Most media buyers launch new creative based on intuition. They guess what will work, spend two weeks testing, and discover they guessed wrong. Meanwhile, Meta publishes every ad your competitors are currently running — the exact creative, the exact copy, how long it's been running, and where it's being shown — for free, in a tool called the Meta Ads Library. The buyers who use it correctly skip the guessing phase entirely. Here's the exact system we use to extract competitor intelligence and turn it into winning creative in under an hour.
Meta Ads Library (facebook.com/ads/library) is a public database of every ad currently running on Facebook, Instagram, Messenger, and Audience Network. It was created in response to political transparency laws, but it applies to every advertiser — not just political ones. Any brand running a Meta ad has all of its active creatives visible here.
You can filter by country, platform, ad category, and specific advertiser name. For each ad, you see the creative (image or video), the copy, the landing page destination, and the date the ad started running. For political and social-issue ads, you also see audience demographics and spend ranges — unfortunately not for commercial ads, but the creative data alone is gold.
You can't see performance directly — but you can infer it. Ads that have been running for 60+ days are almost certainly winners, because no advertiser keeps losing creatives live. Ads that started running yesterday are tests. This signal alone is enough to reverse-engineer a competitor's best creative.
"We audited a competitor's Meta Ads Library and found one video ad that had been running continuously for 147 days. That was the unicorn — the creative that was carrying their entire account. We reverse-engineered the hook structure, applied it to our client's offer, and within 3 weeks had our own 90-day winner using the same pattern."
Here's the exact process we run at the start of every new client engagement. It takes 10-15 minutes and produces more actionable creative insights than a month of guessing.
Start with 5-10 direct competitors. Not "anyone in the same industry" — specific businesses that sell to your target customer. For a local med spa, that's other med spas in the same metro plus 3-4 national med spa chains. For a SaaS, that's direct alternatives in the same category.
Then add 3-5 adjacent competitors — brands that aren't direct competitors but sell to similar customers. These often reveal angles your direct competitors haven't thought of.
Go to facebook.com/ads/library, set the country to your target market, set the ad category to "All ads," and type the brand name. You'll get a list of every active ad from that advertiser. Sort by "Start date" to see the oldest ads first — those are the long-running winners.
For each competitor, look for ads that have been running for 60+ days. These are the winning creatives carrying their account. Screenshot each one and save the following in a spreadsheet:
This is where the real insight comes from. Once you have 20-40 winning ads from 5-10 competitors in a spreadsheet, patterns emerge. You'll notice things like:
These patterns are the distilled wisdom of every competing media buyer in your category. They spent years and hundreds of thousands of dollars figuring out what works. You get it in one afternoon of research.
Don't copy the exact ad — that's plagiarism and usually illegal. Instead, adopt the structural patterns: the hook format, the angle, the offer type, the CTA style. Then apply your own brand voice, your own creative assets, and your own unique value proposition.
Example: If every winning ad in your category opens with a customer testimonial clip and ends with a discount offer, your next test should also be a testimonial-opener with a discount offer — but using your actual customers and your actual offer. You're not copying, you're applying a validated pattern to your specific account.
Here's the exact checklist we run on every ad we save. Each of these is a signal of what's working in your category.
Does the ad open with a question, a statement, a visual pattern interrupt, a price, a statistic, or a promise? Track the hook type for every winning ad in your spreadsheet. You'll see one or two hook types dominate — those are the ones your category audience responds to.
What outcome does the ad promise? Is it weight loss, time savings, money savings, a transformation, a status bump, a feeling? Angles are what separate winning ads from losers, and the patterns across competitors will show you the 2-3 angles that actually work in your category.
Is it a discount (percentage or dollar amount), a free trial, a free consultation, a "first session" deal, a bundle, a payment plan, or something else? Winning offer structures cluster heavily by category — local services often use "first session $X" while ecom uses "X% off first order."
Is the ad static image, video, carousel, or collection? How long are the winning videos? Are they UGC-style or studio-produced? Do they have text overlays or voiceover or both? Answer these for every saved ad.
Most winning video ads on Meta are 15-30 seconds. But in specific categories (finance, insurance, SaaS), winners can be 60+ seconds because the educational component justifies the length. The pattern in your category will tell you which length to test.
Click through and look at where the ad actually sends traffic. Is it a dedicated landing page, a product page, a quiz funnel, a lead form, a Shopify PDP, or a full homepage? Direct-response brands almost always use dedicated landing pages or quiz funnels — generic homepage traffic is a losing pattern.
Switch the sort order to "Recently added" to see what a competitor is currently testing. New ads that appear and disappear within a few days are failed tests. New ads that appear and stay running are new winners. Watching a competitor's test cycle in real time is one of the most valuable uses of the Library.
Use the "Platforms" filter to see only ads running on Instagram Reels. This is crucial because winning Reels creative is structurally different from feed creative (vertical, sound-on, first-frame hook). Don't analyze them together.
Check the same competitor weekly and note how many active ads they're running. A jump from 5 to 50 active ads usually signals a scale-up or a new campaign launch. A drop from 30 to 3 signals they've killed a testing round and consolidated to winners. Both are strategic signals you can respond to.
Instead of describing a video to an editor in abstract terms, send them 3-5 screenshots of winning competitor videos from the Library. Tell them "match this hook structure, match this pacing, match this text overlay style." Editors work 5x faster with references than with briefs, and the output is more on-target.
When a competitor launches a new product, their first week of ads in the Library shows you exactly which hooks and angles they're testing. You can't directly compete, but you can learn from their research spend. Bookmark their Library page and check weekly.
Don't do it. Not only is it legally risky (ad copy is copyrightable), but Meta's algorithm penalizes near-duplicate creatives and the ad will underperform. Copy the structure and the angle, not the words.
An ad that started running yesterday is a test, not a winner. Only ads running 30+ days should be treated as validated. The Library is full of failed tests that look like active campaigns — don't mistake them for successes.
The biggest learnings often come from adjacent categories. If you sell cold plunge tubs, don't just look at other cold plunge brands — look at sauna brands, meditation apps, and fitness recovery products. Their winning angles often transfer directly because the customer profile overlaps.
Big brands run brand-awareness ads that look polished but are terrible blueprints for direct response. Mid-sized performance-driven brands in your category are a better source because their ads are built for CPA efficiency, not brand metrics.
Competitor creative evolves constantly. A quarterly audit catches you up on what's working now. A one-time audit becomes stale within 2-3 months. Schedule 30 minutes every 4 weeks to re-check your tracked competitors.
Here's exactly how we run competitor research at the start of every new client engagement. This takes under an hour and replaces what used to be a week of guessing.
Total time: about 90 minutes. Output: a validated creative direction based on what's currently working in the category, instead of a blank page and a guess.
Meta's creative environment moves faster than it ever has. Creative fatigue kicks in after 2-3 weeks, winning angles rotate quarterly, and what worked in your category 12 months ago is often dead now. The only way to stay current is to watch what's working right now, not what worked last year.
The Ads Library makes this watchable in real time. Every advertiser on Meta is voluntarily handing you their creative roadmap. Most media buyers still ignore it and launch based on intuition. The ones who use it consistently have a permanent advantage: they skip the guessing phase and start tests with validated patterns from day one.
Open facebook.com/ads/library right now, search your top three competitors, and filter for ads running 60+ days. You'll learn more about what wins in your category in the next 30 minutes than in the last 30 days of testing.