This year, Iāve had over 20 long conversations ā some lasting three hours ā with senior marketers managing annual digital ad spends between $5 million and $20 million. Theyāre seasoned professionals, yet every conversation hits the same nerve:
RoAS from digital advertising isnāt just low. Itās unacceptably low.
Many* digital campaigns are returning only 30 to 70 cents for every dollar invested.
(*Campaigns my firm doesnāt handle. Of course Iād say that, wouldnāt I?)
Why?
Because marketers are increasingly relying on automated tools like Metaās Advantage+ and Googleās Performance Max (PMax).
And the results arenāt pretty.
So why are so many smart people trusting algorithms so blindly? Because the āexpertsā insist they should. Just look at what Mark Zuckerberg said during Metaās earnings call in May 2025:
ā[AI will] improve recommendations, so that any business [...] can just come to us, not have to produce any content, not have to know anything about their customers.ā
He also stated:
āOver the last 5 to 10 years, weāve basically gotten to the point where we effectively discourage businesses from trying to limit the targeting.ā
And:
āWe believe at this point that we are just better at finding the people who are going to resonate with your product ā than you are.ā
Sounds like someoneās been hitting the Kool-Aid pretty hard.
Two quick observations:
1. Meta isnāt even the biggest player ā Alphabet remains #1 globally, with most clients allocating their largest ad budgets to Google.
2. Metaās Advantage+ ā their flagship AI-driven toolset ā consistently underwhelms. For many of our agencyās new clients, audits of their existing work routinely show returns between 30 to 70 cents per dollar spent. Negative returns are commonplace.
Zuckerbergās pitch essentially says, āGive us your cash, and weāll do the restā. But turning over your budget to automation without oversight isnāt strategic ā itās gambling.
Ask yourself honestly: Would you trust anyone encouraging you to take a similar approach with your bank or stock market investments? Press a button, sit back, and you can watch the guaranteed 2x or 3x returns flow in?
Exactly. Me neither.
Algorithms arenāt ready to fully replace us because digital marketing isnāt black and white ā itās nuanced shades of grey. AI still needs human insight to guide it, not vice versa.
Hereās how to realistically aim for a 2x to 3x RoAS:
Step 1: Get uncomfortable
Audit your campaigns honestly. Ask your media team detailed questions: How much budget is in PMaX? Advantage+? What's the actual RoAS for each?
Step 2: Master new thinking
Human intelligence currently outperforms machines in strategic direction. You need to define clearly:
Who are your audiences?
What specific micro-affinities resonate most?
What messaging engages each micro-audience?
Where and When should these messages appear?
Which landing pages will speak clearly to these micro-affinities?
Are you asking the six Ws?
Who are your audiences
What micro-affinities resonate most?
What messaging engages each micro-audience?
Where and
When should these messages appear?
Which landing pages will speak clearly to these micro-affinities?
Step 3: Demonstrate your worth
Be the marketer who can explain what you're doing and why. Algorithms like PMax or Advantage+ are black boxes ā even AI experts canāt fully explain them. CEOs and CFOs respect clarity, not mystery.
Step 4: Move up faster
Blend human insight with AI support when appropriate. Marketers clear on their brandās purpose, who understand precise micro-audience fit, consistently outperform competitors relying solely on automation.
āOrganizations amplify intelligence by integrating human insight with AI precision and robotic efficiency.ā (Accenture, 2024).
āGenerative AI reshapes marketing, but alignment with brand purpose requires human oversight.ā (Gartner, 2024).
āAI transforms targeting and personalization; human marketers craft narratives that resonate emotionally.ā (Forrester, 2024).
āCompanies integrating human-AI collaboration outperform peers relying solely on automation.ā (Accenture Research, 2024).
If youād 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.
Reply