Stop Targeting: Why Broad Wins on Meta Ads
Here's a take that makes most media buyers uncomfortable: the best way to target your ideal customer on Meta is to not target at all.
Broad targeting (running ads with zero interest filters, no lookalikes, just age, gender, and country) consistently outperforms detailed audience stacks for ecommerce brands spending $50K/month and above. According to Meta's own 2024 performance data, advertisers who consolidated into fewer, broader campaigns saw a 32% drop in cost per acquisition compared to fragmented setups. A 2025 AppsFlyer report found that 70–80% of Meta ad performance comes from creative quality, not budget or targeting settings.
So why are most brands still obsessing over interest stacks? Because they're playing by 2019 rules in a 2026 algorithm. This guide breaks down why broad targeting works, how to set it up, and the exact framework for making the switch without tanking your ROAS.
Table of Contents
- Why Interest Targeting Is Dead
- Interest vs. Broad: The Efficiency Gap
- How Broad Targeting Actually Works
- The Pixel Seasoning Prerequisite
- How to Set Up Broad Targeting Step by Step
- The 3-Campaign Architecture for Scaling
- The 48-Hour Testing Framework
- 3 Mistakes That Kill Broad Targeting Performance
- Frequently Asked Questions
- Key Takeaways
Why Interest Targeting Is Dead
If you're still targeting "People who like Lululemon" to sell activewear, you're paying what we call a Manual Labor Tax. Three things have fundamentally broken the old model:
1. Data Privacy Killed the Signal
Since iOS 14.5, Meta's visibility into third-party app behavior has collapsed. The interest data you're relying on is often 6–12 months old or based on incomplete signals. You're not targeting "fitness enthusiasts." You're targeting people who once liked a gym meme.
2. Meta's Andromeda Algorithm Changed Everything
In late 2024, Meta rolled out Andromeda, a completely new ad delivery system. During beta testing, it delivered a 5% increase in ad conversions on Instagram. By Q3, that improvement had doubled. Andromeda doesn't need you to tell it who to target. It reads your creative (the text, the visuals, the hook) and finds buyers automatically.
And as of June 2025, Meta is actively consolidating interest categories into broader groupings and removing the ability to exclude users based on certain interests. The platform itself is pushing you toward broad.
3. You're Bidding Against Yourself
If you're running 5 ad sets with 5 different interest groups ("Yoga" vs. "Peloton" vs. "Fitness"), you're not testing different people. You're competing against yourself in the same auction. Same high-intent users, higher CPMs, faster creative fatigue.
Key insight: Narrow targeting creates a performance ceiling. You might see 4.0 ROAS at $500/day, but the moment you push to $2,000/day, frequency spikes, the audience saturates, and your ROAS collapses.
Interest vs. Broad: The Efficiency Gap
Here's the data side-by-side:
| Metric | Interest Targeting | Broad Targeting |
|---|---|---|
| CPM (Cost Per 1K Impressions) | $25–$45 (competitive auction) | $12–$22 (open auction) |
| Audience Size | 500K–2M | 20M–250M+ |
| Fatigue Rate | High (days to weeks) | Low (months) |
| Primary Performance Lever | The "button" (audience settings) | The creative |
| Scaling Ceiling | ~$2K/day before frequency spikes | $10K+/day with creative diversity |
| Learning Phase Exit | Hard (small data pool) | Fast (massive conversion volume) |
The math is simple. Meta's algorithm needs roughly 50 conversions per ad set, per week to exit the Learning Phase.
With the old "spaghetti" approach (10 campaigns, 5 ad sets each), you need 2,500 conversions per week to stabilize. With a consolidated broad approach (3 campaigns, 1–2 ad sets each), you only need 150 conversions per week.
Less complexity. More data density. Stable CPAs.
How Broad Targeting Actually Works
"But if I don't tell Meta who to target, how does it know who to show my ads to?"
Your creative does the targeting. Every element of your ad acts as a filter. We call this the Creative Sieve Framework:
The Hook (Filters for Intent)
- Bad hook: "Check out our new product!" → targets everyone, attracts no one
- Broad hook: "Hitting a wall at 3 PM every single day?" → only resonates with people who have that specific problem
The algorithm sees who stops scrolling. It finds more people like them.
