Four Hard Truths About the Future of Paid Media

The ad industry is changing quickly enough that a playbook which worked six months ago carries no guarantee now. Four shifts every operator, buyer, and founder should be planning around.

1. The biggest way agencies waste budget is running campaigns that are not incremental

This is the most common and most expensive failure, and it hides in plain sight: campaigns that look excellent in Ads Manager while creating no new revenue.

The specific patterns we find week after week:

• Brands with 40% of spend on branded terms in PMax
• Retargeting mislabeled as prospecting
• Suppression lists that do not actually exclude anyone

Each of these recycles existing customers and takes credit for conversions that would have happened anyway. None of them look like errors on a dashboard — they look like strong performance.

The fix:

• Build true new-customer campaigns with clean suppression
• Audit branded spend, especially on Google
• Measure incrementality rather than ROAS

If your platform numbers look strong but total revenue is not moving, this is usually why. The measurement side of the problem is covered in attribution after last click, and the suppression-leak mechanics in beyond blended ROAS.

2. The biggest lie is that what worked once will work everywhere

Too many teams walk into a new account and paste the playbook from their last win.

Scaling is not a template. It is timing, context, and feel. Every brand has its own power curve, audience behavior, and seasonality, and a structure that succeeded elsewhere encodes assumptions about economics that may not transfer at all.

Running ads is like flying a plane: you do not start with stunts and barrel rolls. You learn the controls first.

There is a second failure that follows the first. Even when teams find something that works, they coast — hit target ROAS, stop testing, and settle. The best operators stay curious, test continuously, share learnings across accounts, and incentivize buyers to push past the benchmark rather than park on it.

Scaling is not about finding one system. It is about knowing when to break it.

3. Thumb-Stop Rate is the leading indicator most agencies ignore

There is one metric I check before anything else when reviewing creative performance.

Thumb-Stop Rate = 3-second video views ÷ impressions.

It measures how often your ad makes someone pause. Most agencies focus on ROAS, CPA, or CTR — all lagging indicators that describe what happened after attention was already captured. Thumb-Stop Rate is a leading indicator: it tells you whether the creative is strong enough to earn attention at all.

The benchmarks we work against:

If your ads do not make people stop scrolling, nothing downstream matters. This pairs with hold rate and completion rate as the closest available proxies for whether creative is doing brand work as well as conversion work — see the memory economy.

4. AI creative buys speed and can cost you trust

AI-generated ads are everywhere. They are polished, fast, and increasingly repetitive.

What we observe in accounts: older audiences respond well to AI creative because they often cannot tell the difference. Younger audiences, Gen Z especially, identify it instantly — and trust drops the moment they do. That asymmetry means the same asset can perform acceptably against one cohort and actively damage credibility with another.

Behind that sits the Dead Internet Effect: as AI learns from AI, quality degrades. The more automated content floods the web, the harder genuine differentiation becomes.

Meta is already responding by identifying duplicate ad variations and classifying them as the same entity, which forces advertisers toward creative diversity rather than raw volume. The mechanics of that are in creative pipeline math.

Long term, two parallel tracks emerge:

1. Automation at scale — platforms generating dynamic ads tailored per impression.
2. Authenticity at scale — storytelling, creators, and memorable brand moments separating real from noise.

AI can automate output. It cannot automate meaning.

What thriving looks like

• Focus spend on incremental growth, not recycled revenue
• Stay curious, and never coast on a benchmark
• Build creative that earns attention before it asks for a sale
• Balance automation with authenticity

In a market saturated with automated content, real strategy and creative instinct are the durable moats.

FAQ

What is the most common way ad budget gets wasted?

Running campaigns that are not incremental — spending against people who would have converted anyway. The three patterns we see most: heavy branded-term spend in PMax, retargeting mislabeled as prospecting, and suppression lists that fail to exclude existing customers. All three report as strong performance while contributing no new revenue, which is what makes them persistent.

What is Thumb-Stop Rate and what is a good benchmark?

Thumb-Stop Rate is 3-second video views divided by impressions — how often your ad makes someone pause. Above 30% indicates strong creative, 20% or better is healthy, and below 10% means the ads are not connecting. It is valuable because it is a leading indicator: ROAS, CPA, and CTR all describe what happened after attention was won, while this measures whether attention was won at all.

Does AI-generated ad creative actually work?

It depends on the audience. Older cohorts respond reasonably well because the difference is not obvious to them. Gen Z tends to identify AI creative immediately, and trust declines once they do — so the same asset can be efficient against one segment and harmful to credibility with another. Speed of production is real; the risk is repetitive output in a feed already saturated with it.

Why do agency playbooks fail on new accounts?

Because a playbook encodes the economics of the account it was built on — its acquisition costs, payback window, seasonality, and audience behavior. Transplanted to a brand with different constraints, the same structure produces different results. Process is worth respecting; the skill is recognizing when the context has changed enough to break it.

What is the Dead Internet Effect in advertising?

The degradation that follows when AI systems increasingly train on AI-generated content, lowering overall quality and making differentiation harder. In advertising it shows up as feeds full of competent, interchangeable creative — and as platforms responding by grouping near-identical ad variations as a single entity, which penalizes volume without diversity.