Meta Wants to Run More of Your Marketing — But Should They?

​Want to know which parts of your Facebook and Instagram ad campaigns Meta AI can handle better than you can? Wondering how to use Meta's newest algorithm changes and AI tools to lower your ad costs and improve lead quality?

​In this article, you'll discover how Meta's latest updates to ad placements, creative AI, the pixel, and business agents are reshaping what marketers need to control and what they should hand over to the algorithm.

#1: Why Meta Is Removing Manual Placement Controls and What Advantage+ Placements Actually Does

Meta is rolling out changes that remove advertisers' ability to exclude individual placements at the ad set level. This includes the ability to target only mobile users, desktop users, iPhone users, or Android users. The change has started appearing in Ads Manager for some accounts and is expected to become universal.

​This isn't a surprise for experienced advertisers. Meta has been nudging people toward Advantage+ Placements for years, displaying pop-up warnings whenever someone tried to deselect placements manually. Advantage+ Placements lets Meta's algorithm decide which ads appear on which placements and when, using its own data to optimize delivery.

​The typical spend distribution under Advantage+ tells the story: around 60% of ad spend (sometimes higher) goes to Instagram, a large share goes to Facebook, and a very small percentage goes to the Audience Network, which places ads on third-party websites. Threads, WhatsApp, and Messenger also receive ads automatically when the ad format is compatible.

​Tara Zirker has tested this extensively across client accounts. Choosing Advantage+ Placements consistently produces better results and higher-quality leads than manual selection. Removing placements manually, even the ones that seem like wasted spend, tends to make ads more expensive and reduce overall lead and sales quality. Meta's detection systems have improved to the point where the algorithm simply won't place an ad in a spot where it won't render properly, whether the font is too small or the format wouldn't translate.

Who Still Needs Placement Control

Account-level controls remain available even as ad-set-level options disappear. At the account level, advertisers can still tell Meta to exclude Audience Network or other placements entirely.

​The businesses most likely to retain granular control are those in Meta's categorized industries. Enterprise healthcare accounts, financial services, and credit-related businesses often have compliance requirements dictating where ads can and cannot appear. Meta is keeping placement options open for these categorized accounts. Health and wellness is one example of a category that will continue to have access to more placement controls.

​For everyone else in the general, uncategorized bucket, the recommendation is straightforward: leave Advantage+ Placements on and let Meta's algorithm handle distribution.

#2: The Andromeda Algorithm and Why Creative Diversity Is Now the Competitive Lever

Everything Meta is doing right now traces back to a larger trajectory. The company started with Advantage+, which was the entry point for AI-managed ad decisions. The destination is an algorithm called GEM, a generative model where advertisers input an image and a budget, and the AI generates all creative, iterates on its own, and optimizes delivery without human intervention.

​The current era sits between those two points. Meta's newest algorithm, Andromeda, rolled out globally in October and is now active in every ad account. Andromeda works like an organic algorithm layered into the paid system. It favors high-quality content, directs the highest-quality traffic to accounts running the highest-quality creative, and rewards ads that generate positive engagement.

​This means pay-to-play now has tiers and gates. Spending money alone no longer guarantees access to the best audience segments. The quality of creative determines what tier of traffic an account can reach and what costs it pays.

​The practical takeaway is straightforward: the more creative diversity an advertiser provides to the algorithm, the more effective the advertising becomes. Meta is demanding creative diversity across accounts because it needs massive amounts of data to build toward GEM. Advertisers who provide varied creative formats, angles, and styles give the algorithm more to work with, which in turn unlocks better performance at lower costs.

How Meta's AI Creative Tools Are Advancing

The AI creative features inside Ads Manager are advancing faster than most advertisers realize.

​A beta feature observed on a client account demonstrated what's coming: Meta's AI generated two complete video ads with AI avatars using only static images that were already on the account. No script was provided. The AI pulled from the account's data to address the client's pain points, walk through benefits, handle objection overcoming, and present the intro offer. The avatars also performed a virtual walkthrough of the physical facility, constructed entirely from the uploaded images. The output was visually convincing with no detectable glitches.

​Advertisers do need to review and approve AI-generated creative before it runs. The approval interfaces vary across accounts as Meta tests different approaches, but the review step exists.

​Tara recommends turning on all AI creative options. The quality has reached a point where these tools consistently produce strong results, and the Andromeda algorithm rewards accounts that use them.

#3: How the AI-Enhanced Pixel and Product Catalogs Change the Data Game

Meta's pixel has undergone a quiet transformation. The code hasn't changed, but the pixel is now AI-enhanced, making it, as Tara puts it, “ten times smarter.”

​Previously, e-commerce advertisers had to manually pass rich product data to the pixel, including detailed product descriptions and SEO-level descriptors, to help Meta match products with the right audiences. The AI-enhanced pixel reads that information directly from the website. It scans product offerings, website content, and SEO data, then delivers that data back to Meta in a usable format for audience matching.

​The action item for advertisers is straightforward: clean up pixel errors. Most accounts have configuration issues, whether it's missing access levels, broken event tracking, or incomplete setup. The cleaner the pixel installation, the more the AI features can do with the data it collects. Every time an advertiser logs in, Ads Manager surfaces these errors. Fixing them is now more consequential than ever.

Product Catalog Enhancements

Product catalogs are becoming a core input across all sales campaigns, not just catalog-specific ad formats. Meta's AI can now automatically generate and optimize shopping ad formats in real time, personalizing them for each consumer based on the data Meta holds about their preferences and behavior.

