How Meta Ads Attribution Actually Works in 2026

For the past two years, Meta has been stripping away manual controls. Interest targeting became a “suggestion.” Detailed targeting expanded by default. Advantage+ took over audience selection. The message was clear: trust the algorithm. Then, in June 2025, Meta did something that seemed to contradict all of that. They gave advertisers value rules. Value rules let you tell Meta’s algorithm that certain audience segments are worth more (or less) to your business, and to adjust its bids accordingly. You can bid 60% more for women aged 25 to 34. Or bid 40% less for users in a specific country. Or deprioritize a placement where your conversions have a high refund rate. All without creating separate ad sets or fragmenting your campaign structure. As Jon Loomer put it in his deep dive on the feature: value rules address a real weakness in Meta’s optimization. The algorithm is designed to get you the most results possible. But it doesn’t inherently care about the value of those results. It optimizes for volume, not quality. Value rules let you layer business intelligence on top of algorithmic efficiency. The feature launched for Sales and App Promotion campaigns in June 2025, expanded to all campaign objectives by August 2025, and received significant enhancements in 2026 including placement-specific rules and device platform adjustments. But here’s the thing Meta tells you upfront in the setup screen: “When you use value rules, you may see more conversions from your preferred audiences, but your overall cost per result may increase.” That warning isn’t decoration. Meta’s own documentation says value rules can increase your cost per result by 20 to 1,000%. Not a typo. One thousand percent. So the question isn’t whether value rules are powerful. They are. The question is whether you know enough about your business data to use them without lighting money on fire. How Value Rules Actually Work (With Examples) The logic is straightforward. Value rules are bid multipliers applied at the audience segment level within Meta’s auction system. Every time your ad is eligible to appear, Meta calculates a Total Value Score: Total Value = (Advertiser Bid x Estimated Action Rate) + Ad Quality + User Value When you set a value rule, you’re adjusting the “Advertiser Bid” component for specific audience segments. A +50% rule means Meta bids 50% higher for people matching that segment. A -40% rule means Meta bids 40% less. What you can target with value rules: Age ranges GenderLocation (countries, regions, states) Mobile operating system (iOS or Android) Device platform Ad placement (Feed, Stories, Reels, Audience Network, Marketplace) You can combine up to 2 criteria per rule (for example, “women aged 25 to 34” or “iOS users in California”). You can create up to 10 rules per rule set, and up to 6 rule sets per account. Rule priority matters. When a user qualifies for multiple rules, Meta applies only the first matching rule in the sequence. So if Rule 1 is “+20% for women in California” and Rule 2 is “+50% for iOS users,” a woman in California using an iOS device gets the +20%, not the +50%. Order your rules from most specific to least specific. Example 1: E-commerce LTV optimization Your data shows women aged 25 to 44 have an average lifetime value of $850 over 12 months. Your overall average is $530. Women in this age range are worth about 60% more to your business. Setup: Rule 1: Increase bid +60% for Women, Age 25-44 Rule 2: Decrease bid -20% for Men, Age 55-65+ Your current CPA is $45 across all demographics. With the +60% bid increase, you’re willing to pay up to $72 to acquire a woman aged 25 to 44, because her LTV justifies it. Meanwhile, you’re bidding less for a segment that your CRM data shows has a high return rate and low repeat purchase rate. Example 2: Geographic performance differences You run a B2B SaaS product. Leads from major US metro areas convert to paid customers at 3x the rate of leads from smaller markets. Setup: Rule 1: Increase bid +40% for Location: New York, San Francisco, Chicago, Los Angeles, Austin Rule 2: Decrease bid -30% for Location: [lower-converting regions] Meta’s algorithm still reaches all audiences. But it bids more aggressively for impressions in high-converting metros, steering more budget toward the leads that actually close. Example 3: Placement quality differences Your data shows that Feed conversions have 2x the average order value compared to Audience Network conversions. Setup: Rule 1: Increase bid +30% for Placement: Facebook Feed, Instagram Feed Rule 2: Decrease bid -50% for Placement: Audience Network Instead of excluding Audience Network entirely (which reduces the algorithm’s delivery options), you’re deprioritizing it through bid adjustments. Meta will still deliver there when it’s extremely cheap, but your budget concentrates on placements where the conversion quality is higher. How Value Rules Impact Campaign Pacing and Spend This is where things get practical and where most advertisers underestimate the effect. Value rules don’t change your daily budget setting. Your budget stays the same. What changes is how aggressively Meta