Editorial Team A recent eMarketer survey from 2025 found that only 13% of marketers think their social media targeting is “highly effective.” That number tells me most people are still stuck on basic demographic segmentation and, frankly, burning through cash on impressions that never convert. So how do you actually stop shouting into the void and start connecting with people who are ready to buy?
Key Takeaways
- Turn on Meta’s Advanced Matching feature. It can improve your Custom Audience match rates by up to 20% which directly makes your retargeting campaigns more efficient.
- Start using Value-Based Lookalike Audiences by Q3 2026. This means you stop chasing just any new customer and start finding people who look like your most profitable ones, based on CLTV.
- Get Offline Conversion Tracking integrated for at least 60% of your campaigns by the end of the year. You need to feed real-world sales data back to the algorithm so it can optimize for what actually makes you money.
- Roll out Dynamic Creative Optimization (DCO) on at least 40% of your product-focused campaigns. Let the system figure out the best ad combos instead of guessing yourself.
Only 13% of Marketers Are Highly Effective: The Illusion of Basic Targeting
That 13% figure from eMarketer is a huge warning sign. Too many advertisers are still building their Facebook ads around broad interests, age, and location, a strategy that worked five years ago but is now completely saturated. With the number of advertisers on the platform and how smart Meta’s algorithms have gotten, targeting “people interested in ‘online shopping'” is just setting your money on fire. Meta’s AI needs better signals than that to find the right people for you.
From what I see, a lot of marketing teams are just running the same old playbook because it’s comfortable. They haven’t adapted to the new realities of data privacy or the algorithm’s potential. The outcome is predictable: your CPMs go up and your returns go down. To get better results, you have to give Meta richer data and much clearer instructions. The goal is finding the right people with surgical precision.
Advanced Matching Boosts Custom Audience Match Rates by Up to 20%
Advanced Matching is one of the most powerful features inside Meta’s platform that almost everyone seems to ignore. When you flip it on, it lets Facebook use hashed customer info from your site (like name, address, phone, email) to get a much better match rate for your Custom Audiences. For e-commerce clients, I’ve personally seen this one setting increase the size and accuracy of their retargeting audiences by as much as 20%.
Imagine you’re trying to retarget cart abandoners. Without Advanced Matching, Meta might only be able to match a user if they used the same email. With it turned on, other data points like a phone number or even part of an address (all securely hashed, of course) help Meta confidently find that same person even if they’re on a different device. You miss fewer people and your retargeting net gets much tighter. It’s a simple toggle in your pixel settings inside Events Manager, but so many accounts have it off. Activating it is a basic requirement for building remarketing funnels that actually convert.
Value-Based Lookalikes Outperform Standard Lookalikes by 3x on ROAS
Standard Lookalike Audiences are fine. You give Meta a source audience (like a customer list) and it finds similar people. But Value-Based Lookalike Audiences are on another level. This feature lets you upload a customer list that includes a specific value for each person, like their customer lifetime value (CLTV). The algorithm then hunts for new users who don’t just look like your average customer, but who look like your best customers. According to Meta’s own internal case studies, advertisers using these have seen a 3x higher Return on Ad Spend (ROAS) than with standard Lookalikes.
This is how you shift from simply buying customers to buying profitable customers. For example, a SaaS company shouldn’t just upload a list of every single trial sign-up. You should segment that list and upload only the users with a high annual contract value (ACV) or those who’ve stuck around for more than two years. By feeding the algorithm that specific, high-value data, you’re telling it, “Go find more people exactly like *these*.” Yes, it requires more data prep on your side, but the ROI almost always justifies the work. I tell every business with enough transaction data that they absolutely have to be testing this audience type.
| Feature | Advanced Matching | Value-Based Lookalikes | Offline Conversion Tracking |
|---|---|---|---|
| Improves Custom Audience Match Rates | ✓ Up to 20% | ✗ | ✗ |
| Requires CLTV/Purchase Value Data | ✗ | ✓ | ✗ |
| Boosts ROAS Compared to Standard Lookalikes | ✗ | ✓ 3x higher | ✗ |
| Feeds Real-World Purchase Data to Algorithm | ✗ | ✗ | ✓ |
| Improves Overall Campaign Performance | ✓ Indirectly | ✓ Indirectly | ✓ Up to 15% |
| Simple Toggle in Events Manager | ✓ | ✗ Requires data prep | ✗ Requires data upload |
| Focuses on Acquiring Profitable Customers | ✗ | ✓ | ✓ Indirectly |
Offline Conversion Tracking Improves Campaign Performance by 15%
Your digital ads don’t operate in a bubble. Businesses with physical stores or sales teams often have no idea how to connect their ad spend to actual revenue. That’s exactly what Offline Conversion Tracking is for. By uploading your offline sales data (like in-store buys, phone orders, or deals closed in your CRM) back into Meta, you give the algorithm the full story. Meta’s own Business Help Center says that advertisers who do this can see their campaign performance improve by up to 15%.
