Updated November 2020 with expanded coverage of AI for marketing use cases.
Have questions about how to use artificial intelligence (AI) for marketing? This guide covers AI for copywriting, AI for creativity, and machine learning, along with the basics of natural language processing and natural language generation.
Let’s start with a primer on why CMOs are using AI in the first place.
AI offers CMOs and their teams the ability to deeply understand consumer wants; confidently create targeted, personalized campaigns that resonate; provide predictive customer service; and generate the growth C-suites expect. Today, a machine can understand what stories to tell, what emotions to convey, and how to combine them into a message that will systematically drive impact across every channel at every customer touchpoint.
I believe deeply — to my bones — that the most important development in the history of marketing is machine learning…it will fundamentally change our relationship with consumers.”
-Kristin Lemkau, JPMorgan Chase CMO
AI allows marketing teams to quickly analyze vast amounts of customer data to predict a customer’s needs and wants and improve the customer experience. Advanced AI systems allow a brand to better understand their customers and understand how to better communicate with them. With this deeper insight, marketers can deliver the right message to the right customer at the right time.
What if you could review a marketing campaign, point to two or three words, and say, “those words are worth $200,000 a week to our business,” backed by the confidence of mathematical certainty?
The worst feeling for a CMO in today’s data-driven marketing climate is to be called out by a CEO for a decision that was not based on data but should have been. After all, according to a recent McKinsey study, 83% of global CEOs view marketing as a major growth driver. And marketing growth today is data-driven. CMOs oversee an area no longer considered just the organization’s creative messaging arm—it’s now an expected revenue generator with financial goals to meet.
But which words and what creative contribute to that growth?
Many CMOs have experienced a CEO who asks the simple question: “Why did you put this content in the market?” The subtext is: “What data and insights drove your decision to go with that version of creative?”
Unfortunately, seven out of ten CMOs report they have little to no confidence in answering that question. This in an age of vast data on consumer preferences and multichannel optimization. CMOs report investing more than 20% of their budgets in creative. They invest millions of dollars in matching products and services to audience segments and people to drive more personalized marketing. In the end, the process of product, service, or creative messaging relies more on subjective guesswork than on objective, data-driven insights.
What if a marketing executive could confidently report that an entire Super Bowl creative campaign was based on mathematically proven insights? What if those insights were used by all creative designers, copywriters, and marketing teams, whether in-house, outsourced, or offshore?
This guide details key use cases and best practices for using AI to drive marketing and business results. It explains how AI is a powerful, indispensable, and scientific tool that brings predictability to the art of creativity.
AI in marketing is the application of machine learning, deep learning, natural language processing, and other artificial intelligence technologies to solve key marketing challenges. As customers demand high levels of relevant, meaningful personalization and brands reorganize around a customer-centric operating model, AI and machine learning become critical tools needed to develop marketing creative, improve customers’ relationships, and drive personalization.
Visionary leaders are figuring out how to use AI to augment and amplify their creative work to let the human creatives do their best creative work.
–Alex Protopapas, Persado Chief Content Officer
Essentially, AI is a machine that completes the tasks that involve a certain degree of intelligence that was previously deemed only to be available to humans. There’s even a test for humans to discern the difference between human and machine-generated output called the “Turing Test.” Named after British mathematician and computer science pioneer Alan Turing, the test determines a machine’s ability to show intelligent behaviors equal to or indistinguishable from actual human behaviors.
Consider Alexa or Siri. Those systems talk with humans in a way that users cannot easily discern if the responses are machine-generated rather than human-generated. Advanced chatbots are another prime example.
CMOs and marketing leaders will most likely use a combination of these AI-related technologies in their marketing strategies:
In short, machine learning and deep learning are fundamental subsets of AI.