competes in auctions for specific segments. When you increase bids for a segment, Meta enters higher-priced auctions to reach those people. That means: You win more auctions for that segment. More of your budget goes toward the people you value most. You pay more per impression for that segment. Higher bids mean higher CPMs. Your budget depletes faster if a large portion of your audience matches the high-bid rule. When you decrease bids for a segment, the opposite happens. Meta is less competitive in auctions for those people, which means fewer impressions but lower cost per impression. The net effect on pacing: If your high-value segment is a small portion of your total audience (say, 15%), the pacing impact is manageable. The budget shifts toward that 15% without dramatically changing overall delivery. But if your high-value segment is 60% of your audience and you’re bidding +50%, you’ve effectively increased your average bid across most of your delivery. Your budget will pace faster and may exhaust earlier in the day. What to watch for: Check hourly delivery after activating value rules. If your budget is exhausting by 2 PM instead of running through the full day, your bid increases are too aggressive for your budget. Monitor impression share by segment. Are you actually getting more delivery to the segments you increased bids for? If not, competition might be too high at that bid level. How Value Rules Impact Cost Per Result Let me be blunt about this. Value rules will almost always increase your average cost per result. Meta tells you this outright during setup. The question isn’t “will costs go up?” They will. The question is “will the value of the results go up by more than the cost?” Here’s the math that matters: Without value rules: 100 conversions at $45 CPA = $4,500 spend Average customer value: $530 Total customer value: $53,000 ROAS: 11.8x With value rules (+60% bid for high-LTV segment): 85 total conversions at $55 CPA = $4,675 spend 50 of those conversions are from the high-LTV segment ($850 average value) 35 conversions from other segments ($530 average value) Total customer value: (50 x $850) + (35 x $530) = $61,050 ROAS: 13.1x Fewer total conversions. Higher CPA. But higher total customer value and better ROAS. That’s the trade-off. Value rules sacrifice volume efficiency for value efficiency. If you’re optimizing for top-line conversion count, value rules will look like they’re hurting you. If you’re optimizing for revenue and LTV, they can look very different. Without Value Rules With Value Rules Conversions 100 conversions 85 conversions CPA $45 CPA $55 CPA Avg Value $530 avg value $640 avg value Total Value $53K total value $61K total value ROAS 11.8x ROAS 13.1x ROAS Fewer conversions. Higher CPA. More revenue.  Value Rules vs. Narrow Audience Targeting Before value rules existed, the standard approach for targeting high-value segments was audience fragmentation. You’d create separate ad sets for different demographics, each with its own budget. Women 25 to 34 in one ad set. Men 35 to 44 in another. Different budgets reflecting different segment values. This approach has three problems in 2026: 1. It fragments the learning phase. Each ad set needs 50 conversion events per week to exit learning phase. If you split one campaign into 4 demographic ad sets, each one needs to generate 50 events […]

Everything You Knew About Creative Testing Is Wrong Now! Two years ago, the winning playbook looked like this: find one killer image, write 10 headline variations, split them across 5 interest-based ad sets, and let the winner emerge. Rinse and repeat. That playbook is dead. And the people still running it are the ones posting on Reddit asking why their CPMs doubled overnight. Here’s what happened: Meta deployed Andromeda globally between late 2025 and January 2026. It’s not a minor tweak. It’s a ground-up rebuild of how ads get matched to users. The old system started with your audience selections. Andromeda starts with your creative. It reads the visual, the audio, the copy. It decides who should see it. Your targeting inputs are suggestions at best. The result? Brands testing 20+ new ads per month are seeing 65% higher ROAS than brands testing under 10. The top-performing advertisers run roughly 395 live ads versus 296 for the bottom third. Creative volume and creative diversity are now the primary scaling levers. But “test more creatives” isn’t a strategy. You need to understand what Andromeda actually looks at, what GEM does with that information, and how to build a testing system that feeds the machine the right signals. That’s what this article covers. The Andromeda Pipeline: How Your Ads Actually Get Delivered Before we talk about testing, you need to understand the delivery pipeline. This breakdown from Search Engine Land is the best plain-language explanation I’ve seen, and here’s my condensed version. When someone opens their feed, three AI systems work in sequence to decide what they see: Stage 1: Retrieval (Andromeda) Andromeda scans tens of millions of eligible ads and pulls out roughly 1,000 candidates for this specific user at this specific moment. It does this by analyzing your creative