How AI Agents and Custom GPTs Are Transforming Marketing Workflows and Strategy – Dan Sanchez

If you’re like most marketers (or business owners for that matter), you’ve probably noticed AI has gone from hot buzzword to daily headline — and yet, for a lot of folks, it’s still a bit of a black box. That sense of “am I missing out, or is this just hype?” is natural. I’ve been there! So, here’s what I find works for me, and what you can try yourself to start actually using AI in ways that drive, not just talk, real results in your marketing workflow.

How to Approach AI Without the Overwhelm

Let’s be real: AI is more than a fancy optimization. It’s a whole-new-toolbox moment, kind of like social media was a decade back. Most of us struggle at first to see where it fits in — until the practical use cases emerge. It’s easy to get stuck thinking “this is just a faster way to write emails” or “it’ll just automate some menial bits.” But in practice, there’s a smarter approach: take it in baby steps, notch by notch.

I think about it in three key phases:

Faster: Use AI to speed up what you’re already doing
Better: Make your output higher quality, more personalized, and more valuable
Smarter: Actually use AI to level up your own skills and strategic thinking

Let’s break these down, practical-style, and you can start applying them immediately.

Step 1: “Faster” — Let AI Turbocharge the Mundane

Find Your Repetitive Tasks

We’re creatures of habit — in life and at work. So, the fastest wins for AI almost always come from your most repetitive routines. Here’s what’s worked for me:

List out your daily, weekly, monthly, and quarterly tasks. Seriously, just jot them down.
Spot the work that eats up the most time or feels most manual. That’s your “automation gold.”
Break down the steps inside those tasks — where does the data come from? What slows you down?
Ask yourself, “Could AI take the first pass on this step, or even automate it all?”

Speeding Up Analysis, Reporting, and More

A favorite example: weekly survey analysis. I know folks who used to do this by hand—summarizing survey data, copying out responses, writing tailored reports for management. Total energy suck. With AI now, new survey results get summarized, insights extracted, and reports generated automatically, freeing up time for you to actually interpret and use the findings (instead of just shuffling the data).

Processing survey responses: Use AI tools to auto-summarize and pull out key quotes or stats.
Drafting standard reports: Let AI generate first drafts, pulling from raw data, so all you do is polish.
Monitoring routine metrics: Set up AI automations to alert you only if something’s off, not every time.

The big win: When you offload this “busy work” to AI, you’re not replacing the human touch — you’re actually making room for it. That means more time engaging with staff, customers, or strategic thinking. (Pro tip: Don’t automate authentic human conversations. Bots still can’t fake real rapport!)

Step 2: “Better” — Leveling Up Quality with Personalization

Once AI’s handling the heavy lifting, you can start experimenting with making your output not just faster, but better. And the name of the game here is hyper-personalization.

How to Move from Bulk Output to Hyper-Personalization

Let’s say you’re running an email course with video lessons. Instead of sending the same summary to everyone, you can now:

Ask each participant a handful of questions (job title, industry, what they sell, etc.)
Feed both the course transcript and those answers into your CRM/AI assistant
Have AI generate customized, actionable steps for each person — so a CMO at a beauty brand gets different playbooks than a B2B software marketer

The bonus? Once set up, this can run unattended — people keep getting real value, and you get to focus on higher-level projects. It’s set-and-forget that actually serves your audience better than “one size fits all.”

Real-World Applications

Testimonial Integration: Upload all your customer testimonials into a custom AI. Tell it, “Find and use the most relevant client feedback for this blog post, email, or campaign.” Now, you’re speaking your customer’s language — straight from the source.
Dynamic Content: AI can tailor copy, case studies, and even social proof to each segment, using real data — not just generic personas.
Automated Best Practices: Ditch the snoozy, static “best practice” docs. Replace them with living, breathing AI-powered guides that adapt examples and advice for each industry, customer tier, or even user behavior.

By putting in a little work up front, you reap outsized results down the line. One personalized setup continues delivering long after you’ve moved on.

Step 3: “Smarter” — Let AI Coach and Uplevel You

This is when things get exciting. AI isn’t just an assistant — when used right, it becomes a coach or thinking partner that helps you improve faster than ever before.

Ways I’ve Used AI to Get Smarter (and You Can Too):

AI Coaching: Set up a custom AI with your goals, values, and priorities. Ask it to coach you through challenges — one reflective question at a time. It can’t always “read the room” like a human would, but it gets you most of the way there, and it’s available 24/7.
Skill Feedback: Upload a podcast transcript, a sales email, or a draft presentation and request targeted feedback—“How could I tighten this up?” or “What examples would make this clearer?”
Big Picture Strategy: Build a custom GPT (or Gemini Gem, or Cloud Project) that understands your company’s mission, values, and direction. Use it as a sounding board for major decisions or crossroads, always keeping you aligned with your purpose.

Bonus tip: Sometimes just talking things out with AI — articulating the challenge — sparks new ideas or clarity, even if the answer isn’t perfect. It’s a pressure-free mental workout that sharpens your thinking.

Custom GPTs Versus Prompt Libraries: Organize for Scale

One surprisingly useful trick I rely on: custom GPTs for repeat work. Rather than copying and pasting endless prompt libraries, I create specific AI agents/GPTs for each area — client, project, process, or even persona type. Here’s how I break it down:

Custom Instructions at the Account Level: These are the non-negotiables (e.g., always use my writing style, always add bulleted lists, speak in first person). Set them once, and all future GPTs/projects inherit them.
Client/Project-Specific Custom GPTs: Each client or brand gets its own “brand bot,” loaded with their audience, products, goals, and context. No cross-contamination, less time spent repeating yourself, and outputs always stay on-message.
General vs. Specific GPTs: Have one general brand bot with key context, but also create topic-specific GPTs (for things like SEO, social, or product launches) if work is repetitive and unique to that domain.

If you’re collaborating with multiple businesses, teams, or clients — separate GPTs or project bots are the way to keep knowledge clean and outputs consistent.

Final Thoughts: Start Simple, Then Expand

The best way to approach AI right now? Start small, pick a pain point, and let AI make that easier. Once your workflow is humming faster, layer in personalization. Then, push the boundaries and let AI help you (and your business) get smarter, not just more productive.

You don’t have to master it all at once. But by stacking “faster, better, smarter” approaches into your daily routine, you’ll be amazed at how much leverage you get — and how your own role shifts from frazzled tactician to future-ready strategist.

Now it’s your turn to try: What’s one task in your marketing today that could be faster, better, or smarter — with AI lending a hand?

Episode Transcript & Magic Chat

Coming soon.