new! Please listen to the article
I meet with marketers every month in a small group called “Marketing Therapy” and as an outlet who can speak openly about the challenges we face in our work. We talk about leadership, strategy and industry change. But it's not surprising. Recently, everyone wants to talk about AI.
All marketers in these groups are already using AI in some way. Usually they chat with at least one of the popular LLMs as a kind of collaborator.
It's great to learn while people push themselves, try something new and try and move. Unfortunately, all the motivations behind it seem like they are afraid to fall behind. I'm not confident in the opportunities that AI presents.
How can I know that? Most members cannot clarify their AI strategies. Instead, they are responding to their demands of “more AI” or their perception that others are doing it more. As a result, there is a lot of use of AI's “spaghetti” method, which may seem innovative and advances, but does not actually move the needle.
To quote Sun Tzu, “Tactics without strategy are pre-defeat noise.”
I don't pretend I didn't fall into the same trap as myself. Also, there is no path to AI-Adoption maturity that works for all marketing leaders. But I can tell you how to find a way for AI to work for you without spending time spinning your wheels.
The cleverest thing you can do now is to be slower. Think critically about where AI fits your goals and make a plan before you start throwing all the prompts you hear in your chosen LLM.
Why is there a gap between AI adoption and value?
The AI was promoted and adopted incredibly quickly. Many people use it.
But most of those people simply chat with LLM, so far not many have gained a lot of business value. Some people have come across the “trough of disillusionment” stage in Gartner's hype cycle. Their interest is waning as AI fails to fulfill its (very high) expectations.
If you reach this point, consider acknowledging my verification of your feelings. Please continue now.
Using AI in your marketing workflows has meaningful incremental value. I see it in person every day. But if you haven't seen its worth yet, it's normal. You haven't found your way yet.
Currently, only 30% of brands and agents have fully integrated AI into the entire media campaign lifecycle. Half of those who haven't said they'll achieve full integration by next year have no strategic roadmap to get there.
However, adding AI to all existing workflows to the campaign lifecycle is not enough to get past experiments and durable value.
Three changes need to be planned to create a real impact.
- Employ AI throughout the campaign lifecycle.
- We embrace AI as a collaborative strategic partner, not just a glossy version of the legacy tool.
- A transition from reactive execution to a real-time, future-oriented mindset.
These are not minor adjustments. They are fundamental changes that require intentional planning and long-term commitment, and to lead you where you want to be.
The AI roadmap starts with goals
Starting with tech, it's not a problem or a desirable business outcome, but a big mistake people make.
Before inserting AI at any stage, you need to know what success looks like. This allows you to map your current processes and goals to the best AI tools for the results you want to achieve.
To build momentum and confidence in your Ai-Adoption journey, start with ideas that are safe to test, but also have strategies behind them. Don't set goals that require a complete rewiring of your tech stack, or set goals that require you to invest a huge amount of time or money to develop.
Below are two examples of what a realistic goal might look like.
- Increase the number of creative tests you run weekly using Generated AI by 20% to iterate and launch approved creatives quickly.
- Using cross-platform bidding and budget optimization AI, it reduces by 10% compared to Meta and Google baselines.
So, once you have identified your goals, how can you plan?
1. Mapping the Complete AI Campaign Lifecycle Integration
To truly optimize your ads, AI must be adopted throughout the entire campaign lifecycle, rather than isolated pockets.
As you narrow down the goals you want to achieve throughout your entire campaign, you will find that you cannot rely on one type of AI to do everything. Any type of AI plays a clear role and using the wrong AI, or just one type, leaves a big gap in your strategy.
Take GroupM as a practical example.
GroupM noticed bid and budget inefficiencies in many client campaigns, but could not use just one solution to fix them all. Instead, machine learning was used to analyze customer behavior to uncover trends and segment audiences more effectively. These insights have led the team to use AI to fine-tune media spending in real time, maximizing each client's budget and fostering stronger engagement between segments.
The possibilities for AI in GroupM's campaign optimization are unstoppable. Generic AI allows you to turn raw creatives into hypersegmental creatives tailored to the audience's grander understanding. Once the entire process is dialed in, Agent AI can act on several steps without the need for manual intervention.
From insights to execution, it is not actually a step-by-step campaign lifecycle where AI can't streamline and expand its efforts.
2. Rather than a quick car, we treat AI as a collaborator
The true value realized from AI comes from when it makes us more keen, creative and more strategic. Don't think of it as a partner as a tool that uses AI (especially LLM).
Imagine telling ChatGpt that your brand's meta ads are performing poorly. The prompt gives the AI a dataset that shows where the metrics are flat, and asks how to troubleshoot the problem. You'll probably get a response about where to move your budget, but that may not be correct and may not be what you need.
Meanwhile, engage in conversations with AI and provide more context (techniques used, rotation creatives, past campaigns were winners), and develop plans to coordinate issues, refine theories, and adjust them. Here, AI identifies the root of the problem. This is creative fatigue. Additionally, the action will be taken to generate new creatives in the A/b test and see how they resonate.
The hack I want to use is the Socrates method. Not only should you give your AI an answer, but rather tell them to ask questions that will help you clarify your answer. You are still in the driver's seat, so we can guarantee you will leave with better insights.
3. Go from reactive to proactive
For years, marketing AI has been equated with glory automation, and we have seen the ideas flow in how marketers choose today's AI tools. Think of it like this: if you need to manually fix the “AI” every time something changes, it's not AI. And the reactive approach of adjusting your approach every time is not sustainable.
Winning brands are brands that focus on taking advantage of opportunities, and their tech stack needs to do the same. Many AI tools use probabilistic and causal models trained with marketing-specific data, making them suitable to assist with predictions and preemptive changes that otherwise remain flat.
You cannot design your playbook for every event, including tariffs, supply chain shocks, or new rival models. However, you can always predict new potential outcomes to design real-time learning systems and drive aggressive marketing strategies prepared for resilience.
Migration from AI-enabled to AI-Native
We often ask digital marketers how we can realize how to fast the value of AI. It is possible and always happens. But speed None Strategies only lead to failure.
In many cases, AI is making shoe holes in marketing efforts. It's not going well.
Clearly, the roadmap is the path to goals, not the ultimate goal itself. Treat it as a living experiment and use what you learned along the way to bring what you learned closer to your purpose and maintain accountability.
Even if you focus on the outcome, your biggest concern should not fail. Because everyone fails. This is a new realm for all of us. However, those who fail will also have the opportunity to learn and do better next time.
Put the fear of failure aside, focus on today's achievable outcomes, and learn from them so that you can tackle your more ambitious goals than tomorrow.
More resources on AI adoption and marketing use
“Human-friendly marketing”: The power of human synergy
Why Marketers Don't Wait for Perfect AI: Lessons from Apple's Lag
AI belongs to the Marketing Toolkit, but saves space for humanity.
Navigating AI adoption and marketing use: A strategic approach
How CMOS uses AI to create a career-changing strategic business impact