using computer vision and AI audio analysis, then matching it against the user’s behavioral patterns and intent signals. This is the make-or-break stage. If Andromeda doesn’t pull your ad into the shortlist, you don’t exist in that auction. Your budget, your bid, your targeting, none of it matters. You need to get through the gate first. Stage 2: Ranking (Meta Lattice) Those ~1,000 candidates enter the ranking stage. Lattice calculates expected value for each one: eCPM, predicted CTR, conversion probability, competitive bids. It picks the winner. According to Meta’s engineering team, Lattice delivered 10% metric gains and 6% conversion improvements. Stage 3: Learning (GEM) GEM (Generative Engagement Model) is the feedback engine. It’s 4x more efficient at driving performance than what came before. When someone converts (or doesn’t) after seeing your ad, GEM uses that outcome to improve future predictions. It also fills signal gaps when privacy restrictions block data by comparing your ad’s performance against billions of historical data points. What this means for you as a buyer: Andromeda decides IF your ad gets a chance. Lattice decides WHO wins. GEM decides how the system LEARNS from the result. Your job is to give Andromeda enough diverse creative signals so your ads pass the retrieval gate across many different user segments. Not just one. The Entity ID Problem (And Why 30 Ads Can Count as 1) This is the concept that changed how I think about creative production. And it’s the one most buyers still haven’t internalized. Andromeda doesn’t look at your ad count. It looks at conceptual uniqueness. Meta assigns each creative an internal identifier called an Entity ID based on its visual fingerprint. If you upload 30 ads that share the same template, same background, same visual structure with different text overlays, Andromeda collapses them into one Entity ID. One Entity ID = one ticket to the retrieval auction. If that single ticket fails for a particular user segment, your other 29 “different” ads never get a chance. They don’t exist in that auction. Performance data from admetrics.io suggests Creative Similarity Scores above 60% trigger retrieval suppression. 303 London’s diversity guide recommends keeping the index below 40%. This is huge. It means the old approach of “take winning image, test 15 headlines” actively hurts you now. Meta’s visual recognition models see an image with slightly different text overlays as essentially the same image. According to Social Media Examiner’s breakdown of the algorithm changes, if the system perceives a lack of diversity, it punishes your account with higher CPMs. The practical framework for ensuring unique Entity IDs. Before you build a new creative, ask three questions: Is the message different from what’s already running? Is the visual execution different (not just text on the same template)? Is the format different (static vs video vs carousel vs UGC)? If the answer is “no” to at least two of those, you’re probably getting grouped under an existing Entity ID. GEM, Lattice, and What They Mean for Your Testing Most articles about Andromeda stop at “creative is targeting now.” That’s true but incomplete. GEM and Lattice add two layers that directly affect how you should design tests. GEM learns from context, not just clicks. GEM doesn’t just track whether someone clicked or converted. It models the entire user journey. As this Medium breakdown explains, GEM compares your ad’s performance against billions of historical data points to estimate directional lift, even when privacy restrictions block the direct signal. For testing, this means early signals matter more than they used to. GEM starts forming opinions about your creative within the first few hundred impressions. A bad hook doesn’t just waste those impressions. It teaches GEM that your creative isn’t worth showing, and the system deprioritizes it going forward. Lattice evaluates across attribution windows. The Logical Position playbook explains that Lattice blends attribution windows at the architectural level. It evaluates success differently for high-ticket leads vs low-friction purchases because the system understands that timing and behavior vary by objective. For testing, this means you need patience with high-consideration products. A creative selling a $2,000 product might look terrible at day 3 but solid at day 14 once the longer attribution window kicks in. Killing it early means you never see the real performance. The Creative Similarity metric. Social Media Examiner reports that Meta now exposes Creative Similarity as a metric in Ads Manager. High similarity = higher CPMs because Andromeda views repetitive content as fatiguing. It also surfaces “Top Creative Themes” so you can see which angles are resonating (humor, social proof, nostalgia, etc.). Fair warning: because these metrics are new, Tara Zirker advises against over-optimizing for a specific score right now. Use them as directional signals, not hard thresholds. The Testing Framework That Works Under Andromeda Here’s the framework I use. It’s not theoretical. It’s what I run on my own campaigns and what I built TheOptimizer’s launching workflow