But marketers have important questions about AI use in marketing:
AI also makes businesses more human in the digital age. It enables marketers to better understand and predict what words resonate with consumers, which emotions to use, and how to deliver more relevant messaging on any channel to increase human connection. JPMorgan Chase CMO Kristin Lemkau recently noted that “machine learning is the path to more humanity in marketing.”
One of the best ways to make AI more approachable, tangible, and meaningful is to apply the lens of specific marketing use cases, says Paul Roetzer, Founder and CEO of The Marketing Artificial Intelligence Institute. Roetzer noted that “I quickly realized the simplest way to understand what AI does is to look at the individual things a marketer does in a given day or in a given month and understand, ‘How could AI help them do that?’”
Roetzer explains that AI for marketing is about marketing predictions — predicting what to write about, what headlines will work, what image will work best, and how much money to spend on which channel.
Use cases enable marketing leaders to answer key questions and prioritize their investment in AI for marketing.
In 2019, the Marketing Artificial Intelligence Institute surveyed a number of marketing professionals to understand what they used AI to help them do. The top five use cases out of the 60+ listed were:
The easiest way to get started with AI for marketing is to experiment with specific use cases that impact specific parts of the business to quickly see what works and what doesn’t, and how to grow and scale from there.
From content personalization, copywriting, email subject line optimization, and adding the correct emotions to marketing messages, AI helps CMOs and their teams better engage with consumers.
By Assaf Baciu, Persado Co-founder and Senior Vice President of Product and Engineering
CMOs and senior marketers want the ability to have an always-on conversation with customers, but they often meet resistance from customers who want that conversation to be authentic and personalized. Regardless of the technology they use, it feels as if brands are talking to their customers with their backs to them.
Imagine if you were meeting a friend or colleague at a cafe and they turned their back to you when they sat down. How would you feel if your friend started talking away from you and you started talking to their back?
Not a very polite — or effective — way to have a conversation, either in-person or online!
Yet this is how most marketers talk to their customers today, despite having exabytes of data on preferences and predictions. For a function that owns the voice of the customer and is arguably the best poised to deliver a customer-centric experience, marketers lack the context and understanding of what words resonate with customers and are often unable to act on any customer-level insights across the customer journey.
Marketers talk to customers without listening to their responses. AI changes this dynamic and allows marketers to listen to their customers and respond in a way that is appropriate and effective for the conversation and the moment. AI is the only way to do this because CMOs need computational power, memory, and the algorithms to choose the right words, adjust the words, and react to what customers actually respond to. That’s how they can propel their brands into the future.
Persado allows brands to turn and face their customers, listen to the reactions, and adjust the choice of words every single time they talk to customers. AI communication can become authentic, just like talking with an actual person.
Creatives can use past experiences or gut instincts to develop messaging. That’s the art and unpredictability of the current creative process. But AI collects and organizes enormous amounts of marketing campaign data and produces high-impact machine-generated messages to consistently drive proven results. That’s the science and predictability or AI-powered creative.
In 2020, AI for marketing slid into the Trough of Disillusionment according to Gartner’s 2020 Hype Cycle for Digital Marketing and Advertising. The overuse of AI as a buzzword has created a scenario where marketing and sales departments have enabled their own confusion about what AI is currently vs. the future ideal of what marketers want it to be and do. Is AI machine learning? How is machine learning different from deep learning?
Today’s CMOs must have the ability to create the right personalized experiences and messaging to the right customers at the right time. Only AI can empower them to deliver at the scale of today’s markets. Only AI can use big data analytics, machine learning, and deep learning to remove the guesswork from messaging campaigns and target the messaging to increase the return on investment.
AI is already transforming what has always been considered the exclusive domain of humans: the creative process. Is it possible for AI to deliver new and even better ideas to creative teams? How are creative teams using AI for creativity today?
Creativity is hardwired in the human DNA. Only humans could see a piece of marble, use their imaginations, and create the statue of David, columns for the Parthenon, or the façade of the Taj Mahal. Yet in the age of digital transformation, is it possible for a machine to leverage millions of customer data points to inform, and transform, how ideas are generated and the creative process itself to create even better results than humans could achieve?
