Generative AI in marketing

Generative AI is revolutionizing the way people work in nearly every industry. It’s providing new tools that can take on the kind of creative, analytical, and organizational work that in the past was strictly the domain of humans. 

This is especially true for marketers — and content marketers in particular. Content marketing involves content creation and design as well as personalization, targeting, and cadence planning. These are all tasks at which generative AI excels. 

While there are myriad applications for generative AI, it has its limitations and can be effectively and safely used only with strict human oversight. 

What is generative AI?

Generative AI is a form of machine learning, a branch of the science of artificial intelligence. Whereas earlier forms of AI were used mainly to analyze data, generative AI is used to create — or “generate” — many different types of written, visual, audio, and video content. The technology can also be used to answer questions, interpret data and draw conclusions, write computer code, design new pharmaceuticals, or solve complex real-world problems. 

While generative AI technology is not new, recent advances have made some models — the technical term for an AI tool — easy to use despite their underlying complexity. These models can mimic the way humans speak, write, draw, plan, and strategize by using “deep learning,” a tactic inspired by the way the human brain forms associations.

Generative AI for marketing

Generative AI is already widely used among marketers and is here to stay:

• A study by the MIT Technology Review in 2022 found that only 5% of marketing organizations considered generative AI “critical” to their operations, and just 20% were making wide-scale use of it across different use cases. But by 2025, 20% of marketing executives plan for generative AI to be a critical part of their department’s function, and another 44% plan to use it across various applications.
• A 2023 study by Deloitte found that 41% of marketing, sales, and customer service organizations had adopted generative AI in a limited or at-scale implementation — a number second only to IT and cybersecurity departments. 
• A 2023 Salesforce survey of 1,000 marketers found that more than half are currently using generative AI and another 22% plan to implement generative AI in the coming year. 
• A 2023 Statista survey of 1,000 B2B and B2C marketing professionals found that 73% of them are already using some form of generative artificial intelligence.
• A 2023 Boston Consulting Group cross-industry survey found that 67% of marketing executives are exploring generative AI for personalization, 49% for content creation, and 41% for market segmentation.
• The same survey found that among those already using generative AI, between 40% and 50% are using it for social listening, predictive analysis, generating custom product descriptions, and chatbot marketing.

Generative AI use cases in marketing

Marketing organizations are finding myriad use cases for generative AI. Below are some of the most common.

Content creation

Generative AI models can write copy from an outline or prompt, and they’re handy for short-form content like blog posts, emails, social media posts, and digital advertising. But their content creation ability goes beyond straightforward text generation and includes:

• Generating images from a text description
• Conducting research and finding citations for data-driven reports
• Translating content into different languages
• Creating charts and graphs from descriptions or data sets
• Producing audio from written text, a process known as text-to-voice
• Producing short-form video 
• Composing royalty-free music
• Summarizing long-form content 
• Refining messaging or rewriting it in a different tone or voice
• Generating prospect-facing product descriptions from technical information
• Researching SEO keywords, longtails, and topic clusters
• Optimizing existing content for SEO

Content personalization

Generative AI can personalize marketing messaging at scale by leveraging customer personas and data. It can facilitate the A/B testing of marketing messaging by generating multiple versions of marketing assets. It can also adapt content based on user preferences, geo-location, or trending hashtags. 

Based on parameters like past website behavior, content and product preferences, and company interactions, generative AI can also assist in developing customer personas that drive content personalization requirements. This personalization can be extended to the customer journey itself: generative AI can help map more engaging customer journeys.

Content ideation

When they come up against writer’s block, some content marketers use generative AI models for inspiration and help with creative thinking. By collaborating with generative AI models, marketers can generate ideas for new campaigns by prompting for content that aligns a target persona with their sales and marketing goals.

Automated customer service and support

Website chatbots have long been able to handle simple, scripted “conversations,” but generative AI enables chatbots to have human-like interactions. They can offer concise, relevant, and precise replies to customer and technical support questions in many different languages. 

Generative AI models can also evaluate information —and even the tone of the question — through “sentiment analysis,” helping craft more discerning responses. Chatbots using sentiment analysis can also monitor and respond appropriately to activity on social media channels.

Market research and data analysis

Generative AI can analyze vast amounts of unstructured data and extract insights. Organizations then use these insights to guide business decisions about market segmentation, campaigns, advertising, or product and feature development. 

One example of this use case is predictive forecasting, which uses past trends to form a hypothesis about future behavior. It’s used to forecast churn rates, demand patterns, and how well campaigns or ads may perform.

