Why Facebook Ads Targeting Is Different From Google
On Google, the audience finds you. Someone searches for your product or service, and your ad appears in front of them at that moment of intent.
On Facebook, you find the audience. Nobody on Facebook is searching for your product — they're browsing their feed. Your job is to identify who your ideal customer is, reach them before they're actively looking, and give them a reason to care.
This difference has a major implication: Facebook targeting requires more creative testing, more audience refinement, and a willingness to experiment with different segments before finding what works.
Meta Ads Manager has three core audience types, each serving a different purpose:
Core Audiences (Interest Targeting) Built using Facebook's demographic and interest data. You choose characteristics like age, location, interests, behaviours, and job titles. Best for reaching cold audiences who've never heard of you.
Custom Audiences Built from your own data — website visitors (via pixel), customer email lists, video viewers, app users, or people who've engaged with your Facebook/Instagram pages. Best for retargeting and re-engagement.
Lookalike Audiences Built by Meta's algorithm to find people who resemble your best existing customers. Requires a Custom Audience as the source. Best for scaling once you have proven conversion data.
Interest targeting is where most advertisers start — and where most of the budget leaks occur.
In Ads Manager → Create ad set → Audience section:
Layer interests using AND logic for precision: Using "AND" between interest groups means the person must match all of them. Example: People interested in "Google Ads" AND who are "Small business owners" AND who have the behaviour "Engaged shoppers". Narrower audience, higher relevance.
Use OR logic for reach: Add multiple interests within the same targeting box to expand reach. All interests in the same box use OR logic — the person only needs to match one.
Test broad vs narrow: Counter-intuitively, very narrow interest targeting sometimes underperforms broad targeting with strong creative. Meta's algorithm is good at finding buyers within a large audience — your creative does more of the qualification work than your targeting does.
Avoid overlapping audiences: If you're running multiple ad sets targeting overlapping interests, you'll compete against yourself in the auction. Use Meta's Audience Overlap tool to check overlap between audiences before launching.
Facebook's interest categories are based on engagement signals — pages people like, content they interact with, links they click — not declared preferences. Someone interested in "digital marketing" might be a student, a freelancer, or a CMO. Interest targeting can get you in front of the right category of people; it cannot pre-qualify their buying stage or intent.
Custom Audiences are built from people who've already had some interaction with your business. They're almost always more cost-effective than cold interest targeting because the audience already has some familiarity with you.
Requires the Meta Pixel installed on your website. Allows you to target people who visited specific pages, spent a certain amount of time on site, or completed a specific action.
High-value website audiences to build:
|
Audience |
Targeting rule |
Time window |
|---|---|---|
|
All visitors |
All website traffic |
30 days |
|
High-intent visitors |
Visited pricing/services page |
14 days |
|
Non-converters |
Visited site but did NOT trigger Lead event |
30 days |
|
Past customers |
Triggered Purchase/Lead event |
180 days |
Upload a CSV of customer email addresses (or phone numbers). Meta matches them against Facebook profiles and creates a targetable audience.
Best uses:
Match rate: Meta typically matches 50–70% of uploaded email addresses to Facebook profiles. A list of 1,000 emails might produce a usable audience of 500–700 people.
Built from people who've interacted with your content on Facebook or Instagram — without needing them to visit your website.
Useful engagement audiences:
These audiences are particularly valuable before your pixel has enough data to build reliable website audiences.
Lookalike Audiences tell Meta's algorithm: "Find me more people like these." Meta analyses the profiles in your source audience and finds new people with similar characteristics across Facebook and Instagram.
1% Lookalike: Closest resemblance to your source. Smaller, higher quality. Best starting point.
2–5% Lookalike: Larger, slightly less similar. Good for scaling once 1% is proving out.
6–10% Lookalike: Large audiences, less precise. Use when you need volume and have strong creative.
The better the source, the better the Lookalike. In order of preference:
Minimum source size: Meta recommends at least 100 people in the source audience, but 1,000+ produces meaningfully better Lookalikes.
Rather than guessing which audience works best, run structured audience tests:
Phase 1 — Test audience types: Run the same creative against 3 different audiences with equal budgets for 7–14 days:
Measure cost per result. Allocate more budget to the winner.
Phase 2 — Test interest combinations: Within interest targeting, test different interest groups at the same budget level. Identify which interest clusters produce the lowest cost per result.
Phase 3 — Scale the winner: Increase budget on proven audiences by no more than 20–30% per day to avoid triggering a new learning phase.
How big should my Facebook Ads audience be? For cold interest targeting, 500,000 to 2 million is a workable range for most campaigns. Too small (under 50,000) limits delivery; too large dilutes relevance. For retargeting, audiences as small as 1,000 can work effectively.
Should I use Advantage+ audience or manual targeting? Meta's Advantage+ audience (formerly "Advantage Detailed Targeting Expansion") allows Meta to expand beyond your defined audience if it finds conversion opportunities. It tends to work well for established campaigns with conversion data. For new campaigns, manual targeting with defined audiences gives you more control while you're learning what works.
Does interest targeting still work after iOS 14? Yes — interest targeting is not pixel-dependent and was not directly affected by iOS 14. What iOS 14 affected was conversion tracking accuracy for pixel-based events. Interest targeting continues to work as before.
How often should I refresh my audiences? Custom audiences based on pixel data update automatically. Lookalike audiences should be refreshed every 1–3 months as your customer base evolves. Interest-based audiences don't need refreshing, but if performance declines, testing new interest combinations is worthwhile.
Can I target by job title or company on Facebook? Yes, but it's less reliable than LinkedIn's targeting. Job title and employer information comes from what users self-report on their profiles — coverage is incomplete, especially outside the US. For precise B2B targeting by job function, LinkedIn is more accurate, though more expensive.
Related reading: How to Advertise on Facebook Step-by-Step | Facebook Ads Retargeting Setup Guide | What is Social Media Advertising | Facebook Ads vs Google Ads