ANA Mandate: Marketers Need AI Skills in 2026

The American National Advertising (ANA) Association is telling marketers to build real AI skills, because the future of the job depends on it. This means you have to get your hands dirty with active, practical application. The real question is how you can actually plug AI into your daily work to get results you can measure.

1. Master Prompt Engineering for AI Content Generation

Good AI results start with good prompt engineering. Too many people toss vague requests at tools like DALL-E 3 or Midjourney and then wonder why the output is generic. You have to be specific. Treat the AI like a smart intern who takes everything you say literally. It needs detailed instructions to give you anything worth using. When you’re trying to generate marketing copy, for instance, you need to specify the target audience, the exact tone of voice, the key message, what the call to action is, and even negative keywords for words you want it to avoid. For an email campaign announcing a new SaaS feature, a prompt might look something like this: “Generate three subject line options for an email announcing ‘Project Zenith,’ a new AI-powered analytics dashboard. The target audience is B2B marketing managers in tech companies. Tone: professional, exciting, benefit-driven. Focus on increased efficiency and data-driven insights. Avoid jargon like ‘teamwork’ or ‘sea change.’ Include a clear call to action to ‘Request a Demo’ in the subject.” That level of detail pushes the AI toward a far more usable first draft, which saves you a ton of time on revisions.

Pro Tip: Iterative Prompt Refinement

You won’t get it right on the first prompt. Think of it as a conversation. Start broad, see what the AI spits out, and then tighten up your prompt based on what worked and what didn’t. This back-and-forth trains both you and the model.

Common Mistake: Over-reliance on Default Settings

It’s easy to just use the default personas or styles built into AI tools, but that’s how you get bland content that sounds like everyone else’s. You have to customize the output every time to match your brand’s voice and the specific goals of that campaign.

2. Integrate AI for Data Analysis and Predictive Insights

This is where AI gets really powerful: processing huge amounts of data to find patterns a human analyst would miss. We’re seeing tools like Google Analytics 4 (GA4) build AI insights right into the platform, but the trick is knowing how to read them and what to do next. A practical way to start is by setting up GA4’s predictive metrics. Go to “Reports,” then “Life cycle,” and into “Monetization,” where you can find metrics like “Purchase Probability” or “Churn Probability.” You can actually use this by creating a custom audience of users who have a high purchase probability but haven’t bought anything in the last 7 days. Then you can export that audience directly to Google Ads and hit them with a targeted remarketing campaign. This lets you engage people before they disappear for good and can seriously improve your conversion rates. A 2023 eMarketer report found that companies using AI for this kind of predictive work saw their campaign ROI jump by an average of 15%.

Pro Tip: Don’t Just Look at the Numbers, Understand the ‘Why’

AI will tell you *what* is happening, but it’s your job to figure out *why*. If GA4 flags a segment as a high churn risk, your job is to dig into their user journey and find the friction point. Is a page broken? Was a product update a flop? The AI points to the fire. You have to find the source of the smoke.

Common Mistake: Ignoring Data Bias

An AI model just reflects the data it was trained on. So if your historical data is skewed (maybe it’s all from one demographic), your AI’s predictions will be just as skewed. You have to audit your data sources constantly and look into data augmentation to get fairer, more accurate results.

3. Implement AI-Driven Personalization at Scale

Customers expect personalization. It’s table stakes now. With AI, you can deliver these hyper-personalized experiences across every touchpoint without a massive team doing it all by hand. Imagine a visitor clicking through specific product categories on your e-commerce site. An AI recommendation engine, like the kind you find in Adobe Sensei, can instantly show them related products or trigger a tailored follow-up email. To get this running, you’d typically connect your CRM to a personalization platform and start defining rules based on user behavior (like “viewed product X” or “added to cart but didn’t buy”). The AI then takes over, changing content and product recommendations in real time. If a user keeps looking at high-end electronics, for example, the AI will make sure they see your premium product banners on the homepage instead of the entry-level stuff. To make this work, marketers have to understand the logic behind these AI engines and how to tweak the settings for the best results.

Pro Tip: Test and Segment Your Personalization Efforts

Don’t stop A/B testing just because you’re using an AI. You still need to test your personalization strategies on different audience segments to make sure they’re actually helping. What works for your power users might completely turn off new visitors.

Common Mistake: Creepy Personalization

There’s a line between helpful and just plain creepy. Be careful not to use data that feels too personal or makes people uncomfortable. The point is to deliver value, not to show off how much data you can collect.

4. Develop an AI Governance Framework

As you weave more AI into your operations, you absolutely need clear governance. This is about building trust and using AI ethically, which goes way beyond just checking a compliance box. A good framework needs clear policies for how you handle data going into AI models. What customer data is allowed? How do you anonymize or protect sensitive info? Then, you need a review process for anything the AI generates, which probably means a human has to look at it before it goes live, especially for important communications. You also have to think about the ethics of the output. Is the AI reinforcing old biases? Is it making deepfakes without telling anyone? The IAB’s AI Ethics in Advertising Framework is a decent place to start. I always recommend putting together a cross-functional AI ethics committee with people from legal, marketing, and data science to keep these policies current.

Pro Tip: Document Everything

Document everything. Keep logs of your AI models, the data you’re using, the parameters for training, and every time a human has to step in. When it’s time for an audit or something breaks, that documentation will be a lifesaver.

Common Mistake: Treating AI as a Black Box

Don’t treat AI like a magic black box. You need to understand, at least basically, how it’s getting its answers. Push your vendors for more transparency and get your team educated on the fundamentals of how these models make decisions.

5. Foster a Culture of Continuous AI Learning

The AI field moves fast. What’s new today is old news in 18 months. That means you and your team have to be learning constantly, setting aside time every week to develop skills and try out new tools. Encourage your team to get certifications from good sources. For example, the HubSpot Academy’s AI for Marketing course gives a solid base in how to apply this stuff. Go beyond formal courses, too. Set up internal “sandboxes” where your marketers can mess around with new tools without worrying about breaking a live campaign. I’ve seen teams completely change their game by dedicating just one hour a week to exploring a new AI feature and then sharing what they found. That shared knowledge is the best way to future-proof your team.

Pro Tip: Focus on Application, Not Just Theory

Theory is fine, but you learn this stuff by doing. Get your marketers to apply what they’re learning to small, real-world tasks. The hands-on work is where the knowledge actually sticks.

Common Mistake: One-and-Done Training

A one-off training day is useless. Learning these skills is an ongoing process. You need a real, structured program for continuous development, not just a single workshop. Building AI skills for marketers means changing how you think about strategy, creativity, and getting work done, and it requires a real commitment to constant learning and ethical guardrails.