CMOs and their marketing teams face intense pressure to create lasting brand equity and value.
Gavin Miller, Vice President and Fellow at Adobe Research noted that the demand for creativity and creative content will only continue to rise, but the quality of that creative output is constrained by people, time, and how creative humans can be. Marketing leaders have a wealth of data at their disposal from which to gain key insights to drive powerful customer experiences that reinforce their brand values.
AI for creative development turns that problem into a strength, taking large amounts of data, learning from it, and working with humans to generate the best creative results.
85% of CMOs know their organizations’ future business success hinges upon creativity and big ideas that build the brand and create an emotional connection, according to a recent survey from Dentsu Aegis Network. Only 54% of these same CMOs believe they deliver on that promise. Another 84% said data collection, management, and analytics that drive consumer insight are important for their success, yet only 49% believe they deliver well in this area.
AI helps bridge the gap between awareness and action.
AI for creativity provides scientific insight into customer needs and expectations, personalized messaging, the customer journey, precise customer targeting, effective content, growth opportunities, and targeted messaging campaigns. It helps marketing teams wade through immense amounts of data to get to the actionable insights they need to create effective campaigns.
Machine learning has evolved from simple learnings, such as comparing pictures of dogs and cats and correctly classifying them, to beating Grand Master chess players, writing full-blown fiction novels, creating video advertisements, and generating movie trailers. “Ideation is no longer in the realm of only humans,” noted Persado co-founder Assaf Baciu.
by Alex Protopapas, Persado Chief Content Officer
Humans have had a hard time making sense of AI and wondering how it will affect life as we know it–and because humans love binaries, the narrative has historically been “Human vs. Machine.” It’s new, therefore it must be an enemy.
In marketing, there are activities that are traditionally thought of as the sole and unique province of humans–namely, conceptualizing, developing new ideas, and generally being creative. Human creativity is invaluable and therefore untouchable; the mention of AI in this context is immediately perceived as a threat, as if the purpose is to get to a future where you either have one or the other.
Real life, though, is showing us that this way of thinking about technology is outdated–and it’s holding us back. There’s no human vs. machine battle. It’s the meeting of the two that is changing the way ideas are born. It’s the human/machine partnership that matters. Visionary leaders are figuring out how to use AI to augment and amplify their creative work to let the human creatives do their best creative work.
A real, recent example:
Persado analyzed three years of a retail client’s data spanning over 2 million messages and almost 100 million customer touchpoints to pinpoint the most powerful concepts and messages for this client’s customer base. Those insights informed the entire creative process and guided how the creative agency of record planned and developed one of their largest campaigns of the year. Persado’s AI had a seat at the creative table from the beginning of the process, fundamentally changing how the creative team developed the creative direction and generated content–from in-store catalogs to radio ads, and from web banners to Instagram stories.
The goal isn’t to replace human creativity but to enhance it, make it more sophisticated, and take it to the next level.
Persado uses natural language processing and machine learning to run experiments that predict which words will perform best in a given campaign, brand, channel, or season. Because the system uses machine learning and not simple A/B testing, it has a memory of which words work, when they work, and for whom. That “memory” serves as the backbone of data-driven insights that transform the creative process by fueling creative teams with the specific words, concepts, and ideas that are proven to resonate with customers. If a creative team knows that the concept of “reward,” and all the words, phrases, and images that go into that concept, is what resonates with a specific customer segment, the team is then free to use their innate human creativity to bring the concept to life.
Another example is the firm Textio, which uses AI and machine learning to help companies create job postings that resonate with a more diverse hiring pool through augmented writing. Its software helps HR directors and their teams create job listings using words and phrases it predicts will reliably resonate better with certain groups and remove bias.
