Meta offers two main ways to reach your ideal audience: Suggested Audience, where Meta’s algorithm automatically decides who should see your ad; and Filter Audience, where you manually define who to target using specific filters and parameters.
With the introduction of Meta’s Andromeda AI, early reports suggest that it has significantly improved the algorithm’s ability to match ads with the right users.
So, should you be excited?
In this article, we’ll break down how each targeting option works, explore what Andromeda AI actually does so you can decide which approach makes the most sense for your campaigns.
What is Suggested Audience?
Suggested Audience is Meta’s AI-driven targeting option that automatically determines who’s most likely to respond to your ad, then targets them based on its own data signals and algorithms.
When you create a campaign, Meta analyses several factors to predict which users might respond to your ads. Some of the main factors include:
- Your campaign objective: The goal you choose (in this case, leads) tells Meta what kind of action to optimise for.
- Your ad content: Meta’s system analyses your ad text, images, and videos to understand the topic, tone, and context to match them with users who are more likely to find it relevant.
- Your past interactions: If you’ve already run ads or have pixel data, Meta studies how people have interacted with your previous ads to find similar users.
- Platform signals: It also uses broader behavioural data, such as engagement trends, click history, and interests, to improve predictions over time.
Pros of Using Suggested Audience
Less setup work
With Suggested Audience, you don’t need to spend time defining every audience trait. Meta’s algorithm automatically figures out who to show your ads to based on your campaign goal, content, and previous results.
Adapts automatically
As your ads collect engagement or conversion data, Meta continuously learns and tracks who’s responding and who’s converting. It then prioritises showing your ads to similar users. This built-in feedback loop means your campaign can improve automatically without frequent manual changes.
Great for scaling
Once Meta’s AI understands your ideal customers, it can easily expand your reach by finding new people with similar interests or traits. Suggested Audience can play an excellent role in growing campaigns that already have a good amount of data and conversions
Cons of Using Suggested Audience
Limited control
Under the Suggested Audience option, you can’t directly choose the specific demographics, interests, or behaviours your ad targets. It can feel restrictive for businesses that need precise control.
Needs data to perform well
Suggested Audience performs best when Meta has a good amount of engagement or conversion data to learn from. If you’re new to Meta Ads or your ad account doesn’t have enough pixel data, the system may take longer to find the right audience and might waste some of your initial budget during the learning phase.
Harder to analyse
Since you don’t define or see the exact audience makeup, it’ll be harder to tell why a campaign worked or failed. Even strong results may come with fewer insights into the type of people who converted.
What is Filter Audience?
Commonly known as manual targeting, Filter Audience is a more traditional way of targeting people through Meta Ads. This option lets you build your own audience by applying specific filters and parameters.
Under Filter Audience, you define your target audience based on:
- Location: Choose specific countries, cities, postal codes, or even areas within a certain radius around your business.
- Demographics: Narrow down by age, gender, or language preferences.
- Interests and behaviours: Target people based on what they like, follow, or engage with on Facebook and Instagram.
- Custom audiences: Upload your own customer list or target people who have already interacted with your previous ads.
- Lookalike audiences: Create a new audience that shares similar traits with your best customers or existing followers.
Pros of Using Filter Audience
Full control
With the Filter Audience option, you get to decide who exactly sees your ads. This level of control can be excellent for targeting a certain demographic, for example, women aged 25 – 40 in New York who are interested in yoga and wellness. You can tailor every campaign to match your ideal buyer persona.
Transparency and insight
Because you’ve defined the parameters yourself, it’s easier to analyse which audience segments perform best. You can compare results by age group, interest, or location and make data-driven decisions.
Ideal for local or niche campaigns
Filter Audience can be great for businesses with well-defined customer profiles. For instance, a real estate agent can target homeowners in a specific neighbourhood, or a salon can reach women within a 10 km radius who are interested in beauty and self-care.