​When someone clicks on a product, Meta AI surfaces potential questions the consumer might have. For a lotion, it might prompt “Is it good for sensitive skin?” or “What are the reviews saying?” and provide answers within Meta rather than sending the user to an external website. This keeps the shopping experience on-platform and reduces friction.

​Advertisers with large catalogs also have new control over which products get featured. A department store running a shoe sale can tell Meta to focus only on shoes rather than displaying the entire catalog. These catalog controls give advertisers greater precision, while the AI handles delivery optimization.

​Deeper Shopify integration is also on the way, which should further streamline the e-commerce experience within Meta's ecosystem.

Consumer-Side Personalization Changes

Meta has also changed how personalization works on the consumer side. Previously, users could opt in to personalization separately for their organic feed and their ad feed. Now the choice is streamlined: personalization is either on for both organic and ads, or off entirely. Opting out means the algorithm delivers random content.

​This is a strategic move for advertisers. Most consumers who initially opted out of ad personalization eventually realized they were still getting irrelevant ads. By bundling the choice, Meta is likely to increase the percentage of users who accept personalized ads, which improves targeting quality for every advertiser on the platform.

#4: Using Meta AI at Meta.ai for Account Analysis and Strategy

Meta AI, accessible at meta.ai, can now connect directly to ad accounts, Instagram profiles, Facebook pages, and Google Workspace. The tool is free to use for now, though Meta has indicated that businesses seeking greater usage will eventually be able to subscribe to MetaOne for expanded AI tool access.

​Getting started is simple. Logging in with existing Facebook or Instagram credentials grants access. From there, asking Meta AI to connect to a Business Manager account is a one-click enable. The tool then has access to account data and can provide analysis and recommendations.

Prompts That Deliver Useful Analysis

For an advertiser who is a couple of weeks into a campaign and past the learning phase, the starting prompts are conversational. Asking “What's working, what's not working, and what can I do better?” returns specific recommendations based on actual account data, including spend figures, impression counts, and performance breakdowns.

​For organic content analysis, asking Meta AI to evaluate Instagram Reels from the past 30 days returns a ranked list of top performers, along with an analysis of why each one worked. The tool can also be set to deliver recurring reports, such as a weekly Monday briefing covering key advertising and organic metrics.

​Meta AI has a mode selector similar to other AI platforms. Switching from the default “instant” mode to “thinking” mode allocates more processing power and produces deeper analysis.

Where Meta AI Differs From Other AI Tools

The differentiator is Meta-specific data. Connecting an ad account to Claude or ChatGPT via an MCP connector is possible and useful, but Meta AI pulls proprietary platform data that isn't available through those external connections.

​Tara tested this by uploading a competitor analysis folder to Meta AI for a large client account. The tool identified the competitor's longest-running ad, analyzed its messaging, and offered to create five differentiated ad angles. That kind of competitive insight, grounded in Meta's own ad library data, gives the tool a unique perspective.

​The caveat applies here as it does with any AI platform: recommendations must be filtered through context. Some suggestions may have already been tested, and compliance requirements in sensitive industries may rule out certain approaches. If a recommendation hasn't been tried and doesn't conflict with known constraints, testing it is worthwhile. Tara has been surprised multiple times by results from ideas that initially seemed either too basic or too complex.

#5: Meta Business Agent and the Rise of AI-Powered Customer Interactions

Meta is launching Meta Business Agent, an AI-powered assistant that operates across WhatsApp, Messenger, and Instagram DMs. The agent can answer customer questions about a business, make product recommendations from the catalog, book appointments, qualify incoming leads, and close sales on the e-commerce side. Businesses retain control over when a human team member steps in.

​The operational capabilities extend beyond customer-facing interactions. The agent can summarize customer conversations, provide insights on messaging patterns, and deliver daily briefings on communication activity.

​Tara's team has started experimenting with these agents specifically on lead gen ads, which are the on-platform ads where users don't have to leave Meta to submit their contact details. The agents are qualifying leads and helping people select the correct product. Early results are promising, though still limited in data.

Why AI Agents Give Businesses a Competitive Edge

The core advantage is speed. Many businesses struggle to respond to DMs in a timely manner, and staffing someone to answer them in a way that actually converts is expensive and inconsistent. AI agents handle both problems by providing instant, 24/7 responses that draw on the business's account data, previous chat conversations, and established tone.

​Businesses that sell anything with complexity benefit most. When potential customers have multiple questions before purchasing, whether about product details, logistics, pricing, or suitability, an AI agent that can answer immediately and accurately reduces friction at the most critical point in the buying process. It qualifies serious buyers faster and disqualifies tire-kickers, which is exactly what a sales process should do.

​Businesses that don't implement AI agents, whether through Meta or another provider, face a growing disadvantage. Consumer expectations are shifting toward instant answers. The companies that meet that expectation will convert at higher rates; those that don't will lose prospects to competitors who do.

​For businesses that want to enhance their sales pages specifically, tools like VideoAsk allow embedding a small corner video from a founder or team member that routes to different pre-recorded video answers based on the visitor's question. Combined with AI chat agents, this creates a layered support experience that covers both visual and text-based communication.

Other Notes From This Episode

Where to subscribe: Apple Podcasts | Spotify | YouTube Music | YouTube | Amazon Music | RSS

✋🏽 If you enjoyed this episode of the Social Media Marketing podcast, please head over to Apple Podcasts, leave a rating, write a review, and subscribe.

Stay Up-to-Date: Get New Marketing Articles Delivered to You!

Don't miss out on upcoming social media marketing insights and strategies! Sign up to receive notifications when we publish new articles on Social Media Examiner. Our expertly crafted content will help you stay ahead of the curve and drive results for your business. Click the link below to sign up now and receive our annual report!