Ads Spy Guide 2026: How to Research Competitors Without Guessing

Ads spy is the practice of systematically studying competitor advertising before writing a single line of copy. Not scraping, not guessing — structured observation of what brands already spend money proving works.

TL;DR: Ads spy research means pulling real in-market ads from public libraries across Meta, TikTok, YouTube, and LinkedIn, then extracting the patterns that explain why those creatives survive. This guide covers the full workflow: where to look, what to capture, how to turn raw ads into testable hypotheses, and when to move from manual research to an ads spy tool that does the retrieval for you.

This is not a roundup of tools. That guide exists — you can read ad spy tools compared. This is the methodology: how to read an ad, what signals actually matter, and how to build a swipe file that earns its keep.

If your current ads spy process is "browse the Meta Ad Library for ten minutes and screenshot what looks cool," you're collecting noise. The structured version takes the same public data and extracts information: run length, creative format, offer framing, hook type, audience signals. The difference between those two outcomes is process, not access.

Most teams who approach ads spy research for the first time focus entirely on the creative surface — they see a successful video ad and note its color palette or background music. That's the wrong level. The insight that makes your own ad better lives one layer deeper: the angle, the problem framing, the offer structure. A competitor's specific UGC video is irrelevant; the fact that they've been running UGC problem-solution formats for 90 days without rotating them out is very relevant.

What Ads Spy Actually Means

The phrase gets used loosely. For this guide, ads spy means accessing publicly disclosed advertising — the ad transparency databases every major platform publishes — and analyzing it systematically. No unauthorized access. No scraped data.

Every platform with scale now runs a public ad library:

• Meta Ad Library — facebook.com/ads/library — Facebook and Instagram ads, required by EU transparency law and voluntary global coverage
• TikTok Creative Center — ads.tiktok.com/business/creativecenter — top-performing TikTok ads with engagement signals
• LinkedIn Ad Library — linkedin.com/ad-library — B2B ad transparency since 2022
• Google Transparency Center — adstransparency.google.com — political and issue ads; commercial display coverage growing

These libraries exist because regulators demanded it. That's also why the data is shallow: platforms publish what they're legally required to, not what you'd want as a researcher. Impression estimates are bucketed. Date ranges are approximate. Some filters break depending on your query.

The first thing to understand about ads spy research is that every data point is an estimate. That doesn't make it useless — it makes it directional. You're looking for signals, not certified metrics. The value compounds when you observe the same signal across multiple competitors rather than reading too much into a single advertiser's behavior.

Why Ads Spy Research Has Real ROI

Most teams skip structured competitor ad research because it feels slow. It isn't. Here's the math:

A Facebook ad creative test costs between €400 and €1,200 in learning budget depending on your audience size and bid strategy. A learning phase that exits on a losing creative burns that spend confirming something you could have filtered out by looking at what competitors don't run anymore.

Ads spy research doesn't replace testing. It front-loads the filter. You test ideas that survived competitive selection pressure — ads that ran long enough to prove something. Creative testing then becomes hypothesis-driven rather than spray-and-pray.

The DTC ad intelligence frameworks that work at scale all share this foundation: research before spend, then test fast. Without the research step, you're optimizing the wrong variables.

For competitive intelligence at agency scale, the ROI argument is even cleaner. A single insight — that a competitor dropped UGC and shifted to static — can redirect three weeks of production. Teams that run structured competitor ad analysis consistently report fewer wasted test cycles and higher first-flight win rates on new creative. That's the natural result of testing fewer losing ideas.

Where to Pull Ads Spy Data

Start with Meta Ad Library for most categories. Coverage: every active ad on Facebook and Instagram, plus a 7-year archive for political/social ads. For commercial ads, retention is typically 30-90 days post-deactivation.

What to filter by: advertiser name (gets all their active ads), country (US, UK, DE, AU are densest), media type (video for UGC patterns; image for static offers), and active-only vs. all (deactivated ads tell you what didn't convert).

The weakness of the native Meta library: no keyword search, no sorting by run length, no impression estimates for commercial ads. It's browsable, not searchable. This is where Meta ad library research tools pick up the slack — they add the search, sort, and filter layers the native UI lacks.