Demand and lead generation

Marketers rely on strategic generative AI tools for numerous strategic marketing applications:

• Automating campaign execution, cross-channel coordination, and lead scoring
• Adjusting advertising strategies in real time
• Optimizing customer and market segmentation
• Orchestrating SEO strategies 

Generative AI in marketing: Examples

Generative AI has broad applications for marketers across a wide range of industries. Below are examples of how companies in different industries are employing generative AI in their marketing efforts.

• Determining the ideal bid range for Google AdWords by identifying patterns in historical data and using them to predict the performance of future digital ads
• Reducing churn by personalizing loyalty programs for individual consumers, using incentives tailored to each consumer’s buying patterns
• Sending only those product offers relevant to each customer, rather than using a generic approach, reducing spam and improving customer affinity
• Developing advertising budgets around ideal touchpoints by determining which ones contribute directly to sales
• Using data to customize an interactive customer journey most likely to convert a lead to a prospect in real time
• Automating reporting the ad performance of multiple digital ad programs simultaneously
• Using micro market segmentation to personalize messaging based on multiple customer data points and automatically generating content and imagery for an email nurture campaign

The benefits of generative AI in marketing

The examples and use cases for generative AI listed above yield various benefits for marketing teams. Those benefits can extend to customers and prospects as well. 

Generative AI saves time and resources and improves ROI

By accelerating content development and SEO research, generative AI helps marketing teams make more efficient use of their human resources. It can also give team members more time for strategic work.

Generative AI fosters customer loyalty

Generative AI makes it easier to personalize content at scale. It can also help create, schedule, and respond to social media posts in real time. Relevant content fosters longer-lasting customer relationships. It also improves conversion rates and lead quality.

Generative AI improves marketing strategy and outcomes

Generative AI can automate competitive research, analyze website traffic, interpret consumer behavior patterns, and predict how advertising will perform. This enables marketers to improve conversion rates by:

• Iterating and refining ad user journeys and campaign tactics
• Adjusting ad content and placement, and making better ad buys in real time
• Uncovering new market segments

The risks of using generative AI for marketing

Any application of generative AI comes with risks. For marketers, the most serious risks involve problems with content generation and include:

• Plagiarism and data piracy - The model can replicate something in its dataset word-for-word
• Proliferating intentionally false or misleading information - The dataset may contain propaganda or incorrect information
• Bias - This can be inherited from the data collection and model training used to build generative AI tools and introduced into the content it generates
• Copyright violation - The generative AI model may be built on data the company lacks permission to use
• The violation of users’ data privacy - User prompts or inputs may be collected without the user’s permission
• Unpredictable behavior beyond the tool’s planned functionality - Complex new technologies don’t always behave as planned

Types of regulation

Advocacy groups are pushing for the regulation of generative AI usage, and some governments, as well as public and private companies, have responded. 

The European Union

The EU has prohibited AI usage that poses “unacceptable risks,” but applications like content development are considered “low risk,” and regulated less stringently. Those regulations include:

• Identifying content generated by AI
• Designing models to prevent it from generating illegal content
• Publishing summaries of copyrighted data used to train generative AI models 
• Disclosing the source for content generated by AI

Because EU regulations usually affect non-EU countries doing business in or with the EU, their regulations also extend to American and British companies.

The United States

Because the Federal Trade Commission oversees false and deceptive business practices, it could become involved in regulating content to prevent the creation of deep fakes. U.S. regulations are likely to affect how AI models are developed rather than their usage.

Private industry

Companies that develop and use AI are also considering ways to lobby against as well as comply with impending regulation. They are also readying the implementation of their own safeguards and guidelines. Those companies include tech giants Microsoft, Google, Amazon, OpenAI, Apple, Nvidia, and IBM, which are prime targets of regulatory scrutiny. 

Risks for brands

Currently, brands under scrutiny with generative AI are limited to the companies behind the technology rather than the companies that use the technology. These lawsuits deal with copyright and licensing issues for training data and privacy concerns. 

While brands using AI have not yet been sued, the risk of lawsuits is real. Brands could also lose access to generative AI tools that have become integral to their business. And their brand image could be tarnished by public complaints about their use of copyrighted content without permission or plagiarized text and images.

Best practices for implementing generative AI in marketing

Risk mitigation for using generative AI

Human oversight is the most important safeguard against any of the risks posed by generative AI. AI tools should only be employed with a clearly stated strategy and goals. Marketing leaders should implement and enforce policies and practical ways that regulate how generative AI is deployed and human reviews of AI-generated content should take place. 

