Ad spy tools occupy a strange position in the ecommerce stack. Nearly every brand over a certain size pays for one. Very few can say what they got from it last month, and fewer still can point at an ad they shipped because of it.
Part of that is the category’s fault — the pitch is usually “see what’s working for your competitors”, which is a promise about outcomes made by a product that only delivers access. But most of it is a workflow problem. These tools produce screenshots, and screenshots are not briefs.
This is an honest look at what the category does, what it can’t do regardless of price, and how to decide whether you need one. If you already know your way around the Meta Ad Library, skip to the section on what none of them can tell you — that’s the part that decides the question.
| What you want | Where it comes from |
|---|---|
| A competitor’s current ads | Free — Meta Ad Library |
| Top ads in your category | Free — TikTok Creative Center |
| Ads a brand ran and switched off | Paid archive |
| Discovery across advertisers you haven’t named | Paid search |
| Spend or ROAS for an ecommerce ad | Nowhere — it isn’t published |
| Why an ad works, as a brief | Your own reading, or a research agent |
What an ad spy tool actually is
An index, plus an interface
Strip away the positioning and the category is straightforward: a collection of ads gathered from social platforms, plus search, filtering, saved collections and — in the better ones — alerting when a tracked advertiser publishes something new.
The important structural fact is that the underlying creative is largely public data. Meta publishes every active ad. TikTok publishes top performers. A spy tool’s value is not privileged access; it is the indexing, the retention, and the interface on top. That’s a real product with real work in it, but it changes how you should evaluate one. You are buying convenience and memory, not intelligence — and convenience is worth paying for right up until the point where the thing you actually lack is time to read.
What they’re built on
Two things: the platforms’ own public libraries, and each tool’s own historical archive of ads it captured before they were switched off.
The archive is genuinely theirs and genuinely useful, because the platform libraries only show what’s live right now. Everything else — the current state of any advertiser’s creative — you can see yourself, today, without paying anyone.
Some tools layer on extra signals: which landing page an ad points to, how many active variants a brand is running, when an ad first appeared in their index. Those are useful, and they are all derived from the same public surface rather than from any performance feed.
The free sources most people underuse
The Meta Ad Library
The primary source for Facebook and Instagram, free, no account required, at facebook.com/ads/library. Search by advertiser, filter by country, and you get every ad currently active for that page along with the date each one started running.
Two limitations worth understanding before you conclude a paid tool is necessary. Spend and impression data is published only for political and social-issue advertising — for ecommerce, there is no performance data, from Meta or from anyone reselling Meta. And the library shows active ads only, so a test a competitor ran for three weeks in January is gone.
The first limitation applies equally to every paid tool in the category. The second is the one they legitimately solve.
A few things make the free library considerably more useful than most people find it. Search by page name rather than product keyword — keyword search is unreliable. Check more than one country if your competitor sells internationally, because the creative often differs and the differences are informative. And sort your reading by start date: an ad still live after sixty days in a competitive category has already told you most of what you wanted to know.
The TikTok Creative Center
TikTok’s equivalent leans in a different direction: rather than tracking specific advertisers, it surfaces top-performing ads with engagement signal attached, filterable by region, industry and time window.
That makes it better for reading a category than for stalking a competitor. If your market is creator-driven or skews young, it’s usually the more informative of the two free sources — and the ideas tend to appear there before they show up in a Meta feed, often by a month or two.
One more free source worth thirty seconds: the brands’ own organic feeds. Many test hooks organically before promoting the winners, so a recent TikTok or Reels post often shows you the raw version of an idea before it became an ad — and the comments underneath tell you, in customers’ own words, which objections your version will need to answer.
Set the time filter short. The point is to catch what is working now, not to admire something that won last spring, and TikTok creative conventions move faster than most categories’ underlying arguments do.
What paid tools genuinely add
Four things, and they’re real. Whether they’re worth your money depends entirely on scale.
A historical archive
The strongest argument for paying. Platform libraries are a snapshot of now; a spy tool’s archive is a record of what a brand has tried and abandoned.
That history answers questions the live view can’t: what did they run last Black Friday, how has their positioning shifted over eighteen months, what did they test and drop. For a brand doing serious competitive strategy work, that’s genuine value. For a brand that wants to know what’s working this month, it’s a nice-to-have.
Be realistic about how often you’ll use it. Historical analysis is the feature people buy the subscription for and the feature they open twice a year, usually while writing a deck.
Search across advertisers you haven’t named
Free libraries are built around “show me this advertiser’s ads”. Paid tools let you search across the whole index — by keyword, format, landing page pattern, or category — which surfaces brands you didn’t know were competing with you.
