How it works
Spotledger records radio, finds your adverts inside it and gives every booked spot a status. Here is each step, and what you do.
Step 1 of 3
Spotledger records each station's official online stream in one-minute pieces, day and night. Every minute is marked monitored, silent or down, so a gap is never hidden and never counted against a station.
Step 2 of 3
Your advert becomes two audio fingerprints, a classic one and a neural one that Spotledger trains itself. Both look for it inside the recorded radio, even when a presenter talks over it, and Spotledger works out which cut aired and for how long. The match must line up in time from start to finish before it counts.
Step 3 of 3
Each booked spot is matched to what aired and given a status. Spots not detected in a fully monitored window, cut short, in the wrong version or aired outside the booked band go on a make goods list. A spot is marked inconclusive, never missed, only when the longest stretch we did not hear in its booked band is at least as long as the advert. Reports come as PDF and CSV, with the coverage record and a way to dispute any line.
The product
Real screens, filled from our own recordings: a detection, your plan against what aired, a page of proof and a make good list.
Fanta
Wazobia FM Lagos · 28.8 s · Both fingerprints agree · confidence 0.99
Nescafé
Wazobia FM Abuja · 30.9 s · Both fingerprints agree · confidence 0.99
Fanta
Nigeria Info Lagos · 28.7 s · Both fingerprints agree · confidence 0.99
Detection log
Station, time, length played, which fingerprint found it and its confidence score, for every advert Spotledger hears.
Apple on radio
9 booked spots
87%
Compliance rate
Plan versus actual
A status for every booked spot, and a compliance rate counted over the minutes we were listening.
Proof of airing
Apple
Proof of airing
The time, the station, the length played and a fingerprint of the report itself.
Make good list
1 spot to claim
| Station | Booked | Not detected |
|---|---|---|
| Wazobia FM Abuja | 1 | 1 |
Make good list
Spots not detected in a fully monitored window, by station, ready for a make good request. Cut short, wrong version and outside band spots are on the CSV list too.
Brands shown here are not Spotledger customers. These adverts were recorded from public broadcasts.
What you do
One file per version of the advert, in English or Pidgin, and the plan spreadsheet your agency already has. Spotledger AI reads the columns and you confirm the mapping.
Every minute, on every station, whether or not anyone is watching the screen. Outages raise an alert.
A Morning Brief at 8am Lagos time says what aired, what was missed and what rivals launched. Every number in it comes from your report.
Download the make goods list: spots not detected in a fully monitored window, cut short, in the wrong version or aired outside the booked band. It comes with the coverage record, ready to send to the station or your agency.
The statuses
The same colours and words are used on this site, in the workspace and in every report. We never say a station did not air a spot: we say Not detected in a fully monitored window, and show the minutes we were listening.
Aired in the booked time band, the booked version, in full.
Aired that day, but not in the booked band.
Played for less than 90% of its length.
Another cut of the same campaign aired, for example the 30 second one instead of the 60.
Every minute that could hold the advert was monitored, and nothing matched. The report shows the minutes we were listening.
Aired, but not on the plan.
Part of the booked band was not monitored, for long enough that the advert could have aired there. We cannot say either way, so it is never called missed. Left out of the score.
We were listening for less than half of the booked band, because a stream was down or not recorded. Left out of the score.
The audio is recorded but not checked yet. Left out of the score until it is.
The compliance rate is spots aired as booked, divided by the spots we could check in full. Spots that are not monitored, inconclusive or still being checked are left out of the count, so a gap in our listening never counts against a station.
Honest about limits
We record each station's official online stream. What listeners heard on FM can differ, because a stream can carry different adverts. Every report prints the source it was recorded from. We also show a coverage record for every station and leave minutes we could not hear out of the score.
Confidence is a score from how closely the airing matches the advert in the fingerprints that found it. A person checks any score under 0.80, and any find by the neural fingerprint alone under 0.94. You can also dispute any line in a report, and a person reviews it.
Reports come daily or weekly. Each prints a SHA-256 fingerprint of its own contents in the footer, so any change to it can be detected. Accuracy results are on the home page and in the case study.
Lagos and Abuja, 12 stations. Live reads are checked in Nigerian English.
Free for 30 days. No card and no approval needed.