Textio analyzes an organization’s writing to build a record of how the organization sounds. The more written words, the more it learns and the faster it learns. It creates a numerical score for a written job listing based on a comparison with listings for similar roles in the same geographic area. It then predicts how well the writing will attract the best talent, suggests the word or phrase changes that will more likely attract qualified and diverse candidates, and writes along with the writer to suggest that right word or phrase. Its AI system can predict words and phrases that attract more women or more men, a system that Cisco has used to hire more women and minority employees than it ever has.
Global hotel giant Hilton Worldwide and IBM partnered to create “Connie,” the hospitality industry’s first Watson-enabled hotel concierge robot. The AI-powered system works with Hilton staff and is stationed near guest reception. It merges knowledge from IBM’s Watson with AI-powered recommendation engine WayBlazer to create highly personalized information for guests, such as local tourist attractions and dining destinations. It greets guests upon arrival and answers questions about hotel amenities, services, and hours of operation.
Over time, it only gets better at its job. The more guests interact with Connie, “the more it learns, adapts, and improves its recommendations,” Hilton says. “The hotel will also have access to a log of the questions asked and Connie’s answers, which can enable improvements to guests’ experiences before, during, and after their stays.”
Machine learning is the path to more humanity in marketing.
-Kristin Lemkau, JPMorgan Chase CMO
In both the Textio and Connie cases, AI helps machines create content, foster interaction, and create responses as good as, and better than, actual humans can. Collaborations between humans and machines will only increase and grow more profitable. An Accenture
study predicts the human-machine collaboration could even boost revenue by 38% by 2022.
AI will continue to improve as long as researchers and enterprises fuel its learning with ongoing examples, rich data, and effective algorithms that turn the data into actionable insights. AI is also faster and more efficient at data collection and analysis. Rather than viewing AI as a competitor to the creative process, CMOs should view it as a collaborator that can swiftly produce creative triggers based on copious amounts of data-driven insights.
For instance, a CMO’s AI for marketing platform can create far more personalized web page copy, email subject lines, social ads, headlines, and CTAs at a faster pace than human writers can. These tasks can be offloaded to the AI agent, freeing up cognitive space so the humans on the marketing team have more time and space to be creative and build on recommendations from the AI system. The technology already exists to ensure the AI-generated creative assets adhere to and amplify the brand style.
Two key technologies powering AI for creativity are natural language processing and natural language generation, which enable brands to act on the insights generated by machine learning algorithms.
How long would it take you to write 500 personalized thank you cards after a wedding?
Could you write each one in the tone and style that you know would resonate with each guest based on your conversations with them, ranging from the bride’s weird Aunt Suzie all the way to Uncle Bill, that distant cousin twice-removed on the groom’s side?
Natural language processing enables marketers to generate on-brand, personalized messages at a scale never before seen, so the 500 thank you card project would take seconds, as opposed to several weekends.
The benefit of using NLG-generated marketing messages and powerful statistical methods is that brands can deliver higher performing messages generated by NLG, but also understand how each component performs and which contributes the most to the outcome.
–Panagiotis Angelopoulos, PhD, Persado Chief Data Scientist
Marketing leaders have similar challenges when it comes to delivering the right message at the right time to the right customer, and that’s where AI language technologies such as natural language processing (NLP), natural language understanding (NLU), and natural language generation (NLG) come into play.
Today’s CMOs and their staffs face a Herculean task—to aim creative messaging to large and diverse sets of consumers, but in a way that still creates a better customer experience, generates conversions, and increases product or service revenue. Only AI for marketing can help, which is partially why nearly 29% of CMOs’ 2018 marketing budgets went to marketing technology, according to the research firm Gartner. That number is expected to grow.
AI marketing technologies such as NLP and NLG help marketing teams personalize content and creative messaging at scale and at the rapid pace necessary in a digital world, yet retain the human feel consumers demand and the consistent brand voice marketers require. These approaches marry technical automation and sophistication with a human touch. Computer researchers and academics tend to view NLP as an umbrella field of AI that includes NLU and NLG. To better understand how they work and what they bring to marketing, however, it’s helpful to look at each separately.