For TikTok, the Creative Center is more marketer-friendly: it surfaces trending ads by industry with rough CTR and engagement signals. LinkedIn's ad library is accessible via the "Why am I seeing this" menu on any LinkedIn ad, or via LinkedIn Ad Library native search — limited but functional for B2B category research.

The harder problem is cross-platform: if a competitor runs the same offer on Meta, TikTok, and LinkedIn, which version performs best per platform? The native libraries don't answer that. You'd need to pull each one separately and correlate manually — or use a multi-platform ad research tool that normalizes the data. AdLibrary's unified ad search solves this: one query surface across platforms, with platform as a filter rather than separate research sessions. For teams that run cross-platform strategy, that cuts a half-day ads spy research session to 15 minutes.

How to Read a Competitor Ad

Pulling the ad is step one. Reading it correctly is the actual skill.

Most people study the wrong layer. They copy the visual treatment or the CTA button text. What actually matters is the creative angle: the specific belief or problem the ad is built around.

Break every ads spy find into five components:

1. Hook type — What stops the scroll? Options: pattern interrupt (unexpected visual), direct call-out ("If you run Facebook ads…"), social proof open ("We spent €2M testing this"), curiosity gap ("The one thing your competitors stopped doing").

2. Problem framing — What problem does the ad name? The ad copy structure usually follows: name the problem, amplify it, solve it.

3. Offer angle — Price anchor, trial, guarantee, transformation promise? The specific angle tells you what your competitor believes converts in this category.

4. Format choice — Static, video, carousel, UGC, talking head, screen recording? Format is a hypothesis about what the audience engages with.

5. Run length — An ad running 60+ days on a spend-capable account is a proven winner. An ad that appeared last week and disappeared is an experiment.

When you study ads this way, you're not copying creatives — you're extracting competitive hypotheses. The ad creative you build can look completely different. The insight is: your competitor has been paying to prove "guarantee-led offers work on cold audiences in this category." That proof is worth more than the specific creative execution.

The Ads Spy Workflow, Step by Step

This is the process used in structuring competitor ad research workflows that high-velocity teams run on a weekly cadence.

Step 1: Build a competitor tracker. List 5-8 direct competitors plus 3-4 aspirational comparables in adjacent categories that share your audience. For each, note their Facebook Page ID, TikTok handle, and any YouTube brand channels. Page IDs unlock cleaner access via saved ads workflows.

Step 2: Pull all active ads for each advertiser. Capture: ad creative, copy, CTA, active date, placement, media format. Don't curate yet — capture everything first.

Step 3: Sort by run length. In the native library, this requires manual estimation. In a proper tool, run length is filterable. Goal: isolate ads that have survived ≥30 days.

Step 4: Categorize by creative angle. Group ads by the hypothesis being tested: offer-led, problem-led, social-proof-led, authority-led, curiosity-led. Within each group, note which formats appear most often.

Step 5: Map to your roadmap. For each surviving angle you haven't tested, add it to your creative brief backlog. Don't copy the execution — write a brief for a version built on your brand's assets and claims.

Step 6: Track monthly. Ads that were running last month and stopped are as informative as ads that are still live. Use ad timeline analysis to spot drop-offs. For a detailed SOP format, see competitor ad research strategy.

Building a Swipe File That Earns Its Keep

A swipe file is only useful if it's searchable and tagged. Screenshots in a Google Drive folder are not a swipe file — they're a graveyard.

A working swipe file tags by: hook type (pattern-interrupt, direct callout, social proof, transformation), format (static, UGC video, talking head, carousel, DPA), platform (Meta feed, Meta stories, TikTok, YouTube pre-roll), offer angle (guarantee, trial, price anchor, feature-led), and run length at capture — a 90-day ad is categorically more valuable than a 5-day one.

The creative strategist workflow built around a structured swipe file means research time is pull-from-archive rather than start-from-scratch. When a brief lands, you're searching a curated database of competitive-pressure-tested ideas, not starting from zero.

AdLibrary's saved ads feature handles this in-app: tag, organize, and share saved ads across team members. The tags persist. The creatives don't disappear when the advertiser pauses the campaign. Use the ad budget planner to model what you're currently spending in research hours vs. what automated retrieval would cost at your ads spy research volume — the ROI case usually resolves quickly.

