Researching competitor ads means pulling their live and past creative, targeting signals, and landing pages from public ad libraries and spy tools, then reading the pattern across dozens of ads rather than judging any single one. Done well, it tells you which angles, offers, and formats a competitor keeps running — and running ads is the strongest signal you'll get, because nobody keeps paying for a losing creative. This guide walks through the process end to end, from the free platform tools to the paid spy tools that fill in the gaps.
Before you start
You don't need a big budget or a specialized team to do this — most of the tools below are free, and the paid ones have entry tiers under $100/month. What you do need is a clear question. "What is this competitor doing" is too broad and you'll end up with a folder of screenshots and no conclusions. Narrow it to something like "what hook does this competitor use for cold traffic on Meta" or "which landing page do they send paid search clicks to." That question determines which platform you check first and what you're looking for once you're there.
You'll also want a place to log what you find — a spreadsheet is enough. Track the competitor, platform, ad format, headline or hook, offer, landing page, and the date you saw it running. Without a log, patterns across dozens of ads are hard to hold in your head, and you'll end up re-checking the same competitor next month with no record of what changed.
Step-by-step: how to research competitor ads
1. Start with the official ad transparency libraries
Meta, Google, and TikTok all publish free, searchable archives of ads currently running or recently run on their platforms, and this should be your first stop because the data comes straight from the source with no scraping involved.
The Meta Ad Library lets you search by advertiser name or keyword and see every active ad across Facebook and Instagram, including creative, ad copy, and — for political and issue ads — spend ranges and audience demographics. For most commercial advertisers you won't get spend data, but you will get every live variant they're running, which is often more useful.
Google's Ads Transparency Center covers Search, Display, YouTube, and Shopping ads and works the same way: search an advertiser and browse what they're currently serving. TikTok has its own Commercial Content Library for the same purpose. None of these require an account, and all three are the most reliable single source because there's no risk of the data being stale or scraped incorrectly.
2. Search by keyword, not just by brand
Searching a competitor's name only shows you ads where they've used their own brand as the identifying text, which understates how much they're actually running. Search the product category and offer language instead — "email marketing software free trial" instead of just the competitor's name — and you'll surface ads from advertisers you didn't know were competing with you at all. This is often where the more useful research happens, because your assumed competitor set is usually smaller than your actual one.
3. Pull historical and cross-platform data with a spy tool
The official libraries only show what's live now (or recently, for Meta). To see how long a creative has run, whether it's been reused across platforms, or what a competitor tested and dropped, you need a dedicated ad intelligence tool. Options like AdBeat, Anstrex, BigSpy, or PowerAdSpy index ads across display, native, and social networks and let you filter by advertiser, network, ad format, and date range. Most offer a free or low-cost trial tier, which is enough to do a focused pass on two or three competitors.
What you're looking for here is longevity: an ad that's run for eight weeks with no creative changes is very likely profitable, because ad budgets get pulled from anything that isn't earning its spend back. A tool that's run for two days tells you almost nothing — it might have been killed for underperforming.
4. Check where the traffic actually lands
Click through to the landing page for each ad you log (using a private browser window, since retargeting pixels will otherwise start following you). Note the offer, the call to action, whether it matches the ad's promise, and whether it's a dedicated page or the competitor's homepage. A mismatch between a specific ad promise and a generic homepage is a signal the campaign might be early-stage or under-optimized — worth knowing if you're deciding whether to compete on that angle.
Tools like SimilarWeb or SEMrush's traffic analytics can round this out by estimating a competitor's overall paid vs. organic traffic split, which tells you roughly how reliant they are on paid acquisition for that page.
5. Look at the pattern across the set, not any single ad
Once you have ten or more ads logged for a competitor, look for what repeats: the same three hooks reworded, the same discount threshold, the same format (video vs. static) dominating. That repetition is the actual finding — a single clever ad might be an experiment, but a hook that shows up in six variants over two months is a validated angle. This is the step people skip, and it's the one that actually produces a usable insight instead of a swipe file.
