This article outlines key concepts of Natural Language Generation (NLG) and how leaders are using this AI technology to deliver business impact. Includes multiple use cases across banking, telecommunications, insurance, media, and retail.
There comes a time in every executive’s career when she has to step outside the day-to-day and quarter-to-quarter duties of leadership and do the hard work of proactively learning how new technologies will transform her business unit or function. Former JPMorgan Chase CMO Kristin Lemkau proactively dove into how machine learning and AI was transforming marketing. Ferrero North America CEO Paul Chibe told his entire executive team across finance, sales, marketing, and operations that they needed to understand how technology was going to impact their functional area and have a point of view on how each of them was going to leverage that technology to impact their business.
NLG makes personalized digital engagement possible at scale, which is a critical part of the value it delivers.
Whether it’s lunch and learn with the CIO or a mandate from the CEO, effective leaders need to understand the technologies that will be key to growth going forward. Natural Language Generation is one of the technologies that is already impacting businesses in material ways, and yet many senior leaders aren’t well-versed on the strategic benefits of this AI technology.
Computer scientists tend to view Natural Language Generation as part of a larger umbrella field of AI called Natural Language Processing (NLP). Simply put, NLP technology converts human language into structured data that a computer can interpret. Natural Language Understanding (NLU) is the technology that interprets human language to identify what the customer needs and that addresses the challenges of slang, mispronunciation, and syntax. NLG technology produces verbal or written text that looks and sounds like a human wrote it.
NLG has been gaining prominence across business applications because it’s the sharp end of this trifecta of technologies — it’s the last mile of engaging a human after a lot of language processing and computation (see recent headlines generated around the NLG model called GPT-3).
When used in conjunction with NLP and NLU, NLG generates natural, context-appropriate, and helpful responses to a customer question or request. NLG makes personalized digital engagement possible at scale, which is a critical part of the value it delivers.
Many of the business-oriented guides to NLG are rooted in specific use cases and tactical applications of NLG, which we’ll get to. However, when you look across these myriad use cases and applications for NLG, there is a common thread at the strategic level: The outputs of NLG engage people at scale.
NLG makes content personalization at scale a reality.
NLG makes content personalization at scale a reality. NLG makes personalized financial reporting across multiple segments a reality. And NLG makes all that data locked in enterprise data warehouses actionable at scale for analysts and end users alike. The NLG/ human interface is where people interact — through words — with the results of all that algorithmic analysis, content creation, and experimentation. Without NLG, the analysis and computations that happened in NLP and NLU are locked in servers or algorithms, but aren’t actionable for specific business problems.
Many executives invest in NLG to reduce costs across human endeavors (automating repetitive content creation tasks), reduce the time needed to complete routine tasks (creating industry reports from financial data), and increase sales, since NLG tests, optimizes, and experiments with content 24 hours a day, seven days a week. Here are a few select NLG applications by use case:
In banking and finance, NLG converts many types of financial data into human-understandable content, including financial reports, regulatory filings, executive summaries, and suspicious activity reports. BNP Paribas uses NLG to deliver a customizable executive summary of their (100+ page) Management Information Services book, which is an actionable report of the client’s global custody activity with information about assets under custody, settlement, corporate actions, and income. Chase used NLG to create more effective messaging for its digital mortgage acquisition campaign that increased actual applications by 82%.
In retail and wholesale, NLG can generate product descriptions for online shopping and e-commerce and also help personalize customer communication via email, web, and chatbot — all at a scale and speed that executives need to compete effectively today.
In media, NLG can produce summarized content for sporting events and financial news.
In insurance, NLG can provide customer-specific underwriting and claims data that is actionable by field personnel. Allstate uses NLG to arm over 10,000 field sales reps with customer data that pays claims faster and improves sales KPIs.
In general, as people increase their time and activity online due to COVID-19, and as digital transformation pushes deeper into markets and economies, NLG will have greater and greater impact in every area. Besides the accelerated need for generating content (e.g. product descriptions) for online shopping and other areas mentioned above, there are also emerging areas that require more content generation, for example, generating personalized educational material for online education or personalized health records for online health services. And the need for actionable insights and engaging words at scale will only increase as time goes by.
To get started leveraging natural language generation for business impact: