A few years back, we realized we needed to move houses so my daughters could have more space to grow up. 

We searched for ages, but we couldn’t find the perfect home. So I decided to build one myself. That was when I discovered the joy of power tools. 

One thing I learned was that a nail gun is a whole lot more useful than a hammer. (When it comes to nailing things together.)

But what if your challenges aren’t all nails? No amount of nail guns are going to pour your concrete, or fit your windows. They’d be useless. In fact, they could do a lot more harm than good. 

It’s like that with AI. It’s a massive power tool. 

Use AI in the right way, and it helps. Put blind faith in it and you’re likely to end up (very) disappointed. 

Like when you’re buying performance media.

Google recently pushed out yet another update to its Performance Max (PMax) algorithm. 

Everyone’s all-in – some accounts I’ve seen allocate up to 50% of their search budget to PMax. 

I get it: Less manual tinkering, more ‘set it and forget it.’ 

But there’s an ugly truth: sometimes, often, PMax delivers 50 cents on the dollar.

Why would anyone accept that? Worse, why would anyone invest a significant amount of their budget on that? 

Sure, 50 cents on the dollar might beat 25 cents on a dollar (with little optimization), but it’s still losing you HALF YOUR MONEY. 

Kinda like in life, where we chase flashy shortcuts even when they don’t truly solve our problems. 

Negative returns shouldn’t be acceptable – yet we let it slide. Why? 

Because autopilot feels easier, and we crave anything that makes us look and feel ‘productive’. 

Digital marketing mirrors life: if a quick fix gives us a slightly better result than we expect, we often ignore the bigger underlying issues. 

But is 50 cents on the dollar REALLY THE BEST we can do? Why SETTLE for that? Maybe it’s time we start asking ourselves some hard questions…

Why do we cling to these so-called ‘easy buttons’? 

AI is amazing – I use a custom ChatGPT to help polish this newsletter. 

But note, I still do the research and writing myself. AI augments; it doesn’t replace my critical thinking. 

Same story with PMax. 

We all know it’s a black box - heck, even AI researchers admit they don’t fully understand why these models do what they do. Sound familiar? 

In life, too, we often hand over control to ‘experts’ and just hope for the best. PMax can deliver that quick fix: jump you from an abysmal $0.25 RoAS to a slightly less cringe-worthy $0.50. 

But it stalls out because it’s designed for scale, not surgical precision. 

And life, like marketing, demands nuance. 

We want better than ‘less awful’ – we want real progress. 

So how do we stop letting PMax (and life’s other shortcuts) do our thinking for us, and start regaining control?

Here’s the ruthless-but-loving plan:

Step 1, Step Out of Autopilot – You’re the strategist, not the algorithm. Stop praying PMax will magically fix everything. Work out what makes the things that work for you work. Why do some ads pay back at 1:1 or better?

Step 2, Learn the ‘Whys’ – Don’t focus on your underperforming ads. Focus on what works and why. Think BMW. People buy for performance, design, status...you name it. Can you identify at least 10 ‘why’ buckets for your customers?

Step 3, Focus on Personalization – Not 1:1 for millions of people, but at least segment messages for your top 3? 5? 10? audience clusters. That’s how you move from $0.50 to above $1 RoAS.

Step 4, Own Your Growth – Mastering segmentation and messaging will do more for your career (and sanity) than blindly scaling AI-driven ads.

(By the way, these four steps make up my Return on Marketing Career (RoMC) framework—more on that here.)

Think with Google: 87% of marketers reported initial performance lifts from automation tools like PMax, but 42% felt uneasy about the lack of control.

McKinsey’s Next in Personalization Report: Companies that excel at personalization can boost marketing ROI by 15% or more – reinforcing the power of granular audience targeting.

Harvard Business Review: Warned about AI’s ‘black box’, cautioning that blind trust in algorithms can undermine strategic thinking and innovation.

Epsilon: Found that 80% of consumers are more likely to purchase from brands offering personalized experiences – highlighting the human touch that AI alone can’t replicate.

If you’d like to discuss your career journey with me one-to-one, please feel free to email me at [email protected] or message me on LinkedIn.

Thank you for reading.

If you know someone who you think would appreciate this newsletter, please forward it to them.