.svg)
Today, when it comes to marketing and comms, Generative AI stands out as Silicon Valley’s latest and greatest hype job. We’ve been told it will revolutionize work and the workplace, from data analysis and campaign management to research and copy. But is it happening?
Yes and no. As someone who believes AI and Generative AI does have the power to transform marketing and comms, it’s important we’re honest with ourselves about where things really stand. Otherwise, it will be that much harder to make the progress that’s truly possible.
Let’s start with a quick rundown of how marketing and comms pros use and don’t use AI.
USE AI
Digital Media
Nearly every spend relies on Meta and Google’s algorithms, systems that in theory learn as you go, more efficiently finding the right audiences and consumers and customers. Of course, this is a pre-hype, pre-Chat GPT phenomenon and not generative AI.
Research
Teams generate quick summaries from Chat GPT, Perplexity and other LLMs on just about any subject. More advanced players use multiple LLMs to check and validate facts, catching hallucinations.
Content, Copy + Assets
Teams produce first drafts of memos, articles, and proposals. However, the more complex the piece, the more editing it needs. Voice is generally generic. Also, designers and others are generating AI images, mostly for social media. It’s very efficient, but prone to making odd errors (an extra finger) that require careful review to catch.
Now, what’s missing, where are Chat GPT, Meta AI, Perplexity and others not playing a regular role? Turns out it’s on the thinking and creative levels, the task where the most intelligence is required.
DON’T USE AI
Strategy
Most marketing and comms pros develop strategy as they always have — intuitively with as much supporting data as they can muster about audiences, product-market fit, corporate issues, etc. They work off existing business plans and generally haven’t found a role for AI beyond research (carefully checked).
Campaigns + Creative
The same goes for building a digital or PR campaign. Sure, generative AI can come up with some copy or pitches, but pros don’t turn to LLMS for brand identity, campaign themes or creativity. It’s still done through old fashioned brainstorming, conceptualizing, drafting, revisions, production, and post-production…
Analytics
Here’s the most surprising miss. LLMs are veritable fonts of data, yet marcomms pros don’t yet see relevant uses cases for analytics. It’s partly because most busy folks don’t have the time and teams necessary to upload datasets and train LLMs to find what they want.
Don’t just take my word for what’s working and what isn’t. A recent Medium piece by Cezar Gesikowski summarized the broad strengths and weaknesses of Generative AI, drawing off a Gartner analysis. The conclusions: high usefulness for content generation, low for decision intelligence and forecasting.
Missed Opportunities: Bridging the AI Gap
Yet it’s already possible to put AI to work when it comes to building strategy, messaging, campaigns, and analytics. It cannot be done with a single query and may need a Python pro or other specialized help. But it’s not a custom software build by any means. With a little work, efforts will pay dividends. A few examples:
BRIDGING THE GAP
Small Languages Models: Use Chat GPT, Perplexity or another LLM to ingest a data set, sort it, access it, and produce new content. It’s a perfect short cut to producing press releases, content, and internal communications drafts you can revise and refine.
Artificial Intelligence Optimization (SEO 2.0): The golden age of SEO, when great content ranked fast, is behind us, but there’s a new, potentially bigger opportunity: AIO. Structure your brand’s content for consumption by information hungry LLMs. Make sure the machines can find the facts, ideas and insights that matter.
Lookalike Audiences — Upload publicly-available data on who’s engaged with your content, analyze it with LLMs, and ask Meta to create a lookalike audience. A new and great way to extend thought leadership beyond the initial engagement to the right audiences and still protect privacy.
So, if we take a step back, Silicon Valley is right. Generative AI will transform every aspect of work — at least when it comes to marketing and communications. The catch is that it requires time, knowledge, experimentation to solve for strategic, creative, and analytical questions. In Generative AI’s next chapter, the winners may be the teams, creators and entrepreneurs who understand users’ needs and create simple solutions. The great news is you can do it with small teams with limited to no coding experience. What are you waiting for?