Essentially, NLP algorithms are the text or voice input “readers,” while NLU and NLG are the “understanding” and “writing” components, respectively. Everything starts with NLP and then flows into increasing levels of the nuanced complexity that is human communication.
by By Panagiotis Angelopoulos, PhD, Persado Chief Data Scientist
Persado created a unique and comprehensive database for marketing language that, far greater than just a dictionary, groups and tags words and phrases into hierarchical categories with words and phrases associated with each. For example, to express “Gratitude,” Persado collected thousands of ways to express it: “thanks for your interest,” “we appreciate your loyalty,” “our way of saying thank you,” etc.
These categories form the backbone of Persado-generated messages. By changing the positions, values, or words, Persado can easily produce a tremendous amount of language combinations.
Persado needed a statistical method to validate natural language generation’s efficiency, so it developed a proprietary machine learning method that uses experimental design with a small set of combinations and evaluates how each language category (emotion, position, format, length, etc.) and its values impact the outcome. Data scientists can not only view the performance of the overall NLG-generated message, but can also understand how each component in a message performs and which contributes the most to the outcome.
Persado has collected an enormous amount of marketing campaign data after working with clients from different industries for many years. And that enables the data science team to create predictive algorithms that produce high-performing machine-generated messages.
What is natural language processing?
Natural language processing is a convergence of computational linguistics and computer science. It is the primary method to analyze human language and break it down into smaller components, such as part-of-speech tagging, word segmentation, terminology extraction, speech recognition, or lexical semantics. Any middle school student who has diagrammed a sentence has witnessed firsthand how NLP algorithms approach their language analysis tasks.
What is natural language understanding?
Natural language understanding is the next step in the NLP process that leads to NLG. Once a machine analyzes human language at the “parts” level, the algorithm attempts to “understand” the communication’s meaning. That includes trying to interpret the sender’s emotion. NLU seeks to answer the question, “what is the intent of this message?”
What is natural language generation?
Natural language generation moves NLP from syntax analysis into the world of semantics: it produces meaning through content creation. The algorithms create words, phrases, and content based on processes similar to human methods, such as determining what information to include, logically ordering information presentation, using the appropriate grammar, and choosing the “right” verbiage.
Examples of natural language processing
NLP is the pervasive base that anyone and everyone who has to analyze text must use, including Siri, Alexa, text-to-speech capabilities (such as trying to send a text message using voice commands), Google Translate, email spam filtering, etc.: the list is almost endless. So, most consumers use NLP every day without realizing it when they interact with their devices’ speech-to-text translation functions. NLU is also used in any and all systems where humans interact with machines via human language protocols.
For example, Siri has to understand several aspects of what someone asks it to do. “Siri, where’s the nearest bank?” must be parsed so it understands what a “bank” is, comprehends that the command is a question, accesses geo-location, compares your location with the location of the closest bank, and then generates a response. As long as Siri doesn’t have problems deciphering word pronunciation (heavy accents, background noise, or speech impediments can present problems), then it will likely reply with an accurate response, thanks to natural language generation.
NLG continues to gain traction for use in generating long and short-form content. In another example, The Associated Press has partnered with Automated Insights, the developer of the Wordsmith NLG platform. Prior to using Wordsmith, AP reporters generated only 300 stories on corporate earnings each quarter. After adopting the Wordsmith NLG, the AP now “writes” over 4,400 quarterly earnings stories.
Wordsmith doesn’t perform these functions completely on its own. AP and Automated Insights worked together to map editorial decisions and bake those into the Wordsmith algorithms. Over time, when machine learning if-then rules give way to more intelligent capabilities–such as generating its own editorial rules as AI detects shifts in reader responsiveness–the reports will become more personalized.
