- Marketing Herald
- Posts
- Your Agency Is lying
Good Morning
Table of Contents
Marketing News
Any agency pitching a "proprietary process" for Meta or Google ads is lying.
Let me save you a sales call:
Before platform automation took over, you could genuinely create outlier performance through technical expertise:
- Custom audience structuring
- Advanced exclusion rules
- Strong bid optimization processes
- Sophisticated account architecture
That era is dead.
In 2025, Facebook and Google have simplified everything.
The platforms ignore most audience exclusions, auto-expand audience targeting, and do a whole host of other behind-the-scenes optimization tricks you can't control.
When an agency claims they have some "proven system" for these walled gardens, they're likely selling you outdated tactics that no longer work.
Whether that's because they still believe they will work or because they think it makes a better sales pitch is anybody's guess.
The real expertise now is much more nuanced.
You need someone with a solid technical background, who understands marketing at a high level, who is also able to interpret your specific business model and how it should be reflected in your ad accounts.
There are no "media buying secrets" anymore. Anyone claiming otherwise is trying to extract a 15% fee for work that's become mostly automated.
Most brands kill their best ads too soon.
They launch, see results, then move on, thinking they need something new.
But the truth is, scalable ads aren’t created, they’re engineered.
Here’s what makes an ad truly scalable:
- It runs for 60+ days without fatigue
- It consistently ranks among top 10 spenders in the account
- It’s duplicated across multiple campaigns, concept vehicles and audiences
We don’t start from scratch every time. We build on what’s already working.
For example, when scaling Nothing Fishy, we took a static ad that already performed well and:
- Create 10 variations of the winning hook (“Did you know 80% of people lack Omega 3?”)
- Created 10 variations of the subject in the image (UGC shot with a PR headline).
No guesswork. No wasted spend. Just strategic iteration.
Scaling isn’t about making more ads. It’s about making better versions of the ones that already work.
If your ads keep dying after a few weeks, it’s not the algorithm, it’s your approach.
Ad frequency is one of the most misunderstood metrics.
Some reasons why:
❌ Meta Advertisers are looking at frequency numbers at the campaign or ad set level instead of the ad level. The ad level may show lower frequency numbers for each ad when there are multiple ads within one ad set.
❌ Meta Advertisers forget about the fact that an ad impression is already measured from the moment an ad appears on screen (greater than 0 pixel and greater than 0 seconds). The user may not have actually seen it, still it's counted by Meta.
❌ Meta Advertisers forget about how 'scroll speed' and 'attention span' impact a real 'ad impression'. The user may not have actually noticed it or paid attention to it.
❌ When working with dynamic ad content, the content itself varies, making exact ad frequency lower.
Would you spend $1M on a channel you can't track properly?
After running hundreds of incrementality tests, my answer is yes.
Most marketers and especially CFOs hesitate when faced with investing in channels without clear attribution.
The pressure to demonstrate ROI pushes teams toward platforms with clean tracking, precise metrics, and impressive-looking dashboards.
This is fundamentally flawed thinking.
The core issue with modern marketing measurement is that channels with the cleanest attribution are often the least incremental to your business. They excel at taking credit for conversions but struggle to prove they actually caused those conversions.
In my experience, channels like connected TV and traditional television deliver some of the highest incrementality despite having no direct click mechanism.
When these channels are properly tested through geo-based experiments, the results challenge so-called "conventional" performance marketing wisdom.
And this "phenomenon" is extremely consistent with consumer behavior:
People don't watch TV with high purchase intent.
But they do absorb your message, develop awareness, and later, when they're ready to buy, they search for your product.
At that point, Google captures them through branded or category search and claims 100% of the credit - despite not initiating the customer journey.
This creates a problematic cycle.
Marketing teams shift budgets toward platforms showing high attribution metrics (typically Paid Search & Paid Social), reducing investment in upper-funnel channels that actually drive demand.
Performance temporarily improves as you capture existing demand more efficiently, then plateaus or declines as you've exhausted the audience that already knew about you.
The platforms themselves aren't being dishonest - they're simply operating within their limited view of the customer journey.
If the goal is genuine business growth rather than impressive-looking dashboards, incrementality must become the primary metric.
It answers the essential question:
"Would this sale have happened without this marketing channel?"
For brands ready to move beyond attribution limitations:
Dedicate a significant portion of your budget to channels that drive actual demand, even when their performance looks suboptimal in traditional metrics.
Test everything, trust the incrementality data, and invest in channels that drive real results...even (and sometimes especially) when those results are harder to track.
The brands that understand this principle are systematically outperforming their competitors and will continue to do so.