Such procedures should include:

• Carefully researching generative AI models to ensure they’ve been built from accurate and legally obtained data
• Implementing an AI roadmap that limits generative AI applications to legitimate use cases
• Testing use cases before rolling them out enterprise-wide
• Reviewing content for bias and other ethical considerations
• Including a disclaimer for AI-generated content
• Implementing a comprehensive data strategy, a data infrastructure for data collection, and activation that vets data to manage risk 

Generative AI tools for marketers

Tools for content generation

These three content-related generative AI tools below are popular among marketers and becoming integral to their martech stack: 

Open AI’s ChatGPT

ChatGPT, a free generative AI tool, helps marketers brainstorm, research, outline, generate, and summarize many types of content by responding to questions, or prompts. Its ease of use and versatility made it the fastest-growing consumer application in history. 

Google’s Gemini (formerly Bard)

Gemini’s functionality and interface are similar to ChatGPT’s; the main difference is the data sources used to build them. Gemini will generate multiple versions of a response to the same prompt, which gives users more flexibility. It can search for images on the internet based on natural language prompts.

Open AI’s DALL-E

DALL-E generates imagery in many different styles. It can also retouch and create varied iterations of existing images based on natural language prompts. 

Other content-generation tools include:

• Stable Diffusion - an open-source image generation technology
• Progen - a content generator for professional communications with links for social sharing and built-in safeguards against plagiarism
• GAN.ai - used to personalize videos at scale
• Anthropic’s Claude - used for summarization, search, creative and collaborative writing
• Omneky - used to customize advertising creative across all digital platforms
• Hypotenuse - a platform that generates product descriptions and advertising captions automatically
• Flick - a social media tool that creates posts, targeted hashtags, and optimizes post schedules

Strategic generative AI tools

There are dozens of generative AI tools to help with all aspects of marketing strategy, some of which may overlap with platforms already part of a company’s martech stack. Because generative AI tools can be more intuitive and easier to use, they provide access to information and functionality previously reserved for specialists and data scientists.

Among the most commonly used are:

• Alteryx - a platform for automated data engineering and analytics reporting
• DataRobot - an open platform used for developing organizational growth strategies
• Skai - an omnichannel marketing platform that automates the optimization of an organization’s ad spend
• Braze - a customer engagement platform that orchestrates customized, cross-channel journeys using dynamic market segmentation and personalized messaging

Custom generative AI tools for marketing

Marketing organizations can also build their own AI models based on proprietary or task-specific datasets. It can be a large undertaking, but this allows them to narrow the scope and increase relevance for the analysis. A custom dataset — using brand guidelines or data collected from previous campaigns — can also help generate industry-specific content or company-specific campaigns more accurately. 

Controlling the data on which an AI model is trained also manages the risk inherent in third-party tools whose data may not have been adequately vetted for bias, inaccuracies, misinformation, or copyright issues.

Amazon SageMaker is a development platform used to build, train, and deploy customized machine learning models using proprietary data sources. 

However, developing proprietary generative AI models is beyond the capabilities of most companies. Most organizations typically begin with publicly available models like GPT-4 and Gemini. Then, the organizations refine the models and update them regularly using prompt engineering, reinforcement learning based on human feedback, and other techniques. 

What is generative marketing?

Generative marketing applies generative AI to marketing workflows. Generative marketing platforms are a type of composable customer data platform (CDP) that employs generative AI capabilities. Generative marketing tools are generally campaign-related and used to improve the accuracy of a campaign’s target audience and personalize the audience journey. Those tools address many of the applications and use cases listed above. 

What’s next for generative AI in marketing?

Productivity vs. strategy

Companies’ current use of generative AI improves efficiency and productivity while reducing costs. According to a 2023 study by Deloitte:

• 91% of business leaders surveyed said they expected generative AI to improve their organization’s productivity. 
• Only 29% were employing generative AI strategically. 

But a shift from focusing on productivity and cost-savings to strategic benefits like innovation and growth is underway. That shift will likely play out in marketing organizations as well, as generative AI technology spreads from content generation applications to strategic marketing applications, according to the same study.

Interactivity and autonomy in generative AI

As natural language processing capabilities improve, other expected developments in generative AI include:

• Discussing an image or video interactively as fluently as you can discuss text-based content today.
• Interacting with generative AI interfaces built directly into applications. 

Autonomous AI — which can operate without any human interaction — is the next step beyond generative AI. Its best-known use is for self-driving cars. In the near future, it’s possible that marketers will find a way to employ these generative AI technologies as well.

‍