From as-run logs to answers: reporting that actually works

Every playout server writes an as-run log: a line-by-line record of what actually went to air, when it started and how long it ran. It is the most honest document in a television operation — the schedule describes intent; the as-run describes reality. Overruns happen, items get dropped, live segments run long, an operator makes a manual change at two in the morning. The as-run captures all of it, without opinion or excuse.

Yet most channels treat that record as exhaust. It lands as a text or XML file in a folder on the playout server, named by date, written in codes, read by nobody — until the day somebody urgently needs to prove something. Then it becomes the most requested file in the building.

The first person who needs it is your advertiser. Agencies do not pay for spots that were booked; they pay for spots that aired. Proof-of-play — evidence that a specific commercial ran in a specific break at a specific time — is a standard condition of media invoicing. If the log takes a week to produce, the invoice takes a week longer to send, and a dispute over one missing spot can hold up payment for an entire campaign.

The second is your regulator. Broadcasting authorities across Europe can demand records of transmitted content: what aired, when, under what age rating, and how much advertising ran per clock hour. When a complaint arrives about something transmitted three months ago, 'we'll check the logs' needs to be a five-minute answer, not a forensic project.

The third is the rights holder. Distribution and licensing deals are frequently priced per airing, and collecting societies want accurate counts of usage. Undercount and you breach the contract; overcount and you overpay. Either way, the as-run is the settlement document — the number of times a title actually played is the number on which real money moves.

Here is how that reconciliation usually goes. Sales promises an agency a proof-of-play report. Operations opens a folder of daily log files — one per channel, per day, thousands of lines each — and starts searching for a code. The code is cryptic, the spot ran under three different versions, and the month spans thirty-one files. An hour later there is a half-built spreadsheet; a day later, a report nobody fully trusts.

Multiply that by every advertiser, every regulator query and every quarterly royalty statement, and as-run reporting quietly becomes one of the largest hidden labour costs in a small broadcast operation. Worse, it fails precisely where failure is expensive: one missed line in a manual search is one missing airing on an invoice.

FeedVision treats the as-run as first-class data. Because the platform delivers playlists directly to professional playout servers — VSN and SL NEO class systems among them — it closes the loop on the return side too: every airing is logged with its date, time, channel and content code, automatically and continuously. There is no folder of daily files to babysit. The record simply exists, complete, for every channel you run.

Context is what makes it useful. In FeedVision, the code on a log line is not an opaque string; it resolves against the same media library that drives scheduling — more than 20,000 clips with titles, episodes, ratings and categories. An airing record is never just 'this code ran at 21:42'. It is this episode of this series, this version of this commercial, on this channel, at that moment.

The same discipline applies upstream. FeedVision's nightly automation keeps an audit trail of every playlist send, so the transmission story is traceable end to end: what was scheduled, what was delivered to the server, and what the server played. When an airing is questioned, you are not reconstructing events from memory — the evidence chain already exists.

Reporting works at two altitudes. At the top, aggregate views answer the broad questions instantly: airings per title, per campaign, per channel, per period — the shape of a month at a glance. From any total, you drill straight down to the individual events behind it, each with its exact transmission time. The chain from 'this spot aired 94 times' to 'here are all 94 timestamps' is two clicks, not two days.

And because the people who consume these reports live in spreadsheets and signed documents, FeedVision exports both ways: Excel for the agency analyst who wants to pivot the data, PDF for the regulator, rights holder or client file that needs a clean, fixed record. Every export is generated from the same data the dashboard shows — one source of truth, several formats.

Then there is the part that changes who can do this work at all. FeedVision's reporting includes an AI assistant that answers plain-language questions about airing history. 'How many times did this spot air last month?' 'Which days did this series skip in March?' 'Compare this campaign's airings across our two channels.' The assistant queries the as-run record and returns the answer — the numbers, not a paragraph of hedging.

That matters because the person who needs the answer is rarely the person who knows the database. A sales manager on a call with an agency should not have to file a ticket with operations and wait two days. They ask, they get the count, they close the point while the client is still on the line.

The as-run is where broadcasting keeps its promises — to advertisers, to regulators, to rights holders. FeedVision's position is simple: if the platform builds the day, sends it to air and watches it play out, it should also be the place that proves what happened. Every airing logged, every question answerable, every report a few clicks away. That is reporting that actually works.