|
|
Competitive Ad Research: What Media Analysts Track and Why
Advertising is one of the few business activities that competitors publish in public. Every campaign leaves a trail of creative, copy, and landing pages, and analysts who read that trail systematically can describe a rival's strategy long before any earnings report does.
What analysts actually track
Most competitive ad research comes down to a short list of signals. Volume shows how many ads a brand has live and how many it launches each week. Format mix shows whether a team leans on video, static images, or carousels. Messaging shows which problems, offers and proof points appear again and again, and run length shows which ads survived long enough to suggest they were worth paying for.
Analysts add context around those signals. They note which countries and placements a brand targets, how the ad leads into a landing page, whether pricing or discounts feature in the creative and how often the brand refreshes its look. A pattern across thirty ads says more than any single standout.
Platforms built for this work collect public Meta ads into a searchable library. GetHookd describes more than 22 million live ads, refreshed daily, with a Brand Spy feature that follows a chosen competitor and reports each launch with its date and the number of ads live.
Explore the platform at https://www.gethookd.ai/.
Why so much ad data is public
The openness is partly regulatory. Under Article 39 of the EU Digital Services Act, providers of very large online platforms that present advertisements must keep a searchable, public repository of them, available through a search tool and through programming interfaces. The repository has to hold the content of each ad, the person on whose behalf it was shown, the person who paid for it if that is someone else, and the period during which it ran.
It must also state whether the ad targeted particular groups of recipients and the main parameters used, and the total number of recipients reached, with breakdowns by Member State for any targeted groups. The repository must be kept for the whole time the ad is presented and until one year after it was last shown.
Voluntary tools such as Meta's Ad Library and Google's Ads Transparency Center predate or sit beside those rules. Together they make advertising one of the most observable parts of a competitor's business, and the observable part is what analysts work with. In practice that means a European brand's campaigns on a very large platform can be inspected in more detail than almost any other part of its marketing, including who paid for them and roughly how many people they reached, though the detail varies by platform and by the category of ad.
Reading signals without overreading them
The main danger in ad research is mistaking presence for performance. A long-running ad may be a winner, but it may also be an evergreen piece on a modest budget, a forgotten campaign or a deliberate brand ad that nobody expects to convert directly. Run length is a reason to look more closely, not a verdict.
Volume has similar ambiguity. A brand that launches fifty variations a week may be testing aggressively, or it may be producing cheap variants because its system makes them easy. Look at whether the variations differ in hook and offer or merely in color and crop.
Good analysts write their conclusions as hypotheses. A statement such as the brand appears to be testing discount framing against social proof is honest about the evidence, and it points directly at something the analyst's own team can test. It also protects the analyst's credibility when a later look at the data shows a different picture.
Building a repeatable process
Consistency beats intensity. Choose a fixed set of competitors, usually five to ten, and review them on a schedule, weekly for fast categories and monthly for slower ones. Record the same fields each time, including the number of ads live, new launches, formats, main angles and any shifts in landing pages.
Use a shared board or spreadsheet so the work accumulates. Saved ads should carry notes about why they were saved, because a screenshot without context loses meaning within a month. GetHookd keeps saved ads available after the original ads go offline, which preserves a competitor's history even when the public library no longer shows it.
End each review with a short brief for the creative and media teams. One page covering what changed, what it might mean and what to test next turns the research from a report that is read once into an input that changes the work.
Limits to keep in view
Public ad data does not show results. Spend, conversion rates and return on ad spend stay private, and tools that claim to reveal them are estimating. GetHookd's own FAQ says it shows what is observable in the public library, such as which ads are live, how long they have run and the copy, and does not estimate revenue or invent spend figures.
Coverage has edges too. The platform focuses on Meta, meaning Facebook and Instagram, and its FAQ says coverage is thinnest in smaller countries and in service or information niches. Analysts who need Google or TikTok must use other sources.
Ethics matter as well. Studying public ads is routine, but copying a rival's copy, claims or visual identity is not, so research should feed original work. A short written policy on what the team will and will not reuse keeps that line clear when deadlines press. Review the policy at least once a year.