around. Step 1: Build 8 to 12 conceptually distinct creatives. Not variations. Concepts. Use the PDA framework: Persona: Different buyer personas respond to different messages. Desire: Different motivations (save money, save time, look better, avoid risk). Awareness: Where they are in the journey (problem-aware, solution-aware, product-aware). Our guide on creating 10 angles for the same offer walks through this in detail. Step 2: Launch into a testing campaign (ABO). One creative per ad set. Clean data, no internal competition. Equal daily budgets ($20 to $50 per ad set). Broad targeting. Let Andromeda decide who sees what. Same optimization event as your scaling campaign. Step 3: Evaluate after 7 days using multi-metric scoring (see formulas below). Don’t just look at CPA. Under Andromeda, a creative with a high hook rate and decent engagement might be worth keeping even if the CPA is slightly above target on day 7. GEM is still learning. Step 4: Graduate winners to your scaling campaign (CBO). Move proven creatives into a CBO campaign with broad targeting and let Meta allocate budget across the winners. Step 5: Monitor for fatigue. Replace before the cliff. Under Andromeda, fatigue windows have compressed from 6+ weeks to 2 to 3 weeks. Your pipeline needs to be producing replacements before current winners decline. See our article on detecting creative fatigue early for the specific automation rules I use. 6 Custom Formulas for Evaluating Creatives in 2026 CPA alone doesn’t give you the full picture anymore. Here are the formulas I use to score creatives. Some of these I picked up from other buyers in the community, some I developed from looking at my own data patterns. 1. Hook Rate (video) Hook Rate = (3-Second Video Views / […]

$250+ spent. Zero conversions. Sound familiar? I’ve seen thousands of Meta ad accounts over the past few years. The pattern is almost always the same. It’s never one massive screw-up. It’s 2 or 3 things stacking on top of each other, quietly draining budget while you’re focused somewhere else. And the worst part is that most of these issues are invisible inside Ads Manager. Your dashboard shows clicks coming in. Maybe even a few conversions. But when you check your CRM, your leads, Shopify orders, or your finalized P/L reports the numbers don’t match. Something is off, and you can’t figure out what. Before you blame Meta, blame the algorithm, or start questioning your offer, let me walk you through the real reasons accounts break down. Not the beginner stuff like “pick the right objective.” I’m talking about the issues that experienced buyers run into when accounts that were printing suddenly go sideways. Your Conversion Data Is Lying to You I’m going to start here because everything else depends on this. If your tracking is broken, every optimization decision you make is based on garbage data. And garbage in, garbage out. The tricky part is that broken tracking doesn’t look broken. Your Events Manager still shows events firing. Conversions still appear in your dashboard. But those numbers are inflated, duplicated, or completely disconnected from reality. Here’s what’s actually happening in most accounts I’ve seen: Double-counting from Pixel + CAPI without deduplication. This is by far the most common issue. You set up Conversions API (which you should), but you didn’t implement event_id deduplication. So every purchase fires twice. Meta sees twice the conversions, optimizes for the wrong user profiles, and your reported CPA looks half of what it actually is. Meanwhile, you’re celebrating numbers that don’t exist. Ghost/test conversions from admin traffic. Your dev team, your marketing team, you personally, all hitting the thank-you page while testing. Each visit fires a conversion event. I’ve seen accounts where 15 to 20% of reported conversions were internal traffic. Events firing at the wrong funnel stage. A Purchase event firing on the product page instead of the order confirmation. An Add to Cart event triggering on page load instead of on button click. These seem minor. They’re not. Meta’s algorithm optimizes delivery based on who triggers these events. Feed it the wrong signals and it finds the wrong people. Low Event Match Quality eating your delivery. Check your EMQ score in Events Manager. Anything below 6 out of 10 means Meta is struggling to match your events to actual users. This directly affects how often your ads make it through Andromeda’s retrieval stage. Poor signal quality doesn’t just hurt your reporting. It actively reduces your ad delivery. Browser-only Pixel tracking now misses 20 to 40% of conversions thanks to iOS restrictions, ad blockers, and cookie consent banners. If you haven’t set up CAPI with proper deduplication, you’re flying blind. How to Fix Conversion Reporting Open Events Manager right now. Install the Meta Pixel Helper Chrome extension. Browse your site and watch what fires. Check for duplicates, wrong triggers, and missing events. Then verify your CAPI setup has event_id deduplication enabled. This isn’t optional in 2026. It’s the foundation everything else sits on. Andromeda Killed Your Targeting Strategy If you’re still running 8 ad sets