Visual Cues (Signal Your ICP)
- A founder in a professional office → signals B2B/corporate audience
- A mom in a messy kitchen → signals D2C/parenting audience
- Close-up of a product in-use → signals product-curious buyers
Meta's AI scans the frames of your video. It recognizes environments, objects, and people, then matches them to users who engage with similar content.
Language (Filters for Sophistication)
- Jargon like "electrolyte bioavailability" → targets health-savvy buyers
- Simple words like "stop feeling tired" → targets mass-market buyers
The takeaway: You don't need 5 different audiences. You need 5 different creatives, each speaking to a different avatar. The targeting is in the ad itself.
The Pixel Seasoning Prerequisite
Broad targeting isn't magic. It has one hard prerequisite: your pixel needs data.
If you have fewer than 500 conversions in the last 30 days, your pixel doesn't yet know what a buyer looks like. Going broad with an immature pixel is like giving a GPS directions with no map loaded. It'll wander aimlessly and burn your budget.
The readiness checklist:
- 500+ purchase events in the last 30 days (non-negotiable)
- CAPI (Conversions API) active with an Event Match Quality score of 8.0+ (without CAPI, you lose ~30% of your data)
- Purchase set as #1 priority in Aggregated Event Measurement
- Clean UTM tracking across all campaigns
If you're not there yet, use interest targeting to build your data. Think of it as seasoning the pixel. Once you've crossed 500 conversions, start testing one broad ad set alongside your existing structure.
How to Set Up Broad Targeting Step by Step
The technical setup is simpler than you think:
- Create a new campaign: CBO or ABO, conversion objective set to Purchase
- Set your audience to open: no interests, no lookalikes
- Age: 18–65+ (unless legally restricted)
- Gender: All (unless your product is physically gender-specific)
- Location: Target country only
- Placements: Advantage+ (let Meta decide)
- Load 3–5 creative concepts, each targeting a different avatar through the creative itself
- Set budget to 1–2x your target CPA per ad set per day
- Don't touch anything for 72 hours
That last step is the hardest one. Seriously. The algorithm needs 48–72 hours to stabilize. If you cut a campaign because ROAS dipped for 6 hours, you're punishing the algorithm for learning.
The 3-Campaign Architecture for Scaling
For ecommerce brands spending $50K+/month, you don't need more than three campaigns. This structure separates testing from scaling:
| Campaign | Budget Type | Budget Share | Purpose |
|---|---|---|---|
| ABO Sandbox | Ad Set Budget | 20–30% | Test new creative concepts in isolation |
| CBO Mainstage | Campaign Budget | 60–70% | Scale proven winners with broad targeting |
| Advantage+ Shopping (ASC) | Auto (CBO) | 10–20% | Let Meta's AI find new buyer pockets |
How they work together:
- ABO Sandbox is your lab. One ad set = one concept. 3–5 creative variants per concept. If a concept hits your target CPA after spending 2x CPA, it "graduates" to the Mainstage.
- CBO Mainstage is where the money lives. Only graduated winners go here. Broad targeting, 1–2 ad sets max. Never edit a winning ad. Just add new winners alongside it.
- Advantage+ Shopping is your black box. Feed it your top 5 all-time performers. Set the existing customer cap to 10–15% to force new customer acquisition.
This is the "Anti-Agency" approach. Most agencies hide behind 15+ campaigns and hundreds of ad sets to make their job look complicated. But complexity is the enemy of scaling. Consolidation gives you data density. Data density gives you stable CPAs. Stable CPAs let you scale.
The 48-Hour Testing Framework
When testing new creatives in your ABO Sandbox, here's the exact protocol:
- Launch 3–5 new concepts in broad ad sets
- Set budget to 1–2x target CPA per ad set per day
- Wait 48 hours. No touching.
- Analyze using the Metric Hierarchy:
| Priority | Metric | What It Tells You |
|---|---|---|
| 1st | ROAS / CPA | Did it sell? |
| 2nd | Thumbstop Rate (3s views / impressions) | Is the hook working? |
| 3rd | Hold Rate (ThruPlays / impressions) | Is the content engaging? |
| 4th | CTR (Click-Through Rate) | Is the offer compelling? |
Decision Matrix
| Signal | Action |
|---|---|
| High CTR, low sales | Your ad works. Your landing page is broken. Fix the page, not the ad. |
| Low thumbstop, decent sales | The creative converts but the hook is weak. Reshoot the first 3 seconds. |
| High ROAS, low spend | Winner. Move it to the CBO Mainstage and increase budget. |
| High CPA, low engagement | Kill it. Don't "wait for it to get better." |
3 Mistakes That Kill Broad Targeting Performance
1. The Retargeting Trap
You're spending 40% of your budget on warm audiences. Meta is taking credit for sales that would have happened anyway. The fix: Shift to 80% prospecting, 15% retargeting, 5% retention. With broad targeting, Meta is already naturally retargeting engaged users within the campaign.