Cons of Using Filter Audience
Takes time and effort to define and refine
Building a manual audience and refining it over time requires research, testing, and frequent adjustments. You’ll need to experiment with different combinations of demographics and interests to find what actually works.
Easier to make mistakes
Because everything is manually configured, you risk making your audience too narrow or too broad, especially if you’re new to the platform. If your filters are too specific, Meta might struggle to find enough people to deliver your ads efficiently. On the other hand, if your targeting is too broad, you risk wasting your budget reaching people who aren’t relevant.
Doesn’t learn automatically
Unlike Suggested Audience, Filter Audience doesn’t adapt on its own. You’ll need to monitor performance by yourself and manually adjust your targeting based on results. It’s a more hands-on approach that relies on your understanding of your customer base and campaign data.
| Suggested Audience | Filter Audience | |
| Who controls targeting | Andromeda AI | You |
| Setup time | Quick and simple | Manual and detailed |
| Optimisation | Automatic, improves over time | Manual adjustments needed |
| Transparency | Limited | Full visibility |
| Scalability | Good for growth campaigns | Best for niche targeting |
| Performance with Andromeda | Generally improves with better matching | Largely unchanged |
What is Andromeda AI and how does it influence ad targeting
Meta’s Andromeda AI is the technology that decides which ad gets shown to which user across Facebook, Instagram, and other Meta platforms. While you don’t directly “turn on” Andromeda or adjust its settings, it works behind the scenes inside Meta’s ad delivery system.
In simpler terms, Andromeda AI makes the matching process between users and ads faster and more efficient.
Here are three key ways in which Andromeda affects ad campaigns:
1. Faster learning
Andromeda can process huge volumes of data and evaluate many ad-user combinations. Your campaigns learn much faster than older systems. The “learning phase” is the period when the algorithm figures out who responds best to your ad. During this time, you risk wasting part of your ad budget as the system identifies your ideal customer. Andromeda’s improvements in retrieval speed and model efficiency help shorten this phase, allowing your campaigns to stabilise faster.
When the system learns faster, you see stable ad performance sooner, helping you move more quickly from testing into scaling.
2. Improved audience matching
Andromeda uses advanced machine-learning models and effective indexing combinations to match ads with users who are likely to respond positively.
For you as an advertiser, that means:
- The “lookalike” element of Suggested Audience is more effective. The algorithm might find micro-segments you may have overlooked.
- Even if you put “broad” targeting settings, the internal system can target the right people.
- The relevance of your ad to the viewer tends to increase, reducing wasted impressions.
3. Improved efficiency
Andromeda screens through millions of ad candidates and narrows down to the best fits faster. This helps you get less irrelevant impressions, spend less on “wasted” exposure, and get more conversions out of the same audience-targeting spend.
Things to consider
While Andromeda boosts performance in several ways, keep in mind that it does not eliminate the need for a good foundation. Here are a few things to understand:
- If your tracking, conversion events, or pixel setup are weak, even the best algorithms can’t help.
- No amount of targeting refinement can save an ad whose creative or ad message is poor.
- Because much of the matching is now automated, you may not be able to see the analytics behind a performance. That means you still need to monitor top-level metrics and test purposefully.
- A system like Andromeda works better when you feed it good data, consistent creatives, and proper campaign objectives. You need that as a base.
When to choose Suggested Audience
Suggested Audience works best when Meta has enough information to make smart decisions on your behalf. It can be a great fit if you want to take advantage of Meta’s AI-driven optimisation without spending hours fine-tuning every targeting filter.
Here are the main situations where using Suggested Audience makes the most sense:
When you already have data for Meta to learn from
Meta’s AI relies heavily on data from your pixel, past ads, website interactions, and conversions.
If your account already has a history of engagement, the system can analyse those patterns to identify what types of users convert best.
For instance:
- If you’re running Facebook Lead Ads and have collected form submissions, Andromeda can look at who completed those forms and find similar users.