Example log entry:
Competitor: Acme SaaS
Platform: Meta
Format: Video, 15s
Hook: "Stop losing leads to slow follow-up"
Offer: 14-day free trial, no card
Landing page: /demo (dedicated, matches ad)
First seen: 2026-06-02
Still running: yes, as of 2026-08-15
How to tell it's working
You'll know the research is paying off when it changes a specific decision — not when you have a full spreadsheet. Good signs: you can name the two or three angles a competitor keeps testing, you can point to a landing page pattern you'd want to test yourself, and you can explain why an ad has likely been profitable (mainly: it's still running after weeks) rather than guessing from the creative alone.
If a network like Adsy is part of how you buy or plan your own campaigns, this same competitor data is what you'd bring into that planning — it tells you which formats and angles are already proven in your category before you commit budget to testing them cold.
A weak result looks like a folder of screenshots with no notes on run duration, offer, or landing page — interesting to look at, but not something you can act on. If that's where you've ended up, go back to step 5 and force yourself to summarize the pattern in one sentence per competitor before moving on.
Troubleshooting
You can't find any ads for a known competitor. Confirm you're searching the right entity name — many companies advertise under a parent brand, a regional subsidiary, or a D2C sub-brand that doesn't match their corporate name. Also check whether they run mainly on a platform you haven't checked yet; a company can look inactive on Meta while running heavily on YouTube or native networks.
The ad library shows an ad but no spend or targeting data. This is expected for most commercial (non-political) ads on Meta and for nearly all ads on Google's and TikTok's libraries — the platforms limit spend and targeting transparency to political/issue advertising in most regions. Use run duration and creative volume as your proxy for spend and performance instead.
Spy tool results look stale or duplicated. Third-party spy tools scrape and index ads on their own schedule, so there's a lag, and the same ad can appear multiple times under slightly different IDs. Cross-check anything important against the official ad library before you act on it — the platform's own data is always the source of truth.
You're finding creative but no clear pattern. This usually means you're logging too few ads, or you're logging across too many competitors at once. Narrow to one competitor and pull at least 15–20 ads before looking for a pattern; fewer than that and normal creative variation looks like noise.
FAQ
Is it legal to look at competitor ads?
Yes. Ad libraries from Meta, Google, and TikTok are public by design, and third-party spy tools index only publicly served ads. You're not accessing anything private or gated.
How often should I re-check competitor ads?
A monthly pass is enough for most categories — ad creative doesn't turn over daily, and a shorter cycle mostly adds noise. Increase frequency around known high-spend periods like Black Friday or a competitor's product launch.
Do free tools give me enough, or do I need a paid spy tool?
The official libraries alone are enough if you only need to know what's currently running. A paid spy tool earns its cost when you need historical data, cross-platform coverage, or filtering across many competitors at once.
Can I see how much a competitor is spending on ads?
Rarely, and only in specific cases — Meta discloses spend ranges for political and issue ads, and some spy tools estimate spend from indirect signals like traffic and ad volume. Treat any spend estimate as directional, not precise.
What's the biggest mistake people make researching competitor ads?
Copying a single ad instead of identifying the pattern across many. One ad can be an experiment that failed; a hook or offer that repeats across weeks and variants is the one worth learning from.
Conclusion
Researching competitor ads is a repeatable process, not a one-off scan: pull live creative from the official ad libraries, fill in history and cross-platform coverage with a spy tool, check where the traffic actually lands, and look for the pattern across many ads rather than judging any single one. The strongest signal isn't a clever headline — it's an ad that's still running weeks after it launched.
Key takeaways
- Start with the free, official sources — Meta Ad Library, Google Ads Transparency Center, TikTok Commercial Content Library — before paying for a spy tool.
- Search by category and offer language, not just competitor names, to find advertisers you didn't know were competing with you.
- Run duration is your best proxy for performance; a creative that's been live for weeks is very likely profitable.
- Always check the landing page, not just the ad creative — mismatches are a signal worth noting.
- Log every ad you review with date, offer, and landing page so patterns are visible after ten or more entries, not just one.