Signal vs. Noise in Ad Run Length

An ad running for 7 days proves nothing. The advertiser might be testing. The budget might be small. The learning phase might not have exited.

An ad running for 90+ days on a spend-capable account has survived budget reallocation decisions, creative refresh pressure, and the algorithm's preference for novelty. At that point, the creative has proven itself against real money.

The threshold worth using in your ads spy research:

• <14 days: Experiment, ignore for research purposes
• 14-29 days: Candidate — watch for growth in run length
• 30-89 days: Proven — worth studying the angle
• 90+ days: Control-class — this ad has likely been iterated from a winner

Ad timeline analysis in a proper tool shows the full run-length history: when a creative appeared, whether it was paused and relaunched, and when it was retired. An ad that was live last year, paused for two months, and reactivated is almost certainly a proven control.

The scaling decisions with ad library signals post goes deeper — specifically how to use timeline data to infer competitor budget behavior and creative testing velocity.

Ads Spy Across Platforms: What the Native Libraries Miss

Meta vs. TikTok signals: On Meta, run length and format distribution are the primary signals. Meta's ad frequency mechanics mean long-running ads are either refreshing creatives or targeting cold audiences at scale — both signals of performance. On TikTok, hook rate is the primary signal. The Creative Center surfaces ads by engagement and view-through, which Meta's library doesn't. The first 3 seconds of every TikTok ad are the hypothesis being tested.

TikTok's ad format conventions differ from Meta — spark ads (boosted organic content) look different from pure paid. The IAB's digital video ad format guidelines provide useful context for normalizing format classification across platforms when running multi-platform ads spy research.

What's missing across all libraries: Impression volume (Meta buckets commercial impressions), spend estimates (not published for commercial ads), actual performance data (CTR, CVR), and a normalized cross-platform view. This is the gap that paid ad intelligence platforms solve.

Meta's free API is fine for one platform. The moment you add TikTok, YouTube, or LinkedIn data into the same query, you need something else. AdLibrary's API — the Business-tier offering — addresses multi-platform data normalization specifically. It's not a replacement for Meta's free API; it's the upgrade for when Meta's API stops being enough: richer per-ad fields, cross-platform in a single call, no app-review friction. At €329/mo on the Business plan, it's built for teams running programmatic ads spy monitoring or AI agent workflows that need structured data at volume. For manual research, the ad-detail-view feature surfaces richer metadata without requiring API integration. Use the CTR calculator to sanity-check benchmark assumptions against observed proxy signals.

Common Mistakes in Ads Spy Research

Copying the execution. The insight is the angle, not the creative. If you copy a competitor's UGC testimonial format verbatim, you're creating a knock-off with none of the brand equity that made it convert. Extract the hypothesis; write your own brief.

Only studying active ads. Deactivated ads tell you what your competitors tested and killed — valuable negative data. An angle that multiple competitors dropped probably doesn't convert in this category. That's useful to know before you spend money confirming it yourself.

Ignoring small advertisers. Category newcomers often run highly efficient creative because they can't afford waste. A brand spending €2k/week with a 60-day-old ad has found something that works. Don't only focus on the obvious names in your ads spy sessions.

Skipping the creative brief. Ads spy research without a creative brief output is just note-taking. The brief is where the insight becomes actionable: hypothesis, format, hook, offer, audience, KPI.

Research frequency too low. Monthly is fine for trend-level analysis. Competitive pressure can shift in a week. High-velocity teams run 15-minute weekly scans of their top 3 competitors to catch new creatives and dropped campaigns before the information goes stale.

Ads Spy at Scale: When Manual Becomes the Bottleneck

Manual ads spy research works for 5 competitors × 2 platforms. When you're running agency-level competitor monitoring across 20+ advertisers on 4 platforms, retrieval is the bottleneck — not analysis.

At that point: automated retrieval (scheduled pulls from public libraries), structured storage (normalized data model across platforms, not a folder of screenshots), and an analysis layer (search by angle, filter by run length, group by format) all become necessary infrastructure.

This is exactly the workflow the from ad library research to creative brief in 60 minutes guide is built around. The creative strategist research workflow with an ad library covers the practitioner version — from first pull through brief sign-off.