NLP enables much more than just email subject line optimization
One of the early use cases for NLP was to optimize email subject lines to improve direct response marketing campaign performance. While this is still an important use case and one where many brands find value, NLP has opened up never-before-seen possibilities for delivering value across the enterprise.
Business leaders in healthcare can now generate personalized messages that remind patients to take medication and get their flu shots because the AI system can process, understand, categorize, and generate the right message at the right time. Business leaders in operations can use AI to generate billing notices and collection notices that improve accounts receivable revenue.
The future of personalized marketing at scale depends on natural language generation. Otherwise, marketing, creative, and copywriting teams will spend their time writing the equivalent of millions of thank you cards in the race to be relevant and to achieve the proverbial one-to-one marketing.
NLG is a natural fit to deliver better copywriting and, contrary to the popular narrative that AI is poised to replace copywriters, human copywriters now use AI to augment and improve their creativity to deliver their best work.
What if humans could leverage AI for copywriting to do their best work and deliver powerful, personalized content at scale? What would that look like and is it even possible?
Rising consumer expectations for personalized experiences mean CMOs and senior marketers face increasing pressure to create content quickly and at scale. Copywriting that meets these consumer preferences demands human creativity but is notoriously difficult to scale. It’s an art to set the tone and style as well as tailor messaging. It takes human touch and subtlety to create and understand humor, parody, or satire sometimes found in copywriting.
Even the most talented copywriters do not possess AI’s data-gathering and analytical capabilities.
–Kat Dessenon, Persado Vice President of Campaign Management
It’s now possible to use linguistic science, machine learning, and natural language generation to guide the art of copywriting in a scientific way and create messages that are on brand every time.
AI can be used in copywriting to speed up repetitive processes, create consistency, and increase creativity. For example, AI can remember grammar rules so copywriters are free to just write. Or, it can come up with keyword suggestions so writers have starting points that vastly expand the size and richness of their typical lexicon, instead of having to spend time researching the same set of recycled words on their own.
AI copywriting in action
There are smaller tools, such as Grammarly, the Hemingway App, and Acrolinx, that can help the copywriting process at a simpler level; e.g., on issues with grammar, sentence structure, weak adjectives, overuse of passive voice, and tone detection. These tools assist the human writer with ensuring clarity and improving efficiency. Both are helpful but still limited to specific applications.
by By Kat Dessenon, Persado Vice President of Campaign Management
Persado trained its AI platform to detect emotions, which is one of the most difficult language elements to master. This breakthrough in AI technology involved much more than simply tagging individual words; the AI system categorizes phrases and the various components of language so the same components can be reimagined and reassembled to deliver more compelling messages. The system matches brand voice and audience impact analytics to produce optimal copy throughout every consumer touchpoint.
Persado uses AI to help automate some aspects of content creation to increase efficiency, allowing teams to focus on more strategic tasks that require human judgment. This is the biggest benefit to businesses–AI’s purpose, in marketing and in general, is not to replace the human workforce, but to free up time for work that computers can’t yet do on humans’ behalf.
The larger the organization, the more pervasively AI can be used. Assuming a large, relevant base of content exists, CMOs and their senior marketers are more likely to get more precise and relevant output with AI. A small business that sends email once a week won’t have as much benefit as a larger enterprise that can leverage thousands of data points.
Popular use cases of AI for copywriting
The Associated Press uses the NLG tool Wordsmith to generate elements of its quarterly earnings reports. Major publications such as Reuters, Bloomberg, and The Washington Post have also leveraged AI for copywriting and as news generation and gathering tools.
Bloomberg uses an AI system called Cyborg to help its reporters churn out thousands of articles on company earnings every quarter. Generally speaking, quarterly earnings reports can be formulaic for reporters. Most report a quarterly rise or fall in earnings or net loss by dollar amount and per share amount; report a rise or fall in revenue; offer some company or sector insights that might explain the rise or fall; quote a company or sector expert about the financial results for the period; and then wrap up the story with a look ahead to expectations for the next reporting period.
