Meta Ads just changed forever (and no one is talking about it)

SUMMARY

Charlie, an ad expert with over a billion dollars spent on Meta ads, explains incremental attribution—a new feature optimizing for new customers to scale brands—critiquing common mistakes and outlining adoption timing.

STATEMENTS

Exclusions in advertising block existing customers from ads, raising CPMs and wasting money by limiting warm traffic without truly targeting new buyers.
Launching endless new ads attracts colder, less profitable customers, increases costs, and creates a cycle of fatigue and lower LTV.
Constant offer testing disrupts customer journeys, chases short-term sales, and makes repeat purchases harder, undermining long-term growth.
Incremental attribution optimizes ads for people who wouldn't have bought otherwise, making outdated tactics like exclusions and endless ads obsolete.
Normal attribution trains algorithms for general purchases, while incremental attribution builds a system for scalable new customer acquisition, like upgrading from a horse to a car.
Optimizing for new customers initially worsens metrics like ROAS and CPA, but prioritizes LTV from repeats over first-sale profits.
First-time buyers are more likely to return than repeat buyers, making new customer focus essential for future cash flow.
Tools like Ammera allow direct tracking of new customer events in Meta, providing a cleaner signal than current incremental attribution.
Incremental attribution leverages Meta's vast historical data via Andromeda, surpassing custom event limitations for broader insights.
Spring 2026 is ideal for adopting incremental attribution, after heavy spenders train the algorithm during holidays, aligning with low CAC periods.

IDEAS

Exclusions seem smart for new customer targeting but actually inflate costs by forcing algorithms into inefficient patterns.
Endless ad launches create a vicious cycle where colder audiences lead to higher expenses and diminishing returns on customer quality.
Offer testing fragments optimization efforts, turning a streamlined funnel into chaotic paths that erode repeat business value.
Incremental attribution redefines success by focusing on "would not have bought otherwise," shifting from reactive to predictive ad delivery.
Attribution is often misused as mere reporting, rewarding superficial metrics that mask true growth barriers.
First purchases are the least profitable, yet optimizing for them directly builds the repeat sales that drive real profitability.
New buyers have untapped LTV potential, making them more likely to convert repeatedly than saturated repeat audiences.
Direct new customer tracking via tools like Ammera offers immediate clarity, outpacing Meta's nascent incremental feature.
Meta's machine learning needs "tuition" from big budgets to mature incremental attribution into a dominant tool.
Timing ad strategies around seasons—loading funnels in spring for summer repeats—amplifies profitability during high-CAC periods.
Incremental attribution taps into platform-wide historical events, democratizing advanced data beyond individual business limits.
Abandoning incremental attribution too soon risks missing its exponential scaling potential once the algorithm learns.

INSIGHTS

True ad optimization demands measuring incremental impact over vanity metrics, fostering sustainable growth by prioritizing untapped audiences.
Common tactics like exclusions erode efficiency by constraining algorithms, revealing how perceived cleverness often accelerates financial waste.
Shifting focus from first-sale volume to new customer acquisition unlocks higher LTV, as initial buyers hold the greatest repeat potential.
Machine learning in advertising evolves through data "education," where early adopters subsidize breakthroughs for later, widespread dominance.
Seasonal timing in customer acquisition turns market fluctuations into strategic advantages, balancing low-cost entry with high-margin repeats.
Attribution's core value lies in teaching systems what drives genuine expansion, avoiding the trap of inflated short-term KPIs.

QUOTES

"Bring me people who would not have bought otherwise."
"You don't finish a marathon because of the cup of water at mile 25, but that's how most marketers assign credit."
"The first purchase is always the least profitable transaction you will ever get."
"First-time buyers are far more likely to come back and buy again than people who've already bought two or three times."
"Optimizing for new customers will look worse today, but it's the only way to build the cash flow that fuels your tomorrow."

HABITS

Track new customer events directly using tools like Ammera to provide clear optimization signals in Meta campaigns.
Avoid exclusions by letting Meta's algorithm naturally filter ad exposure to interested audiences, maintaining cost efficiency.
Limit endless ad launches by focusing on quality creative that sustains engagement without accelerating fatigue.
Conduct offer testing sparingly, prioritizing core products to preserve streamlined customer journeys and predictable cash flow.
Time campaigns seasonally, acquiring new customers in low-CAC periods like spring to capitalize on high-margin repeats later.