AI ad enrichment becomes the leverage point at scale: automatic tagging of hook type, offer angle, and format classification. What takes an analyst 3 hours to classify manually takes seconds when enrichment runs on ingest. The Pro plan at €179/mo covers 300 credits/month (search + enrichment). Business at €329/mo adds API access for programmatic retrieval workflows. The media buyer workflow covers how to run this across client portfolios without the research cadence collapsing under volume.

Frequently Asked Questions

What is ads spy and is it legal?

Yes — when it uses publicly available ad transparency libraries that platforms are required to maintain. Meta, TikTok, LinkedIn, and Google all operate public ad libraries by regulatory requirement. Studying, saving, and analyzing public ads is legal. The line is terms of service: tools that scrape closed surfaces or impersonate users to capture private targeted ads operate in a gray-to-black zone.

How often should I run ads spy research?

A weekly 15-minute scan of top competitors plus a monthly deep-dive is the right cadence for most teams. Weekly scanning catches new creatives and dropped campaigns. Monthly analysis identifies pattern shifts — when a competitor moves from UGC to static, or drops a specific offer angle they were running heavily.

What's the difference between Meta Ad Library and a paid ads spy tool?

Meta's free Ad Library lets you browse a single advertiser's active ads on Meta. A paid ads spy tool adds keyword search, sorting by run length, multi-platform coverage, timeline analysis, saving and tagging, and API access for programmatic workflows. The native library handles occasional manual research; a paid tool becomes necessary when you're tracking more than 5-8 advertisers regularly.

Can ads spy research replace creative testing?

No. Ads spy research front-loads the filter — you test ideas that already survived competitive pressure rather than starting from scratch. But your audience, brand, offer, and account history all differ from your competitors'. Research narrows the hypothesis space; testing validates which hypothesis works for you.

Which platforms should I prioritize for ads spy research?

Start with the platforms where you actually buy ads. If you run Meta and TikTok, research both — TikTok surfaces engagement signals the Meta library doesn't. LinkedIn is worth adding for B2B campaigns. YouTube is harder to research natively but accessible via the Google Transparency Center and third-party tools.

From Research to Roadmap: Closing the Loop

Every ads spy session should produce a deliverable. Without one, the research has no operational value. That deliverable is a creative brief.

The shortest version of the pipeline: pull active + recently deactivated ads for 5-8 competitors, filter to run length ≥30 days, categorize by creative angle (3-5 categories), identify angles you haven't tested, then brief one ad per untested angle — hypothesis, format, hook, offer, audience, success metric. A single paragraph per angle is enough to hand to a copywriter. The value is the hypothesis, not the execution spec.

For a deeper look at creative angle classification, guide to analyzing competitor ad creative strategies covers the full framework including how to identify control-class creatives.

Map competitor angles to your backlog. Every proven angle (≥30 day run length) that you haven't tested becomes a backlog ticket. Assign priority by: how many competitors run it (more = stronger signal), how long the longest-running version is, and fit with your brand's claims.

Flag the graveyard. Angles multiple competitors tested and killed are negative signal. Don't add them to the backlog — add them to an "explained away" list with a note on why they probably didn't convert. This is one of the most underused outputs of ads spy research: a curated list of creative angles the market already tested and rejected.

Set a review cadence. Competitive creative research goes stale. A quarterly review to retire old data and refresh recent captures is worth two hours of calendar time.

The creative intelligence that separates high-velocity teams from everyone else isn't access to better tools — it's process discipline: consistent research, structured output, and a backlog that connects competitive observation to test roadmap.

To track competitor ad budget behavior at a deeper level, watch for: creative execution refreshing while core offer stays the same (scaling a winner), multiple similar-angle ads running simultaneously (A/B testing at scale), or rapidly cycling short-run creatives combined with increasingly aggressive offers (a brand struggling to find its control creative). The best Facebook ad intelligence tools guide covers which tools surface this timeline data most clearly; high-performance ad intelligence and creative research platforms covers enterprise-tier options for agencies running ads spy monitoring at portfolio scale.

If you want to run this workflow at scale — tracking dozens of advertisers, normalizing data across platforms, and feeding enriched ad metadata directly into brief templates — start with the Starter plan at €29/mo for manual ideation and creative inspiration, or move to Pro at €179/mo when you're tracking multiple competitors on a weekly ads spy cadence. The research is in the public libraries. The process is yours to build.