- If your pixel has tracked purchases or add-to-cart events, the algorithm can use that data to prioritise users who are more likely to buy.
In short, the more quality data you can provide Meta, the more effective your Suggested Audience becomes.
When you’re testing new creatives or scaling an existing campaign
Suggested Audience is excellent for experimenting and scaling because it allows Meta to explore a broad range of users quickly.
If you’re testing new ad creatives (like different images, messages, or CTA), Suggested Audience can identify how different audience segments respond without you manually defining them. Meta automatically tests combinations and delivers more impressions to the best-performing segments.
Similarly, if you already have a campaign that’s performing well and want to expand your reach, Suggested Audience can help you scale efficiently. The AI will look for new users who share behaviours and characteristics with your existing customers. It’s a great way to extend your results without requiring a complete retargeting strategy.
When you want to save time and let the system optimise automatically
If you’d rather focus on creative strategy and conversion tracking than audience building, Suggested Audience can do the heavy lifting.
Instead of spending time picking demographics, interests, and exclusions, you simply define your campaign goal and let Meta’s system decide who’s most likely to deliver that outcome.
However, keep in mind that it’s not a “set-and-forget” option. You’ll still need to monitor results and adjust your creatives or objectives, even though Andromeda dramatically has reduced the need for manual audience management.
When to use Filter Audience
While Meta’s Suggested Audience can do most of the heavy lifting with AI, there are still many situations where Filter Audience targeting is the smarter choice. With this option you have full control over who sees your ads, where they’re shown, and how your budget is distributed. This information is critical for certain campaigns.
Here are the main scenarios where Filter Audience works best:
When you’re targeting a very specific demographic or local area
If your product or service appeals to a clearly defined group of people, manual targeting gives you the precision you need.
For example:
- A real estate agent advertising a property in a specific city.
- A restaurant or salon promoting to people within a 10 km radius.
- A language school targeting parents of young children in a specific neighbourhood.
In these cases, opting for Suggested Audience could mean wasting impressions on people outside your service area or demographic. By manually setting filters for location, age, gender, and interests, you can ensure your ad budget is focused on exactly where it matters.
When you want clear visibility into who’s seeing your ads
One of the biggest advantages of Filter Audience is transparency. You know exactly who you’re targeting and can easily understand why certain ads perform better than the others.
For instance, if you’re targeting men aged 25 – 40 interested in fitness, or women aged 30 – 50 interested in home décor, you can directly compare which group drives more engagement or conversions.
In contrast, in the case of Suggested Audience, you see the results, but you don’t always know who your ads reached or what made them click.
When you’re running compliance-sensitive or niche campaigns
Some industries or campaigns need tight control over who sees their ads due to regulations, brand sensitivity, or target market limitations.
For example:
- Some financial services or insurance companies must avoid advertising to certain age groups or regions.
- Healthcare or wellness brands must ensure messaging reaches specific health-interest audiences.
- Some niche B2B services want to run ads where only a small, relevant audience (e.g. construction contractors) should see the ads.
In these cases, using a broad, AI-driven audience targeting system can lead to noncompliance or wasted spend on irrelevant impressions. Manual targeting ensures your message reaches qualified, appropriate, or compliant audiences.
Final words
Andromeda AI doesn’t change how you set up campaigns, but it can make Suggested Audiences more efficient over time. If you’re running small, local, or niche campaigns, manual filters still give you valuable control. But if you have good data and want to scale, trusting Meta’s AI can save time and improve consistency.
If you’re new to Meta Ads, start with a Filter Audience to learn which types of people respond best. Once you’ve gathered enough data, switch to Suggested Audience so Meta can find similar users automatically
Set up your Meta’s Conversions API with Privyr to train the algorithm on your conversion data and generate better leads with each campaign. Privyr helps you to feed richer, real-time data back into the system, helping the AI learn which types of users are most valuable to your business.