with different interest stacks, each with a $15/day budget, I need you to hear this: that strategy died in 2025. Meta’s Andromeda update fundamentally changed how ads get delivered. The old system started with YOUR audience selections and then found people within them. Andromeda works in reverse. It starts with YOUR CREATIVE, reads it using computer vision and AI audio analysis, and then decides which users across Meta’s entire 3 billion user base are the best match. Your interest targeting? It’s mostly a suggestion now. Advantage+ Detailed Targeting can’t even be turned off for most performance goals. Meta uses your inputs as “hints” but goes wherever the algorithm thinks it’ll find conversions. This means two things for experienced buyers: First, audience fragmentation is now a liability. Splitting your budget across 5 to 8 narrowly targeted ad sets doesn’t give the algorithm enough data per ad set to learn. You end up with everything stuck in “Learning Limited.” The Confect Andromeda Study (covering 3,014 e-commerce advertisers and 115.7 billion impressions over the full 2025 calendar year) found that consolidated structures with broader targeting consistently outperform fragmented setups. Second, your creative IS your targeting now. An ad about “exhausted moms” will find exhausted moms regardless of your audience settings. An ad about “best SUV deals” will find SUV shoppers. The specificity lives in the creative, not in the audience panel. If your creative is generic (“Buy Now! Great Deals!”), Andromeda can’t figure out who to show it to, so it shows it to low-quality traffic and your CPA goes through the roof. I went deep on this in our article about how Andromeda affects your ad strategy. If you haven’t read it yet, do that after this one. Your Campaign Structure Is Starving the Algorithm Here’s a quick math problem that will tell you if this is your issue. Take your daily budget. Divide by your average CPA. Multiply by 7. If the answer is below 50, you’re starving the algorithm. Meta needs roughly 50 optimization events per week per ad set to exit learning phase. Below that, it never stabilizes, and you’re stuck in a permanent loop of erratic performance. Example: $150/day budget, $40 CPA. That’s 3.75 conversions/day, or about 26 per week. Not enough. You need to either consolidate (fewer ad sets, bigger budget each) or optimize for a higher-funnel event that generates more volume. The 2026 consensus from practitioners like Jodie Minto is pretty clear: the best-performing accounts now run 1 to 3 campaigns. One for testing new creatives (ABO with equal budgets per ad set). One for scaling proven winners (CBO with broad targeting). Maybe one for retargeting. That’s it. If you’re not sure which structure to use, our campaign structure best practices guide breaks down both options with specific examples. You’re Editing Campaigns Like It’s 2022 I get it. Your CPA spiked on day 2 and you panicked. You lowered the budget. Changed the targeting. Swapped a creative. Maybe all three. Congratulations, you just reset the learning phase. Again. Every significant edit triggers a reset. Budget changes over 20% (in general). Bid strategy swaps. New creatives added to an existing ad set. Targeting modifications. Each one restarts the clock and the algorithm has to relearn everything from scratch. The Cometly analysis documents this well: if you’ve been tweaking settings every other day, you’re essentially restarting the learning process each time. The algorithm never gets enough stable data to optimize. And here’s the part experienced buyers miss: the edit doesn’t need to be big to cause damage. Meta’s own documentation considers anything above a 20% budget change as “significant.” Going from $100 to $125? That’s 25%. You just triggered a reset. What actually works: Let campaigns run for 5-7 days minimum before touching them. If you need to adjust budgets, keep changes under 20%. Time them at the beginning of the day in your ad account’s time zone so Meta starts fresh with the new number. I wrote a whole piece on why killing campaigns too early hurts performance because I kept seeing buyers murder campaigns that would have been profitable by day 5 if they’d just waited. Stop resetting learning phases manually  TheOptimizer handles budget changes at the beginning of the day in your ad account’s time zone, keeps increases within safe thresholds, and only adjusts when the data justifies it. Your rules run every 10 minutes. Your campaigns stay stable.   Get Started for Free Your Creative Library Has Zero Diversity You uploaded 20 ads but Meta treats them as 3. This is the Entity ID problem, and it’s the thing most buyers still don’t understand about Andromeda. Meta assigns each creative an internal identifier (Entity ID) based on its visual pattern. If your 20 ads all use the same template, same background, same creator, Andromeda groups them under one Entity ID. In its eyes, you have one ad, not twenty. That means one ticket to the retrieval auction. If that ticket fails, the other 19 never get seen. Your budget wasted on volume that the algorithm treated as duplication. Data from admetrics.io shows Creative Similarity […]