2. Creative Stagnation
You're running the same 5 winners for 3 months. Your CPMs are climbing because the audience is bored. The fix: Ship 10–20 new creative concepts per month. Not just hook swaps, new angles. Think: different avatars, different pain points, different formats (static, UGC, founder-led, comparison).
3. Weak Offer Structure
You're sending cold traffic to a $30 product with no upsell. Your CAC is $25. There's no margin to scale. The fix: Bundle products to push AOV to $60–$80+. A higher AOV gives you room to pay more per acquisition and still scale profitably.
Frequently Asked Questions
Q: What is broad targeting in Meta Ads?
Broad targeting means running ads with no interest filters, no lookalikes, just basic demographics like age, country, and gender. You let Meta's algorithm find your buyers based on your ad creative and pixel data rather than manually selecting audiences. Meta recommends a minimum audience size of 2 million people for broad campaigns.
Q: Does broad targeting work for small budgets?
It depends on your conversion volume, not your budget size. You need at least 50 conversions per week per ad set for Meta's algorithm to optimize effectively. If you're spending under $5K/month and getting fewer than 50 weekly purchases, start with interest targeting to build pixel data first, then test broad once you cross the threshold.
Q: Should I still use lookalike audiences in 2026?
Lookalike audiences are being absorbed into Meta's Advantage+ system. They can still work as a middle ground, but for brands spending $50K+/month, broad targeting with strong creative consistently outperforms lookalikes because it gives the algorithm a larger data pool and avoids audience overlap issues.
Q: How long does it take for broad targeting to optimize?
Give it 48–72 hours minimum before making any decisions. The algorithm needs time to learn who engages, clicks, and buys. Judge performance on 7-day rolling averages, not daily snapshots. Most accounts see stabilization within 7–14 days of switching from interest to broad.
Q: Won't broad targeting waste money showing ads to irrelevant people?
That's the counterintuitive part: no. Meta's Andromeda algorithm reads your creative to determine who should see it. If your ad says "Tired of your skincare routine not working?" Meta will only show it to people with signals suggesting they care about skincare. The creative is the targeting. You'll actually spend less on irrelevant impressions because CPMs are 40–50% lower than competitive interest auctions.
Key Takeaways
- Interest targeting is a liability for brands spending $50K+/month: stale data, self-competition in auctions, and hard scaling ceilings make it unsustainable
- Meta's Andromeda algorithm (2024) fundamentally changed ad delivery. The platform now uses your creative to determine targeting, making manual audience selection redundant
- Broad targeting reduces CPMs by 40–50% compared to interest-based campaigns because you're bidding in an open auction instead of a competitive one
- Your creative is your targeting. The Creative Sieve Framework (hook filters intent, visuals signal ICP, language filters sophistication) replaces audience segmentation
- You need 500+ conversions in 30 days before going broad. "Pixel Seasoning" ensures the algorithm knows what a buyer looks like
- The 3-Campaign Architecture (ABO Sandbox → CBO Mainstage → Advantage+ Shopping) is all you need to test, scale, and discover new buyers
- Consolidation beats complexity. 3 ad sets needing 150 weekly conversions will always outperform 50 ad sets needing 2,500
Ready to Make the Switch?
Switching from interest targeting to broad isn't just a settings change. It's a mindset shift. You stop obsessing over buttons and start investing in creative. You let the algorithm do what it's built to do.
If you're spending $30K+/month on Meta and feel like you've hit a ceiling, the bottleneck probably isn't your targeting. It's your creative volume and your offer structure.
At Zentric, we help DTC brands build the creative engine and campaign architecture that makes broad targeting actually work, with a 3x ROAS guarantee in 90 days. Book a free discovery call and let's look at your account.
Ready to Scale Your Brand?
Book a free discovery call and learn how we can apply these strategies to grow your ecommerce brand.