If you go by the marketing materials of most martech companies, AI powered tools are everywhere. The irony is that marketers themselves still struggle with understanding where and how to apply AI to their strategy, process, stacks and talent to get the most of this fast-developing technology.
Marketing Content Shows Promise for AI
Marketing content — indispensable to the marketing process — is one area that shows promise for AI-powered tools. The vendor landscape has grown from around five known vendors at the turn of the decade to over 50 or more today.
On the demand side, the 2021 State of Marketing AI Report found that content marketing is one of the top areas where marketers are using AI. Of the 10 AI for marketing use-cases, four were related to content creation: creating data-driven content; predicting winning creative (e.g., digital ads, landing pages, CTAs) before launch and without A/B testing; choosing keywords and topic clusters for content optimization; and optimizing website content for search engines.
Related Article: CX Decoded Podcast: Practical Use Cases of AI in Marketing
Where Can AI Impact Content Marketing?
AI generally helps execute repeatable and automated tasks smarter, faster and at scale. Chris Penn, co-founder of marketing analytics firm Trust Insights, recently shared his experiments using EleutherAI’s latest open-source language model, GPT-NeoX-20B, for content use-cases. He found that it did “really well” generating coherent, readable text when given appropriate prompts, and suggests evaluating AI solutions for “specific, narrow tasks and deploying it to fulfill those tasks as rapidly as possible.”
Content Research, Planning and Strategy Use Cases
Even before production, AI can help develop the most relevant and effective content calendar:
- Content research: scan the internet to collect data and generate topics, keywords and content ideas based on current and historical trends.
- Audience audits: trawl through raw unstructured data — especially on social media — to find brand and category mentions, offer topic suggestions based on audience sentiments.
- Identifying best case A/B testing options to improve content performance.
Content Production Use Cases
AI is getting exponentially better at understanding, writing and speaking human-like language, incorporating local language, grammar, punctuation, brand style guides and other parameters. It infuses unfathomable speed and scale into a traditionally time- and effort-intensive human activity like writing. Accuracy, context and nuance are still areas for human intervention, even as the AI keeps learning.
Some other capabilities in this arena include:
- Develop article structure, table of content and first drafts for long-form content, saving a lot of time getting the ball rolling.
- Creating short-form content such as social media posts, paid search and display ads, email subject lines and sales copy and formulaic content such as press releases, and boiler plates.
- Content repurposing at scale, such as turning long-form content into case studies, infographics, product descriptions or even video scripts, translating content into multiple languages at scale.
- Long form technical content such as financial and annual reports, technical manuals.
- Iterative chatbot scripts based on ongoing learning from user prompts.
Content Performance Optimization Use Cases
AI can also boost content performance in many ways:
- Content personalization recommendations.
- Speed and scale of GTM from research to content creation, distribution and measurement.
- SEO performance: incorporating the best keywords and designing SEO strategies.
- Performance intelligence analytics at scale, recommendations to improve content effectiveness.
- Content and workflow standardization: smart content management platforms can help streamline the process from planning to creation and distribution — especially for larger, distributed teams. They can also drive standardization of brand identity and messaging across content formats and functional silos like email marketing and social media marketing teams.
Related Article: 8 Considerations When Selecting an AI Marketing Vendor
5 Considerations for Investing in AI Tools for Content Marketing
Where can marketers get started on the path to discovering AI tools for marketing processes and outcomes?
- You still need to start with content strategy and use-cases: Unsurprisingly, as with all technology, you first need clarity on what you want to accomplish and why. Tools are only the “how,” so ask yourself what tools can help you work faster and smarter to achieve your strategic goals, said Cathy McPhillips, chief growth officer of the Marketing Artificial Intelligence Institute. Identifying the right use-cases is a natural outcome of this process, so the most repetitive and mundane but important work can be done by AI powered tools.
John Cass and Scott Sweeney, co-founders at AIContentGen, which advises marketers on AI-content tool selection, say the three key parameters for tool selection should be quality of content generated, ease of use and the tool’s research capability.
- Resource allocation: Writing resources are expensive, and can be redeployed to where they are really needed. Focus on pushing much of the heavy lifting to AI, and reallocate humans to more value-creation tasks. While the role of marketing copywriters and content agencies is still evolving, McPhillips said they will need to learn how to work with AI. Not so much in terms of using the technology, which is increasingly plug-and-play, but using their newly freed-up time on more important and rewarding tasks such as content context and relevance; being more involved in quality control, editing and fact checking; and tweaking the tone and manner of copy, etc.
- Process realignment: Integrating AI-powered content marketing tools into the overall martech stack is important for efficient and seamless workflows, but equally important is the question of incorporating subject-matter expert, management and legal approval workflows into content created by AI.
- Build an ecosystem of experts: AI in all aspects of marketing is inevitable. CMOs need to think about building internal expertise for optimal use of the technology and getting people on the table for training the AI to generate more relevant industry and brand-aligned content. If AI-written copy is the ultimate form of regurgitation, how can marketers infuse their original brand voice into content?
Penn said large brands can “fine-tune” very large pre-trained models with their own data (for example, all the blog posts they have ever written) to capture the brand voice. This approach is a lot less compute-intensive than building the entire model from scratch, and SaaS vendors may even offer it as an additional “customization” service. Smaller brands may end up with more generic, industry-specific models (AI trained on healthcare, financial services etc.). This may widen the performance gap between smaller and larger firms. Since clever content has been an area that has let small D2C firms punch way above their weight in recent times, it will be interesting to see how AI may balance that out.
- Reassess performance measurement: AI content tools will impact efficiency (opex), and effectiveness (content marketing ROI). Will it change the way the performance of marketers, writers and content itself is measured? Sweeney and Cass said that with AI tools able to research and create content more efficiently, content creators will gain additional stature as they become more productive, and deliver better quality content and improved conversions. For marketers, content also offers the possibility of brand differentiation, so they will want to assess how AI tools can elevate their competitive advantage and drive profitability.
Related Article: AI in Marketing: Use Cases and Examples in Content Marketing
Marketers Still Need to Learn More
The potential of the technology is obvious in terms of reducing costs, accelerating revenue and even enabling differentiated content experiences. What’s holding marketers back, though, is not fear of AI, but rather the need for more knowledge and education on how best to leverage the potential of the technology in a sustained and integrated way.
What’s promising is not just the level of the technology today, but the fact that AI performance improves with use and more data. Ironically, it could be the perfect solution for a world that’s overwhelmed by content and data, while dramatically improving what we are capable of as humans.