This is where the long tail lives, and the long tail is frequently where the newest ideas are, because smaller brands can’t afford to keep running last year’s winner.
It’s also the feature that most often justifies the price on its own. Knowing your five named competitors is table stakes; discovering the brand doing something genuinely different two rungs below you in the category is where the useful surprises come from.
Alerting
Being told when a tracked competitor launches something new, instead of remembering to check. Small feature, disproportionate effect on whether the research actually happens.
Most competitor research programs die of forgetting rather than of disagreement about their value. Anything that removes the remembering step earns more than its feature list suggests.
The caveat: an alert still lands in an inbox, and an inbox is not a workflow. Alerting fixes the noticing problem and leaves the reading problem entirely intact, which is the pattern for this whole category.
Boards, tagging and collaboration. Saving ads into organized collections, annotating them, sharing them with whoever writes the brief. Agencies get more from this than in-house teams, for obvious reasons.
It’s also the feature most likely to become a graveyard. A board with four hundred saved ads that nobody has opened since March is a very common artifact, and it is worth being honest with yourself about whether yours will be different.
If you do use boards, organize them by argument rather than by brand. “Ads that answer the durability objection” is a brief waiting to happen. “Competitor X” is a folder.
What none of them can tell you
This is the part that decides whether the category is right for you, because these four limitations apply to every tool in it, at every price.
Spend, revenue or profitability
There is no commercial performance data available for ecommerce ads. Not from Meta, not from TikTok, not from any third party reselling either. Anything presented as a spend estimate for a commercial advertiser is a model, not a measurement.
You can infer a great deal from duration and repetition — an ad running four months in a competitive category is almost certainly profitable — but that’s inference you can do yourself in a free library. It isn’t something you’re buying.
This is worth being firm about, because the category’s marketing frequently implies otherwise. If a dashboard shows you an estimated spend figure for an ecommerce advertiser, it is a model built on public signals, and you should treat it with roughly the confidence you’d give a guess.
Whether the ad is actually working
Related and worth separating. A tool can show you an ad. It cannot tell you the ad is good.
Big brands run mediocre ads for months out of inertia and brand budget. New advertisers run everything briefly because they’re still exploring. Duration is a decent proxy across a sample of twenty ads and five competitors; it’s unreliable on any single one, and no interface changes that.
The practical rule: never build a campaign on a single competitor ad, however impressive. Build on a pattern that repeats across independent advertisers, because that’s the only version of this signal that survives scrutiny.
Why the ad works
The most important gap. A spy tool shows you that a competitor is running a before-and-after with a founder quote and a 30% offer. It does not tell you which of those three elements is doing the work, what objection the ad is answering, or which awareness stage it’s aimed at.
That reading is the actual research, and in every spy-tool workflow it is unpaid labor performed by you, on a Sunday, in a spreadsheet nobody else will open.
It’s also the part with the highest return. Twenty ads read properly against a fixed checklist — hook, awareness stage, objection, proof, offer, format — will change what you make far more than four hundred ads skimmed. The scarce resource was never access. It was attention, and a search interface does not create any.
Whether it transfers to your brand
The last gap and the one that costs the most money. An ad that works for a market leader with high brand recognition frequently fails for a challenger, because it assumes knowledge your audience doesn’t have.
A competitor’s claim may also simply be unavailable to you: their “clinically proven” is a fact about their product, not a permission slip for yours. Copying it inherits a claim you may have to defend to a platform reviewer or a regulator, neither of whom will care where you found it.
The useful thing to take is always the structure — the objection being handled, the proof type, the pacing, the offer shape. Structure transfers. Claims and assets don’t.
The transfer question also has a price dimension. A hook built around “the affordable alternative” reads completely differently coming from the cheapest brand in a category than from the most expensive, and an ad that works for a $19 product frequently collapses at $89 because the objection being answered isn’t the same one. Read competitor ads with their price point in mind, not just their creative.
The workflow problem
Screenshots aren’t briefs
Here is the honest failure mode of the whole category, and it has nothing to do with data quality.
You open the tool. You save twelve interesting ads. You feel informed. Then the brief has to be written, the designer has to be told what to make, and the twelve screenshots turn out not to contain a single instruction. The gap between “look at these” and “make this” is the entire job, and it sits on your desk.
Bridging it takes a specific, slightly tedious skill: reading each ad for what it is doing rather than what it looks like, then writing that as a sentence you could hand to a designer. “Answers the ‘won’t survive a wash’ objection with a 40-wash comparison shot” is a brief. A screenshot of the same ad is a mood board entry, and nobody has ever briefed from a mood board successfully.
