Natural Language Generation: The Future of Content Management
Demand for content is increasing faster than marketers can produce it thanks to an ever-increasing need for more personalized content and the evolution of content marketing. As a content marketer, I personally feel this pain. The amount of content needed to drive excellent digital experiences has easily doubled, maybe even tripled. The reason? The payoff is big. More current and targeted content leads to improved digital experiences, which in turn leads to deeper customer engagement and increased revenues. It’s easy to see why “content marketing revenues are projected to grow at a 14.4 percent compound annual growth rate from 2017 to 2021” according to Dawn Papandrea of NewsCred Insights.
So how can you keep up with this insatiable demand for content? Many digital leaders are turning to the use of natural language generation (NLG)--an AI marketing tool also referred to as machine learning content generation--to scale content creation and ease some of the pressure on marketing by taking over routine content creation tasks. One of the easiest ways to incorporate NLG into your content creation is through a Digital Experience Platform (DXP) or CMS.
This blog post will provide actionable guidance for marketers and e-commerce managers on how NLG can help:
Auto-generate editorial text with human-like accuracy without needing additional human resources, giving you greater operational efficiency.
Create content variations for unlimited personas for highly personalized digital experiences to influence buying behavior and gain a competitive advantage.
Reduce the time spent creating large quantities of content down to mere seconds, delivering economies of scale and faster time to value.
What is Natural Language Generation?
NLG is an AI-driven software solution that extracts data from a variety of sources to produce naturally worded prose. In this way, it transforms massive amounts of data—raw data and metadata (such as style and color)—into human-readable text. Imagine auto-creating copy for thousands of retail products in seconds, or auto-generating stock performance reports with lightning speed. NLG, in which a computer writes something, is not to be confused with Natural Language Processing, in which a computer reads something, or Natural Language Understanding, in which a computer understands something, as in the case of chatbots. To learn more about the differences, we’ve found this post to be a helpful resource.
According to Kaushal Mody of Accenture, “Natural language generation uses machine learning to mimic the ways human analysts learn from data and provide recommendations for action. As such, the technology turns raw data into human narratives; communicating meaning in the same way people do, and providing complete transparency into how analytical decisions are made.”
Andy Crestodina, Chief Marketing Officer and Co-Founder, Orbit Media Studios, Inc. asserts that, "Effective customer service and personalized marketing to the masses require fast content creation, and natural language generation helps meet this demand. Systems that process natural language from the audience and other sources to generate natural language responses have a huge advantage: scale."
The ability to automatically generate high-quality text (e-commerce product descriptions, promotional text, SEO-relevant descriptions, job descriptions, etc.)—at scale and with no additional human resources—greatly increases productivity and time to market.
According to Mike Gualtieri of Forrester Research, “This technology strives to express information stored and modeled in software in natural language that humans can understand as if they were talking to a native speaker. Applications use NLG technology to speak or converse with humans. For example, intelligent digital assistants such as Amazon Alexa talk back to humans who ask them a question. Enterprises can use NLG to provide employee-less customer service agents such as Amelia from IPsoft and Watson Engagement Advisor from IBM. Enterprises can also use NLG to produce software-written narrative reports.”
How Does NLG Work?
NLG uses linguistic algorithms to render structured metadata into human-readable text that is indistinguishable from content written by people, ensuring that:
Tone is correct, consistent, and coherent; and
Grammar, spelling and syntax are correct without the need for review and approval.
In this way, NLG can be used to automate repetitive tasks, such as frequent updates of the same text or for data-driven mass content production. This cutting-edge capability can be a real game changer for organizations that produce a lot of content.
Already, AI-driven natural language generation is producing reports on sports, weather, financial results, travel, and more, increasing dwell time on site from visitors and reducing writers’ manual workloads. We’ll look more deeply at specific use cases below for marketers. What’s especially interesting is how good AI has gotten in generating high-quality text—on par with human authors according to readers’ perceptions.
For example, in a study conducted by Christer Clerwall, “respondents were subjected to different news articles that were written either by a journalist or were software-generated. The respondents were then asked to answer questions about how they perceived the article—its overall quality, credibility, objectivity, etc.”
“As we can see from the results displayed above, respondents found that the software-driven text was found to be more informative, trustworthy, and objective, while journalists’ copy was more pleasant to read. In most of the other categories of measurement, they’re neck and neck. And this was four years ago; no doubt a similar study today would yield even better results for the ‘robot journalist copy’.”
In September 2017, Digiday UK wrote that The Washington Post had published 850 articles produced by AI. It’s evident that NLG is here today, and will soon be the new normal. Leading marketers are already beginning to tap into the power of NLG to improve their team’s efficiency and impact.
Specific Use Cases
Organizations with structured data from one or more sources and repetitive production patterns (based on schedule or user request) represent a typical use case for natural language generation. If you are struggling with the need to generate large volumes of text with routine frequency, this may be just the boost you are looking for. We like to say that if you are looking for a way to automate the tasks you like the least, look to NLG to help. Let’s dive into several use cases by industry and/or function:
Digital Commerce NLG is perfect for e-commerce marketers needing to create thousands of product descriptions for entire product catalogs.
