Will AI agents take over your marketing job? Just the sh*tty ones

Are AI agents circling our marketing jobs like vultures, or are you letting the fear (or the misuse of AI) get the better of you?

The truth is, AI is definitely impacting your jobs... but for the better. Hand over the sh*tty tasks, before you are about to get a hella lot more productive and creative.

Here's the lowdown from CMI's recent webinar where Shafqat Islam, President and CMO @ Optimizely revealed his predictions.

The AI era of marketing is here, folks

Not to freak anyone out but things are changing pretty fast in the world of AI. Like seriously, ChatGPT only launched at the end of 2022 and it was only May 2024 when Google search results changed SEO and user experience forever by introducing AI Overview.

And now, we've got AI agents that are going to become more and more integrated into our lives, beyond just work. In fact, Shafqat predicts that by 2030, "most online interactions will be driven by AI agents".

It's true, AI transitions are happening all over the shop, including:

Ambiguous > real AI: Practitioners eager to adopt but companies are behind in deploying at scale (a result of one-off point solutions) or realizing real ROI.

Reshaped workforce: AI is an enabler and an extension of your marketing org, not an eliminator. it will re-shape how we work and allow generalists to become specialists.

Agents transform marketing: The opportunity for AI agents is most pronounced in marketing, allowing CMOs to scale beyond imagination.

Okay, let's talk AI agents real quick

Let's just quickly cover what the heck AI agents are, so we're all on the same wavelength:

Autonomous executors: AI agents act independently to perform specific tasks or functions

Real-time decision makers: They analyze large amounts of data and make decisions with any human intervention

Learning and adapting on-the-go: AI agents continuously improve through machine learning and adapting to new information

Multi-functional: They integrate across platforms to perform diverse tasks like communication, data analysis, and content generation

User-centric: Designed to interact with users seamlessly, providing personalized and context-aware solutions

Long story short: they're taking the sh*tty parts of your job

For content creation and general marketing, we need to be interesting and we need to be memorable. The roles in this field will shine because AI can't be totally creative. They have to be guided...and they have to get their other information from somewhere (AKA the content that already exists on the internet).

And things are only going to get crazier and more creative when we all jump on the multi-agent collaboration train. All aboard, choo choo, etc. 🚂

How Optimizely's marketing organization uses AI

Planning and ideation: Optimizely Opal (powered) for ideation, as well as campaign and content briefs

Content creation: Opal's AI content generator for sparking creativity and creating first drafts

AI campaign kits: Generating entire end-to-end campaigns, including briefs, tasks, content and social promotion

Industry marketer: An agent that creates content and messaging to target specific industries, without needing to hire a specialist

Asset storage: AI-tagging, renditions with smart focal point for automated asset management

Personalization: Opal, multi-armed contextual bandits, web experimentation and ODP

Experimentation: Strategy bot with in-app guidance on which experiments to run and how to measure or summarize results

SDR outreach: Research contact, trying new outreach techniques, (but always personalize)

Customer marketing: trained our own model with historic customer data (customer stories database), so any questions can be asked and get descriptive answers

Agents galore: Optimizely X AI agents

We're on the road to building more and more (and more) AI agents that make marketers' lives easier.

You'll find all the aforementioned agents integrated within our Content Marketing Platform product, so less of the copy-pasting from generic tools like ChatGPT or Claude, and more streamlined AI partnership within your content workflow or content supply chain (as we like to call it).

We found that the more specific the agent, the better. In the case of content marketing, a team of agents need to work together to execute on a content marketing program, end-to-end. Every piece of content requires multi-agent workflows chained together... it's called multi-agent collaboration.

For effective use of AI agents, it's not about just one agent. It might look like just one, but in reality, but there are a bunch underneath working hard to make your life easier.

Here are a few examples of agents working together for specific tasks:

Brand compliance agent: Evaluate content against a set of compliance guidelines + surface the results and recommendations to the user.

Project manager agent: Ensure projects run on time and budget + can communicate with other agents (eg. Resource Agent that will reassign work if someone is at capacity).

Campaign creator agent: Generate an entire campaign kit, including a quality brief, assigned tasks to execute each campaign element, on-brand content, social posts for promoting content.

Audience insights: AI agents can segment audiences in real-time + recommend content topics that resonate most with each segment.

Streamlined content creation: Dynamic creation of personalized content versions for different audience segments.

Performance and optimization: AI generates detailed insights into what's working and what's not, allowing marketers to refine their strategies.

Experiment advisor agent: Automatically tests different versions of content + implements the best-performing ones.

How to 'thoughtfully' use AI

⚠️ WARNING—there are risks to a haphazard approach to AI, including:

Poor digital experience: Relying too much on AI—without deliberate human intervention—can result in bland and robotic digital experiences.

Too many point solutions: Overloading your martech stack with AI tools will result in high TCO, under-utilized tech, and overwhelmed marketers.

Brand damage: Biased or inaccurate content by AI can pose a challenge to brand governance and compliance.

So, here's what to do instead:

Personalize with empathy: provide AI-powered by human-driven, tailored experiences that are empathetic and acknowledge customer privacy

Require human oversight: ensure that humans have a step in the content workflow to oversee and correct any inaccuracies and biases

Maintain transparency: consider adding disclaimers on content and marketing efforts (eg. powered by Opal) to garner trust with your audience

Get wild, stop being B2Boring

Stop worrying about AI taking your job, and let it take on the sh*tty stuff so you can have more fun with your day-to-days.

The bar is higher now when it comes to better, more personalized content. So, the moral of this story is: be more interesting, be more memorable.