Every tool for watching competitors sits in one of three lanes: the free transparency tools the platforms publish, the paid spy tools that model and archive on top of them, and done-for-you research where someone else does the work and hands you the read. Most buying mistakes in this category come from expecting one lane to do another lane's job.
Here is the straight version of what each lane delivers, what it actually costs, and where it belongs. We sell in the third lane, and we will say so plainly when we get there.
Lane one: the free transparency tools
The Meta Ad Library and the Google Ads Transparency Center are the primary sources. Not samples, not estimates: the actual ads, published by the platforms that serve them, current as of now. LinkedIn's ads tab on company pages covers B2B, and TikTok publishes its own commercial content library. Everything every other tool in this market sells is built on top of these sources or modeled around them.
Their strengths are exactness and currency. What they show is true and live. Their costs are of a different kind. They show no spend and no performance for commercial ads. They keep no history: a commercial ad that stops running disappears, so a one-time look is a photograph of a moving object. And they are manual. A proper pass across a ten-company set, on the method in our competitor ad analysis workflow, takes hours, and it needs repeating monthly to stay true.
The right user for this lane is anyone doing a first serious read of their market, and any team disciplined enough to repeat it. The tools are free, the method is public, and the primary source beats every derivative of it.
Lane two: the paid spy tools
Above the free sources sits a market of subscription tools in two rough families. Keyword and traffic tools estimate which search terms competitors rank and bid on, and publish cost-per-click and volume figures for those terms. Ad-spy databases archive ad creative over time and let you search it by advertiser, keyword, or format.
What they genuinely add: history, which the platforms refuse to keep, so an archive that was watching last year can show you what a competitor ran last year. Scale, when you track dozens of companies and clicking through libraries by hand stops being realistic. And the keyword tools' cost data is a legitimate input for estimating competitor ad spend, because it prices the market itself rather than guessing at any one company.
What to hold against them: the modeled numbers. Spend and traffic figures in these tools are estimates presented with more confidence than they deserve, and coverage is uneven, so an advertiser can be materially active while barely appearing in a given database. Archives also lag and miss, especially outside the biggest markets. The practical rule: trust these tools on what they observed, the creative archive and the market's price data, and discount what they inferred, the per-company spend and traffic figures. And before paying, check the tool against a market you already know. Coverage gaps reveal themselves in minutes.
The right user is an in-house team that lives in this weekly: a media buyer or SEO lead who will actually open the tool enough to earn the subscription. For an occasional question, the free lane answers better and the model numbers mislead more than they help.
Trust what the tool observed. Discount what it inferred.
Lane three: done-for-you
The third lane is not software. Someone runs the collection on the primary sources, keeps the history the platforms discard, and, the part neither software lane offers, reads the result and tells you what it means: who is committed, what the market's messages agree on, and where the unclaimed ground is.
This is where Muffin Intel sits, so weigh what follows accordingly. Muffin Intel is our in-house tracking system. It reads the same transparency sources described above, continuously, and holds 55,085 tracked ads across 2,823 advertiser identities. It is not magic access: nobody has spend or performance data for commercial ads, including us, and a vendor claiming otherwise is describing a model. What the corpus buys is history the free tools discard, resolved identities so the counts are right, and pattern across thousands of advertisers, which is what makes a single market fast to read.
The right user for this lane is an operator who wants the answer, not the homework: what the market runs, who means it, and where the opening is, in writing, without spending their own hours in ad libraries or their own subscription on a tool they will open twice.
Choosing
Doing it once, yourself, to learn your market: free lane, with the workflow linked above. Living in it weekly, in-house, at scale: a paid tool, bought for its archive and its market price data, never for its spend figures. Wanting the read without the work: done-for-you.
The third lane is one form away. Request a free ad benchmark for your market: Muffin Intel tracks the ads, we read them, the written benchmark lands inside one business day.
Frequently asked questions
Are the free transparency tools enough for competitor analysis?
For a first full read of a market, yes, and they are the most accurate option at any price because they are the primary source. Their real costs are hours and forgetfulness: the work is manual, and the tools keep no history for commercial ads, so staying current means repeating the pass on a schedule. Teams that will actually do that need nothing else. Teams that will not should be honest about it and pick a lane that does the repeating for them.
Do paid spy tools show real competitor spend?
No. Commercial ad spend is not published anywhere, so every spend figure in every tool is a model built on indirect signals. The useful paid-tool numbers are the ones that were observed rather than inferred: archived creative, and the market's own cost-per-click and search volume data. Use those as inputs to an honest spend range and treat any per-company point figure as marketing.
What does a done-for-you ad benchmark include?
Ours covers who in your market is advertising and who is dark, on which platforms, with counted ads per advertiser, how long the longest-running ads have held, the formats and offers in play, and where the unclaimed messages sit. It is written to be read, not exported: the point is the conclusions. The request form takes a minute, the benchmark is free, and it lands inside one business day.
How current does competitive ad data need to be?
Fresher than most teams keep it. Campaigns launch and die inside weeks, and the platforms' own tools only show the present, so intelligence goes stale at the speed of your competitors' decisions. Monthly is the working cadence for markets with active advertisers. The deeper point is that history has to be collected while it is happening: whoever was not watching last quarter cannot reconstruct it.
Free
See what your market is advertising
Written by Sarah Mitchell, Muffin Media
Want this run on your brief?