Cyborg analyzes a 10-Q or 10-K earnings statement and produces an immediate earnings story with the most newsworthy financials. Although AI earnings stories lack the context a seasoned reporter can bring, the financial numbers are more accurate. After all, it’s a machine and it doesn’t make human mistakes, such as typos.
The Washington Post uses its in-house automated storytelling AI system called Heliograf to write bylines for other formulaic news stories, such as local sports. In 2017, it began using Heliograf to cover D.C.-area high school football games. This freed its sports staff to go to only the most important games in person and to create more extensive, in-depth coverage. Heliograf creates stories by gathering information about scoring plays, player statistics, and quarterly score changes. These stories can be later updated each week with box-score data from coaches or other relevant contextual information. The results are tough to discern from those written by a human:
“The Yorktown Patriots triumphed over the visiting Wilson Tigers in a close game on Thursday, 20-14. The game began with a scoreless first quarter. In the second quarter, the Patriots’ Paul Dalzell was the first to put points on the board with a two-yard touchdown reception off a pass from quarterback William Porter.”
How AI empowers copywriters to write better copy
AI can script simple commercials such as the “Driven by Intuition” Lexus commercial, and AI as a poet has made strides in writing poetry that sounds human. From a marketing perspective, AI copywriting tools are the next phase of deploying individualized content. Even the most talented copywriters do not possess AI’s data-gathering and analytical capabilities. Spinning out content to each individual customer through every possible touchpoint is not a realistic expectation for a human copywriter. And human copywriters cannot perfectly capture just the right words at precisely the right time for every one of the millions (or billions) of consumers.
AI gets CMOs and their teams closer to the right words at the right time at scale. Given the increasing sophistication of AI for copywriting tools, their utility in freeing up a writer’s time so they can focus on higher-level tasks, and the demand for hyper-personalized content, marketing departments should begin exploring AI tools, no matter how small. One of the first and most popular use cases for AI and copywriting is the creation of email subject lines.
Today’s marketing teams need a targeted, personalized approach for email subject line optimization that is more captivating than: “Hi , we noticed you left something in your cart.” They need to know the words they choose are the exact right words so their customers take notice and don’t instantly hit the delete button. Email is also a favorite channel for CEOs to ask the dreaded “Why did you choose that phrase?” question for subject lines.
Machine learning and AI have helped email marketing campaigns evolve beyond the spray-and-pray approach. Everything from generating personalized email body content and optimal email subject lines can now be improved with AI and machine learning.
What AI does or what we want it to do…is it allows you to scale so much faster than doing that champion/challenger test. Because champion/challenger will work very well when you are in the early stages of your development, but it will really start to slow you down at a later stage.
–Ben Blake, CMO, Hotels.com
For email subject line optimization, AI arms marketers with the knowledge of what worked and why for all the myriad elements that make language such a powerful communication medium.
Why pursue email subject line optimization?
Email marketing is digital marketing’s workhorse. It’s inexpensive, easy to deploy, leverages first-party data, and has proven reliable over the past two-plus decades. The Data & Marketing Association reports 73% of consumers say email remains their preferred marketing channel. That’s a growing audience that has opted in and is willing to receive a message. Yet nine out of ten email programs still fail, according to Forrester. Using AI for subject line email optimization is one way to help.
by By Frank Chen, PhD, Persado Natural Language Processing Scientist
Persado’s marketing technology and AI allow clients to test a myriad of marketing content and determine the best predicted messages, but not in a way most marketers are accustomed to doing it. A typical marketer uses A/B testing to compare control copy vs. a piece of test copy. A/B tests are useful to show which of a small handful of options performs best, but A/B tests are notoriously hard to scale and don’t have any memory from one test to the next.
