Music data terms every manager should know

Half the arguments in a campaign review are not about the numbers. They are about the words. One person says growth and means monthly listeners, another means followers, a third means streams gained last week, and the room spends twenty minutes discovering it agreed all along.

The vocabulary is worth fixing once. Streaming is now 69.6 percent of global recorded music income, so most of what a manager is judged on arrives as a streaming figure with a name attached, and the name is doing a lot of work.

What follows is a working glossary, grouped the way a project view is grouped. Each entry says what the term measures and, where it earns it, what the term is commonly mistaken for. The entries are short on purpose. Precision is the point.

Audience terms

Monthly listeners. The number of distinct accounts that played the artist at least once in a rolling 28 day window, as reported by the streaming service. It is a level, not an accumulation, so it can fall without anything going wrong. It is not a fanbase, and two people who each played you once count the same as two people who played you daily.

Followers. Accounts that have explicitly asked to hear from you again, on a streaming service or a social platform. Slower moving than listeners, much harder to inflate, and the closest thing to a standing audience. A listener is a visit. A follower is a subscription.

Engagement. Likes, comments, shares and saves on a post or an account over a period. Views are not engagement, because a view is delivery and engagement is response. Backline reports engagement per channel across Facebook, Instagram and TikTok on one project view, so the comparison between channels is like for like.

Audience geography. Where listening or following is happening, reported as cities or as countries. City level tells you where to route a tour. Country level tells you where to buy media. They answer different questions and are not interchangeable.

Streaming terms

Streams. Plays that met the service's qualifying threshold. Every platform counts slightly differently, which is why a total from one source rarely matches a total from another to the digit, and why the trend matters more than the last three digits.

Saves, and save rate. A save is a listener adding the track to their own library or playlist. Save rate is saves relative to listeners over the same period. Streams tell you a track was delivered. Save rate tells you whether the people who heard it wanted it again, which is the single most useful quality signal on a new release.

Playlist reach. The combined follower count of the playlists a track currently sits on. Reach is capacity, not delivery: it describes how many people could plausibly encounter the track, not how many did.

Playlist adds. New placements gained in a period. Adds are an input to growth, not growth itself. A week with four adds and no movement in listeners is a week where four playlists did nothing.

Editorial, algorithmic and user playlists. Editorial playlists are programmed by a platform's own curators. Algorithmic playlists are generated per listener from behaviour, so they scale with performance rather than with relationships. User playlists are made by the public. An editorial add is a decision by a person, an algorithmic placement is a consequence of your own numbers, and treating the two as the same event will lead you to thank the wrong party.

Sound usage. The count of short form videos that use a recording, and the views those videos received. It is not streaming. A recording can be in half a million videos and convert modestly to plays, or the reverse. Backline keeps sound usage on its own footing rather than adding it to a stream count, because the two are different units and summing them produces a number that means nothing.

Catalogue. Everything released more than roughly eighteen months ago, depending on whose definition you are using. It is where most listening lives: 43 percent of United States on demand audio streams in 2025 came from tracks released between 2021 and 2025, which leaves the majority coming from older material.

Campaign terms

Impressions. The number of times an advert or a post was served. Not people, and not attention. Buy more media and impressions rise by definition, which is why they make a poor headline.

Clicks. Times someone acted on what was served. A click is the first genuine signal of interest in a paid campaign.

Click-through rate. Clicks divided by impressions. A rate, so it survives comparison between a week of large spend and a week of small spend, which raw click counts do not.

Attribution. The rule you use to decide which activity gets credit for an outcome. Attribution is a choice, not a measurement, and the honest version of it is written down before a campaign starts rather than argued after it ends. In Backline advertising sits on the same project timeline as streaming, links and ticket sales, so you can see what moved alongside spend and decide credit with the evidence in front of you.

Smart link click. A click on your own landing page through to a chosen service. This is the one campaign figure that belongs entirely to you, and it tells you where an audience goes when it is given a free choice, which is often not where the plan assumed.

Live terms

Capacity. The number of tickets a room can sell for this configuration. Capacity is the denominator for everything else about a show, and a show record without it can only report absolute sales.

Sell-through. Tickets sold as a share of capacity. A 400 capacity room at 92 percent is a stronger result than a 2,000 capacity room at 38 percent, and only sell-through shows that. Backline holds capacity, tickets sold and the pacing curve per show, and rolls dates up to a tour level figure.

Pacing. The shape of cumulative sales between the on sale date and the show. Two shows can arrive at the same sell-through by completely different routes, and the route is what tells you whether to add spend.

Gross. Face value ticket revenue before deductions. It is not net, it is not the guarantee, and it is not what the artist is paid. Quote it as gross or do not quote it.

Data terms

ISRC. The identifier for a recording. Two recordings of the same song, a single version and a live version, are two ISRCs.

UPC. The identifier for a release, meaning the album, EP or single package that the recordings sit inside. One UPC usually contains several ISRCs. Reporting keyed to the wrong one of these two is the most common cause of a catalogue that looks smaller than it is.

Cumulative total, and daily figure. A cumulative total is everything counted since the beginning, so it only ever rises and it reads identically under a 7 day filter and a 90 day filter. A daily or windowed figure is the gain inside a defined period. If a number does not change when you change the date range, it is cumulative, and it does not belong in a review.

Window, or date range. The period a figure covers. Nearly every disagreement about a streaming number is a window disagreement: a trailing 28 days against a calendar month, or a window that includes release day against one that does not.

Reporting lag. The gap between something happening and the data source publishing it. It varies by platform and it means the most recent days on any chart are provisional. Read the shape, not the final point.

Co-movement. Two series moving together in time. Useful, and not proof of cause. Press coverage and a streaming rise in the same week is a reason to look closer, not a finding.

One vocabulary per project

Definitions only settle arguments if everybody is reading them off the same surface. That is the quiet advantage of holding a project's sources together: streaming, audience geography, social, advertising, website and search, smart links, live shows and ticket pacing, press and newsletter all report under one date window in Backline, so a follower figure and a stream figure are describing the same fortnight without anyone having to check.

Ask Backline AI answers in these same definitions, from the project's own connected data, and names the window it used. When a manager and an assistant share a vocabulary, the review can spend its time on the decision instead of on the wording.

If you want the longer version of the first three entries, monthly listeners, followers and streams is worth a read, and playlist adds are not growth covers why the reach and adds distinction matters more than it sounds.