Let’s say you’re running ads for a car dealership. Someone sees your ad online and clicks, but then they show up at the dealership a week later and buy a car. Without offline tracking, Meta sees a click and nothing more. When you upload that sales data, you’re training the algorithm to recognize which online activities lead to someone actually buying a car. It then gets smarter about optimizing your ad delivery to find more people likely to convert in the real world. Any business with offline sales is making a huge mistake by not using this. It connects your digital efforts to actual money in the bank.
Dynamic Creative Optimization (DCO) Delivers 2x Higher Click-Through Rates
Your ad creative is just as important as your audience targeting. Manually building dozens of ad versions for every segment is a slow, inefficient grind. Dynamic Creative Optimization (DCO) automates all of that. You just feed Meta a bunch of creative components, images, videos, headlines, descriptions, CTAs, and the system automatically tests different combinations to build personalized ads for every user. An IAB study showed DCO campaigns can get click-through rates (CTRs) 2x higher than static, single-image ads.
The system is figuring out the right combination of creative elements for each person. For an apparel brand, DCO can test product photos against different headlines (e.g., “Shop Now,” “Learn More,” “Get 15% Off”) and see what works. The algorithm learns fast, identifying which combos resonate with which audiences and then putting budget behind the winners. It’s an incredibly effective way to lift engagement and conversions without the headache of manually A/B testing everything. I see a lot of marketers get nervous about giving up that creative control, but the data is clear: DCO almost always beats human guesswork.
The Conventional Wisdom I Disagree With: “Always Target Broadly to Let the Algorithm Work”
I keep hearing this idea that because Meta’s AI is so good now, advertisers should just target an entire country, set the age to 18-65, and “let the algorithm do its job.” While there’s a tiny bit of truth in that for some accounts, this advice is mostly wrong and will cost you money. It assumes you have a bottomless budget and a pixel that’s already perfectly seasoned with years of data.
In my experience, broad targeting only works for giant, established brands that have massive budgets and are tracking millions of conversions. For most businesses, it’s just a way to burn cash while the algorithm “learns” who your customers are. That learning phase is expensive. Why pay for it? I prefer a hybrid approach: you start with well-defined (but not tiny) Custom and Lookalike Audiences to give the algorithm strong starting signals. Once the campaign proves it can run efficiently, then you can start carefully expanding your targeting. This gives the algorithm a head start, pointing it toward the right users faster and cheaper. You have to collaborate with the algorithm, not blindly trust it.
The notion that you can cast a wide net and hope Meta’s magic finds your perfect customer completely ignores how competitive the ad platform is in 2026. Every piece of quality data you feed the system gives you an advantage. Only a tiny fraction of advertisers can afford the luxury of broad targeting without strong first-party data. For more on getting the most out of your budget, look into some of the new rules for social ad ROI.
Getting beyond basic targeting is mandatory for growth now. It’s not optional. Using features like Value-Based Lookalikes, Advanced Matching, Offline Conversion Tracking, and DCO is how you improve performance and get a real return on your investment. It all comes down to committing to constant testing and letting the data guide your refinements so every dollar you spend works as hard as it possibly can. You should also check out how AI marketing can boost ROAS for another layer of optimization, and make sure you’re not falling for any common Facebook Ads retargeting myths.
What is Advanced Matching in Facebook ads?
This feature lets your pixel send more hashed customer identifiers (like name, email, and phone number) from your website to Meta. Doing this significantly improves the match rate for your Custom Audiences, making your retargeting campaigns more accurate and effective.
How do Value-Based Lookalike Audiences differ from standard Lookalike Audiences?
A standard Lookalike just finds people similar to your source list. A Value-Based Lookalike is much smarter because you upload your customer list *with* a monetary value for each person (like their lifetime value). Meta then searches for new people who resemble your most profitable customers, not just any customer.
Why is Offline Conversion Tracking important for Facebook ad campaigns?
It connects your online ads to real-world sales. You upload data from your store or CRM (like in-person purchases or closed phone leads) back to Meta. This teaches the algorithm which ads are actually driving revenue, so it can optimize for better results and give you accurate attribution.
What is Dynamic Creative Optimization (DCO) and how does it help?
With DCO, you give Meta a library of creative assets, different images, headlines, and descriptions. The system then automatically tests and combines these elements to create personalized ads for each user, serving the winning combinations to get you more engagement and conversions without all the manual A/B testing.
Should I always use broad targeting for my Facebook ads?
No, probably not. While it can work for huge advertisers with tons of conversion data and big budgets, most businesses should start with more defined Custom and Lookalike Audiences. This gives the algorithm better initial signals and is much more cost-effective than paying for it to “learn” from scratch.