It’s difficult to pinpoint the absolute best possible variation of content such as copy on a webpage using A/B testing because you literally have to test each possible combination against every other combination, sequentially, and over time. Do superlatives work best and if so, which one do you use–Best, Ultimate, or Top? You’d have to test Best vs. Ultimate, then Best vs. Top, then Ultimate vs. Top, and so on. Is it better to use the emotion of Gratitude (We appreciate your loyalty) or Achievement (You earned it!) and, once that is decided, which phrase is the best option? A/B testing becomes untenable as a primary testing option at scale.
Let’s say you determined that “You earned the ultimate secret bonus” was the best headline option in an A/B test. Using A/B testing, you still wouldn’t know why or be able to replicate it or transfer the learning to another channel such as social media ads. Because A/B tests have no memory from one test to the other, after a year of optimizing all your content, you’re no smarter about which language elements work and which are a waste of time.
Persado uses AI, machine learning, and experimental design to compare many different elements at the same time to determine which elements work, why they work, and then put the best messages into production to drive more engagement.
The Persado platform uses natural language processing algorithms to “read” messages and break them down into their component parts then compares and analyzes that copy against potential high-performing variations. The platform draws insight and “memory” from a unique knowledge base of over 1 million categorized words. The AI system then proposes variations that incorporate many different language elements, including the specific word or phrase used, the emotion of the words used, the role and placement of emojis, and the effect of all caps, to name a few. The platform then uses natural language generation AI to take all the learning and analysis into account and “write” new and more effective messages that outperform the control message.
AI beats A/B testing every time
AI is a fundamentally different, and better, way to optimize email subject lines. While A/B testing shows you which one of two options performed better, it still relies on human-created messages and may not test the absolute best option, which may not have been thought of yet. Let’s say a marketing team wants to create a subject line to introduce its new product or launch a campaign. The team will brainstorm and use past experience, personal opinion, and “what worked last year” as the guiding design principles for creating the options to test against a control. They may land on “Discover our new service today” as their best option.
The platform then uses natural language generation AI to take all the learning and analysis into account and “write” subject lines that outperform the control subject line.
–Frank Chen, PhD, Persado Natural Language Processing Scientist
Alternately, a marketing team could mine historical data with machine learning AI and use natural language generation to develop a structured statistical experiment that creates entirely new concepts and language forms that far outperform the control and even the best alternative created by the human copywriting team. The A/B test may determine that the copywriters’ best option, “Discover our new service today,” was the winner that drove a 23% improvement in return rate. Yet the AI-generated subject line of “Welcome to a whole new way of doing things :-)” generated a 31% improvement in return rate, and that phrase was not even on the list of phrases created by human copywriters to test.
That’s the power of AI for marketing and subject line optimization.
The future is already here for AI in marketing across copywriting, content personalization, and creativity. It’s already possible to use advanced algorithms to predict customer churn, generate more relevant content for individual customer segments, and even transform the ideation process through actionable, data-driven insights based on actual customer behavior.
The next frontier lies in using AI to augment and extend one of marketing’s core contributions across the enterprise; namely, the communication of ideas and messages using words and language both inside and outside the organization.
Today, a machine can understand what story to tell, what emotions to convey, and how to combine that into a message that will systematically drive impact anywhere words are used across the enterprise, every single time.
–Assaf Baciu, Persado Co-founder and Senior Vice President of Product and Engineering
Words are central to how brands communicate both internally and externally. Customer service prompts, employee communications, health and wellness reminders, billing and collections notices, and even open enrollment messages are all opportunities to leverage AI for better business and better human outcomes. Persado co-founder Assaf Baciu noted that “today, a machine can understand what story to tell, what emotions to convey, and how to combine that into a message that will systematically drive impact anywhere words are used across the enterprise, every single time.”
Leading brands like Chase, Humana, and others are already using AI in areas such as customer service, billing, and employee communications to generate words that resonate. AI plays a central role in developing and generating the words that matter to employees, customers, prospects, suppliers, and anyone who uses words to communicate, persuade and engage.