Five questions to ask before you switch music analytics tools

The pattern is consistent. Somebody has been paying for a tracker for two years, likes it well enough, and is looking at alternatives because a specific job keeps being manual. Usually that job is the monthly report, sometimes it is answering a label, occasionally it is a tour that nobody can explain.

That is worth naming before you compare anything, because it tells you whether a switch will help. If the frustration is a missing chart, another tracker might fix it. If the frustration is that nothing in the tool knows about your ad spend, your ticket sales or your mailing list, no tracker will, and you are looking at a different type of product. There is a longer version of that distinction in platforms compared by what they actually store.

Assuming you have got that far, here are the five questions that predict whether the switch sticks.

1. What history comes with you, and what does not?

This is the one people find out too late. Almost nothing in this category transfers. A new platform typically starts building its own time series the day you connect a source, which means your first month on it has no comparisons and your first quarter has thin ones.

Three practical mitigations:

• Export what you can from the old tool now, while you are still paying for it. Even a flat CSV of monthly totals gives you a reference point.
• Check whether the new platform accepts a history upload. Spotify for Artists exports a timeline of audience and per-song data, and a platform that can ingest that will have real history from day one rather than a flat line. Ours does, deliberately, for exactly this reason.
• Ask which sources rebuild history automatically. Ticketing, email and advertising APIs usually serve a trailing window on connection, so those backfill themselves. Public streaming counters generally do not.

2. How many people need to see it, and at what level?

Count the actual humans: you, a day-to-day manager, an assistant, the PR company for three weeks, an agent who wants tour numbers, a label contact who wants a monthly summary, maybe the artist.

Now ask two things of every candidate. Can each of those people be given access to one project rather than everything? And can you tell the difference between someone who can change settings and someone who can only read? If the answer to either is no, the tool will not be where the team works, because you will not put a freelancer next to your whole roster.

3. Which numbers have to reconcile with something else?

Every manager has one or two figures that get quoted back to them by someone with a different source. Spotify for Artists is the usual referee for streams. The box office is the referee for tickets. Meta Ads Manager is the referee for spend.

Pick those figures, and during the trial check them against the referee, on the same dates. Small differences are normal and explainable: reporting lag, timezone boundaries, whether a source counts a partial stream. A difference you cannot get an explanation for is the thing that will undermine the tool in a meeting six months from now, and it is far cheaper to find in week two.

4. What does the tool do without you?

The daily reality of running several projects is that you look at a dashboard when something prompts you. So ask what the tool sends, and how noisy it is.

The useful version is a small number of alerts with a real baseline behind them: a genuine surge against this project's own recent history, a milestone, a sell-through that has stopped moving, a source that has stopped syncing. The useless version fires on any movement, and gets muted in a fortnight. When you see the alerts settings, look for whether thresholds are relative to the project's own history rather than absolute numbers, because absolute thresholds are wrong for every project except the one they were tuned on.

5. What is the exit?

Ask it out loud during the sales conversation, and notice how the answer lands. Can you export your own data, in a format that is usable, without asking? Is there a documented way to delete a project and its stored data? If you disconnect a source, does syncing stop and does the stored data go?

A vendor who has thought about this answers immediately, because the same machinery serves data protection obligations they already have. A vendor who improvises is telling you something about how they think about your data.

Running the switch

One full reporting cycle, both tools live, and one honest comparison at the end: which one answered the question that made you look in the first place, and how long each took.

Then cancel one of them on purpose. Most teams I know are paying for a tool that nobody has opened since spring, because switching happened and finishing never did.