In this episode of SEO’s Getting Coffee, Emina and Sean are joined by Azahara Corrales, AI advocate, international speaker, and marketing leader, to unpack what AI is really doing to our day-to-day workflows and what it definitely isn’t.
Because once you strip away the hype, the LinkedIn hot takes, and the “I automated my entire agency in a weekend” posts, the truth is far more nuanced.
AI isn’t replacing marketers.
It isn’t magically fixing broken processes.
And it definitely isn’t saving you time if you don’t know what problem you’re solving.
But used well? It can change how work feels and that matters more than most people realise.
Watch Video
Automating Smarter: AI’s Role in Marketing | SEOs Getting Coffee Ep.45
The Two-Speed Reality of AI Adoption
One of the most grounded observations in this conversation is that AI adoption is not moving at a single pace.
Inside digital marketing, it’s easy to assume “everyone is using AI now.” But that’s simply not true. What we’re seeing instead is a widening gap:
- Teams and individuals who are fully embedding AI into daily workflows
- And others who are barely touching it beyond occasional experimentation
Azahara points out that Large Language Models (LLMs) like ChatGPT are almost always the entry point. From there, usage tends to follow a predictable path:
- LLMs for everyday questions and drafting
- Custom GPTs trained on internal data
- Automation and agents for repetitive tasks
- (Eventually) more complex, interconnected systems
The danger is assuming step three is where you start, instead of where you earn the right to be.
Start With Friction, Not Possibility
If there’s one practical takeaway from this episode, it’s this:
Don’t ask “What can AI do?” Ask “What is wasting my time?”
Azahara’s first meaningful automations didn’t come from experimentation for experimentation’s sake. They came from frustration.
- Reporting.
- Benchmarking.
- Copying numbers between spreadsheets and decks.
- Tasks that are necessary, repetitive, and deeply unfulfilling.
These are the workflows AI is genuinely good at supporting, especially when paired with clean data and consistent structure.
And crucially, this isn’t about chasing novelty. It’s about reducing cognitive load so energy can be spent where human judgment really matters.
Automation Doesn’t Remove Responsibility (It Increases It)
A recurring theme in the discussion is quality assurance, and why automation actually raises the bar rather than lowering it.
LLMs hallucinate.
They fill gaps confidently.
They optimise for plausibility, not truth.
Which means:
- Problem definition matters more than ever
- QA can’t be optional
- Blind trust is not a workflow strategy
As Emina points out, many failed automations aren’t AI failures; they’re thinking failures. Automating a poorly understood task just gives you a faster way to be wrong.
AI doesn’t remove the need for expertise.
It exposes the lack of it.
Data Privacy, Responsibility, and the Black Box Problem
Another standout conversation in the episode is about data privacy and responsible AI use.
The temptation to upload everything into a model is real, especially when the outputs are useful. But that doesn’t mean it’s safe, appropriate, or future-proof.
Even when using enterprise tools, the principles remain the same:
- Share only what you need
- Assume you don’t fully control where data may surface
- Put governance in place before something goes wrong
This is one of the quieter but more important parts of AI maturity, often skipped in favour of speed, using it responsibly and reducing AI waste is vital.
Is AI Making Us Lazy, or Just Less Burnt Out?
Perhaps the most interesting part of the conversation centres on this uncomfortable question.
Is AI making marketers worse thinkers?
Or is it simply helping them survive increased pressure, higher workloads, and shrinking margins?
The answer, unsurprisingly, is both, depending on how it’s used.
AI can flatten curiosity if it replaces thinking.
But it can also preserve energy when thinking isn’t the bottleneck.
In an industry where “do more with less” has become the default expectation, AI often isn’t about optimisation; it’s about endurance.
The Future: Agentic Teams or Organised Chaos?
The episode closes with a glimpse into where things might be heading: agentic AI systems, specialised agents, and automated coordination across disciplines.
Exciting? Yes.
Unsettling? Also yes.
The idea that AI could one day handle cross-channel alignment, performance synthesis, and internal communication is compelling, particularly for smaller teams.
But as with everything else discussed, the same rule applies:
Just because something is possible doesn’t mean it’s desirable.
Human oversight, context, and ethical judgment don’t disappear just because workflows evolve.
Room 404: Big Tech Running AI Education
For Room 404, Azahara chose something more serious than a bad marketing trend.
She wants to banish the idea that Big Tech companies should be shaping how AI is taught to children.
With Meta, Google and others increasingly involved in AI education conversations, her concern is simple: the organisations building and monetising these systems shouldn’t be the ones deciding how they’re introduced in schools.
AI literacy matters, but who controls the curriculum matters even more.
Straight into Room 404.
Final Thoughts
This episode isn’t about selling AI as a silver bullet. It’s about using it deliberately, sceptically, and with a clear understanding of where its value actually lies.
AI won’t fix broken processes.
It won’t replace strategic thinking.
And it won’t absolve us of responsibility.
But when used with intent, it can make work calmer, cleaner, and more humane, and that’s a future worth building toward.
If you’re experimenting with AI in your own workflows, we’d love to hear what’s genuinely helped and what absolutely hasn’t.
For more insights and in-depth conversations on the latest in SEO and digital marketing strategies, stay stuned for upcoming episodes of “SEOs Getting Coffee.” Subscribe to our channel for regular updates and expert opinions.