FACTS

Meta's business model inherently filters ad delivery to avoid overexposure to existing customers who don't want to see them.
Advertisers spending millions on ads are currently "paying tuition" to train Meta's incremental attribution algorithm through trial and error.
Spring is the cheapest time of year for customer acquisition costs, while summer sees the highest spikes in CAC.
First-time buyers exhibit higher return rates than those with multiple prior purchases due to unmatured LTV.
Incremental attribution, powered by Andromeda, accesses every historical event on the platform for consistent updates.

REFERENCES

Ammera (tool for passing new customer events to Meta).
Barry Hott (quoted on marathon attribution analogy).
Andromeda (Meta's system powering incremental attribution with historical data).

HOW TO APPLY

Integrate a tool like Ammera to split purchase events into first-time (red) and repeat (yellow) for direct new customer optimization in Meta.
Set up custom events in your Meta pixel to track and prioritize new customer conversions over general purchases.
Monitor initial metric dips like rising CPA when shifting to new customer focus, viewing them as signals of improved long-term LTV.
Begin testing incremental attribution with small budgets in late 2025 to familiarize the system before full adoption.
Align campaigns with seasonal timing: acquire aggressively in spring, nurture repeats through summer, and scale into holidays.

ONE-SENTENCE TAKEAWAY

Optimize ads for new customers using incremental attribution by spring 2026 to scale sustainably beyond outdated tactics.

RECOMMENDATIONS

Prioritize direct new customer tracking tools now for cleaner signals until Meta's incremental feature matures.
Abandon exclusions immediately to prevent cost inflation and preserve algorithmic efficiency.
Focus creative efforts on sustaining high-quality audiences rather than flooding with new ads.
Build seasonal strategies that load acquisition funnels in low-cost periods for profitable repeats later.
Measure success by new customer additions, not short-term sales volume, to capture true LTV potential.

MEMO

In the high-stakes world of digital advertising, where billions are wagered on fleeting clicks, Meta's recent rollout of incremental attribution emerges as a quiet revolution. Charlie, a veteran ad strategist who has orchestrated over a billion dollars in Meta campaigns, warns that this feature could upend how brands chase growth. Long plagued by tactics that promise scale but deliver chaos, advertisers often resort to exclusions—barring existing customers from ads in a misguided bid for freshness. Yet, as Charlie explains, this only burdens Meta's algorithms, spiking costs without genuine innovation.

The pitfalls run deeper. Flooding platforms with endless new creatives might snag short-term wins, but it lures in colder, less loyal audiences, eroding lifetime value and inflating expenses across the board. Similarly, relentless offer testing—pitching bundles and novelties—fragments customer paths, turning reliable funnels into unpredictable mazes that prioritize one-off sales over enduring loyalty. Charlie likens traditional attribution to training a horse for speed: reliable, but limited. Incremental attribution, by contrast, is building a car—clunky at first, but poised for transformative velocity by targeting those who wouldn't have converted otherwise.

At its heart, this shift reframes attribution not as a scorecard of past sales, but as a teacher for future expansion. Borrowing a marathon metaphor from marketer Barry Hott, Charlie notes that crediting a single aid station for victory misses the race's essence. Early adopters may see dips in return on ad spend and rising costs, but these are illusions; the real prize lies in first-time buyers, whose untapped potential for repeats far outstrips saturated customers. Tools like Ammera already enable precise tracking of new versus repeat purchases, offering a bridge until Meta's machine learning, fueled by big-budget "tuition," fully harnesses the feature's power through Andromeda's vast data trove.

Timing, Charlie insists, is everything. With heavy spenders set to refine the system over 2025's holidays, spring 2026 beckons as the sweet spot—when acquisition costs plummet and algorithms peak. Brands that preload their funnels then can weather summer's cost surges, banking on profitable second and third sales. This isn't mere optimization; it's a blueprint for resilience in an era where ad fatigue and economic flux demand smarter plays.

As Meta evolves, Charlie urges a pivot from vanity metrics to meaningful metrics: not sales tallied, but customers cultivated. For disruptors eyeing six-figure daily spends, the path forward involves ditching outdated hacks, embracing data-driven precision, and riding the wave of machine learning's maturation. In this arena, ignoring incremental attribution isn't just risky—it's a fast track to obsolescence.