and machine learning can complement humans to automate repetitive tasks at scale and help make relevant delivery of personalized content a more pervasive reality.
- Prepare their content to be processed by AI/ML by ensuring that content is stored in a structured format that machines can understand.
- Gather your customer data if you haven’t already as this will be a critical set of data for machines as they learn what the preferences are for your prospects and customers.
- Invest in the existing technology that automates mundane tasks and non-creative tasks. For example, machine translation is a mature technology that speeds up time-to-market for translated content.
- Investigate and experiment with emerging technologies such as Natural Language Generation and adaptive personalization and learn how they can help you increase customer engagement.
Companies need to measure the effect of these investments to understand how they impact customer engagement and how they impact internal efficiency.
Both the top- and bottom-line numbers tell you if you’re heading in the right direction. This exploration of automation opportunity and innovation is crucial for companies that want to maintain their lead through AI-assisted processes and tooling.
MTS: Thanks for chatting with us, Peggy.
Stay tuned for more insights on marketing technologies. To participate in our Tech Bytes program, email us at firstname.lastname@example.org