"Creating custom content across a portfolio of tens of thousands of products simply isn't practical for many ecommerce websites. Using Natural Language Generation, marketers can automate the creation of certain kinds of content following the best practices of what has been most successful, saving time, resources and improving performance." Lee Odden, CEO, TopRank Marketing
You could also use it to automatically update category pages (for example, with seasonal promotions or buzzwords that are relevant to specific times of the year; i.e., “Perfect for Valentine’s Day”).
We mentioned personalization earlier, and digital commerce is a prime example of where NLG excels. For example, combining personalization and NLG gives you the means necessary to understand if someone is shopping for themselves or for someone else and to generate the correct language in context; i.e., “Perfect for Your Birthday” versus “Perfect for Her.”
In addition, consider using NLG to update category text; i.e., “Whether you are looking for A, B, or C.”
Finance & Insurance Many news reports on stock market results are already being generated by AI-driven NLG software, and the applications in this industry are both wide and deep. For example, if you were a global banking concern and wanted to have press coverage for all of your local branches, you could rely on NLG to create news releases based on the demand for mortgages in each location, generating insights into the local real estate markets and automating a previously menial task.
NLG can also automate the creation of compliance reports, account statements, and any other data-driven copy helping you bypass time intensive data crunching.
Human Resources Imagine your job is to write up and promote the hundreds or thousands of open positions within your company. How would you ever accomplish a task like that? With NLG, you can not only automate the creation of job descriptions, but also their publication to any channel such as your own website or third-party job website.
Travel & Tourism Just like product descriptions in digital commerce, NLG can automate descriptions about locations, hotels, resorts, their grounds and rooms, as well as bars and restaurants – even events. As long as there are specific data points that can be used, NLG can create the copy.
Brick & Mortar If you are a corporate marketer trying to bridge the marketing and communications gap with your local store managers, NLG might be a great tool for you, as it can easily customize store landing pages with data such as local hours, opening hours, and more. HQ can even add in their own custom messages, such as a nationwide campaign.
Publishing NLG is also perfect for the publishing industry that needs to create thousands of news stories for a particular topic. Kelly Liyakasa of AdExchanger asserts that natural language generation allows publishers to create articles more quickly, cheaply, and potentially with fewer errors than human journalists. “It’s a critical capability for the large-scale news agency, whose content is used by other publications and journalists to develop their own localized editorial.”
Leading publishers and media companies such as Forbes, The New York Times, and the Los Angeles Times are already automating news content. The Associated Press, one of the world’s largest and most well-established news organizations, has been using AI to automate the creation of articles for several years now.
According to Francesco Marconi of the Associated Press, “To give you a sense of the impact of this first project, we went from producing about 300 stories to close to 4,000 each quarter, which was a 12x increase in content output. We also saw a reduction in error rate and were able to free up 20% more of reporters’ time to focus on higher-value [projects].”
Additional Applications The creation of any other copy that results from data can be automated with NLG: weather reports, traffic reports, etc. One of the more interesting applications is that NLG can help tag images for SEO purposes, an undeniably manual task. In this way, it abstracts information from pictures, combines that information with product descriptions, and creates new text to drive better search results. This can also be helpful for creating more descriptive experiences for visually challenged visitors who rely on screen readers.
Benefits of Using Natural Language Generation
As we’ve seen, there are multiple benefits to implementing NLG to boost content creation:
Deliver Better Digital Experiences at Scale Your organization benefits from high-quality, personalized copy that no human author would create ad hoc or cost efficiently. This helps get better search engine visibility leading to an increase in organic traffic, while also increasing engagement and dwell time.
Reduce Spend while Driving Efficiency With NLG, content authors are relieved from repetitive, routine tasks and can focus on their other projects, with more time for creativity, strategy, and exploration. Meanwhile, the organization can increase content output without additional human resources. It can also minimize translation costs, as multiple output languages can be generated simultaneously.
Increase Content Quality As we noted earlier, NLG ensures that spelling, grammar, and structure are correct without the need for review and approval. It also supports the use of the corporate verbal brand as it relates to specific words, voice, and tone.
What to Look for in a Solutions Provider
We’ve talked a lot about NLG; here’s what you should look for when evaluating digital experience platform (DXP) or content management (CMS) vendors who offer NLG as part of their solution:
Look for highly interoperable DXPs in order to connect data silos and consolidate that data. This helps to ensure clean data for better text.
Look for a platform that gives you the ability not only to control the user experience, but also the author experience for the best productivity and efficiency.
Ensure that the DXP is paired with an AI-driven personalization engine to deliver the most relevant, contextually appropriate, and compelling digital experiences for your visitors.
See NLG in action with e-Spirit’s Digital Experience Platform
NLG is already part of the FirstSpirit Digital Experience Platform, thanks to its partnership with Retresco GmbH, announced earlier this year at the company’s annual partner & customer summit.
Want to see NLG in action to help marketers and e-commerce managers boost content creation and impact?