ProductUse cases
Learn
BlogAnalyses, trends & playbooks GuidesIn-depth market guides
PricingFAQ
DE · EN · FR · IT · ES · ROStart now

The DACH supplement ads playbook · Chapter 1 of 12 · 4 min read

What you can and cannot deduce from a public ad

Any ad library shows you thousands of creatives. None shows you which ones make money. The difference between a good analyst and one who fools himself lies in knowing exactly what can be deduced from public data and what cannot — and never crossing that line without declaring it.

This first chapter is the foundation of the whole playbook: the signal doctrine. Everything that follows rests on it.

What you'll be able to do after this chapter

  • read an ad's four public signals and know how much each one weighs;
  • explicitly name what you do NOT see — spend, ROAS, conversions — and refuse conclusions that assume them;
  • use longevity as an honest proxy, with its limits declared.

The principle

Four things can be known about any public ad, and RavenBI puts each on the card:

  1. The age — how many days it's been running. The most honest public signal: brands tend to switch off quickly the ads that don't justify their investment, so what lives 60-180 days has passed a real operational filter.
  2. The status — active or stopped. A stopped ad is a finished lesson; a live one is a decision still being paid for.
  3. The estimated reach — how much audience it has touched. Not budget, but its visible trace.
  4. The variants — how many versions run in parallel. A message in seven variants is a message someone invests in methodically, not a throwaway test.

What can NOT be known, however good the tool: real spend, cost per acquisition, conversions, profit. Any analysis that sounds like it knows them — "this ad makes money" — has exceeded the data. The honest phrasing is: "the ad its owner has reasons to keep alive". It sounds like a nuance; it's the difference between reading the market and telling yourself what you want to hear.

The workflow in RavenBI

  1. Open the Ad Library on your market. Every ad card carries the four signals: age in days, status, estimated reach, number of variants.
  2. Switch to the Exposure sort — it combines longevity with reach, lifting exactly the ads with the strongest public signals to the top. (On EU markets, the default sort is 🔥 Scaling — recent acceleration; on UK and US, where Meta doesn't publish reach, the library opens straight on Exposure. Chapter 10 puts both to work.)
  3. Switch to the Recent sort and observe the contrast: the wave of fresh tests, most destined to disappear. The difference between the two lists is, visually, this chapter's entire doctrine.
  4. Pick three old ads and three new ones and compare their signals one by one — the calibration exercise that starts any market analysis.

The Ad Library in RavenBI

The four signals, on every card: age, status, estimated reach, variants — the raw material of the whole playbook.

How to read the signals

  • Old + high reach + many variants = the triple signal — the ad with the strongest possible public signals; the first object of study.
  • Old + low reach = alive, but on a low flame — perhaps a narrow profitable segment, perhaps inertia; interesting only if its theme repeats at others.
  • New + exploding reach = visibility ramped up fast, not necessarily validation — check in two weeks whether it still exists.
  • Many variants of the same message = the brand believes in the message and is hunting the execution — a signal about the message, not about any particular variant.

The typical mistakes

  • reading reach as profit — it's the trace of the budget, not of the result;
  • analysing a three-day ad like a ninety-day one — they haven't passed through the same filter;
  • forgetting that longevity is a proxy, not proof — brand campaigns also exist, kept alive for other reasons;
  • formulating conclusions the data doesn't support — any sentence with "makes money" requires data nobody on the outside has.

Exercise

Take an ad you were tempted to call a "winner" and write its four signals, one sentence each. Then rewrite your initial conclusion using only statements the signals support. If the final sentence sounds more modest than the first — the chapter has done its job.

Signals only make sense in context: the same age means something different in a crowded category than in an empty one. Before you analyse creatives, choose the terrain — the market, the category, the problem. That's where the next chapter begins.

Guide overview