The research that never gets used
The natural consequence: the tool gets checked enthusiastically for three weeks and then quarterly, then not at all, and the subscription renews for a year past the point of usefulness.
That isn’t laziness. It’s a workflow with a manual step in the middle that nobody has capacity for. Any competitive research process that depends on someone reliably finding a spare morning is a process that will stop.
The test of whether your tool is earning its keep is not whether you log in. It’s whether an ad you shipped last month traces back to something you found in it. If you can’t name one, the subscription is a habit rather than a lever.
Run that check quarterly. It takes two minutes and it is the only honest evaluation of a research tool, because every tool in this category looks valuable while you’re using it and only reveals itself in what shipped.
When you need one, and when you don’t
Reasonable cases for paying
You track more than about ten competitors. You do genuine strategic work with historical creative — positioning shifts, seasonal patterns, category evolution. You’re an agency and need shared, annotated boards across clients. You want discovery across advertisers you haven’t named.
Those are real jobs, and free libraries do them badly. If two or more describe you, the subscription is straightforwardly worth it and the rest of this article isn’t about you.
Even then, treat it as one input rather than the research program. The archive tells you what a brand did; it still won’t tell you why it worked or whether it transfers, and those remain the two questions that determine what you should make.
Cases where the free libraries are enough
You follow five competitors and want to know what they’re running now. You’re looking for creative direction rather than a competitive archive. You already struggle to find time for the reading, in which case a better search interface solves nothing — the constraint is downstream of search.
For most small ecommerce brands, that second list is the accurate one. The free libraries are genuinely sufficient for the job they think they’re paying to do — and the money is better spent on producing creative than on looking at other people’s.
There is also a quieter cost to the paid version: it makes research feel like a destination. Opening a tool full of beautifully organized competitor ads produces a strong sense of progress, and that feeling is not correlated with whether anything reaches an ad account.
The alternative: research that ends in creative
The premise worth questioning is that competitive research is a browsing product at all. What you actually want is not a better way to look at ads. It’s the conclusion, and then the ads.
The premise worth questioning
Every product in this category is built on the assumption that if you could just see enough ads, you’d know what to make. Twenty years of ad libraries suggest otherwise. The brands who consistently ship good creative are not the ones with the best search tools; they’re the ones with a habit of reading a handful of ads properly and a pipeline that can act on what they find within the week.
So the useful question isn’t “which spy tool” but “where does my process actually break”. If it breaks at noticing, buy alerting. If it breaks at remembering, buy the archive. If it breaks — as it does for most small brands — between the screenshot and the finished ad, no amount of indexing helps, because the missing step is interpretation followed by production.
Reading, not indexing
Wisry’s AdClone competitor research runs a research agent across the Meta and TikTok ad libraries and reads what it finds — hook, awareness stage, objection, proof type, offer placement, format — then groups the recurring patterns and returns six campaign angles for your own brand.
Every angle keeps the source creatives attached, so you can check the reasoning before funding it and diagnose it afterwards if it doesn’t work. That last property is the one no board of saved screenshots gives you: when an angle underperforms, you can see exactly which competitor ad convinced you to try it, and retire the assumption rather than the whole approach.
The claim check matters as much as the sourcing. Concepts are written against your own product facts, so an assertion a competitor makes about their product gets flagged rather than inherited — which is the specific risk that makes naive ad-copying expensive.
Ending in an ad, not a folder
The other half is what happens next. Angles feed directly into creative — static ad generation for testing messages at ad-set scale, scene-by-scene video for the messages that prove out — produced against your own brand and product facts, with a claim check so a competitor’s assertion never quietly becomes your promise.
The manual step in the middle disappears. That’s the actual difference, and it matters more than any feature comparison, because the manual step is where competitive research programs die.
How to choose
Ask one question: what will you do with it on a Tuesday morning?
If the answer is “check what four competitors launched and note anything new” — the free libraries do that, and the forty-five minutes it takes is the real cost, not the subscription. If the answer is “search eighteen months of archive across sixty advertisers” — pay for a spy tool, that’s exactly what it’s for. And if the answer is “turn what’s working into ads I can actually run this week” — you’re describing a different category, and no amount of better search will get you there.
There is also a middle path worth naming: keep the free libraries for weekly awareness, skip the spy subscription, and put the money into producing more creative. For a brand under a few thousand a month in spend, that reallocation is almost always the higher-return choice, because you are constrained by output rather than by information.
The tools are fine. The question is whether looking is your bottleneck. For most brands, it isn’t; the bottleneck is everything between looking and shipping.