What to measure when you run ads for a release

Every ad platform gives you a dashboard that says the campaign went well. It is measuring reach, clicks and cost per click, all of which describe the advertising rather than the project.

The question you care about is different: did this money leave the project better off. Answering it means reading the spend against things the ad platform never sees, which is the job Backline does by holding advertising, landing page clicks, streaming, audience and ticket sales for one project on a single timeline.

The three numbers the platform leads with

Impressions. Buy more and it goes up. It is a budget restatement.

Click-through rate. A creative quality signal, useful for comparing two adverts against each other and almost useless as a campaign outcome.

Cost per click. Genuinely useful for buying efficiency, and it tells you nothing about what happened after the click.

None of these is dishonest. They are simply about the advert, and there is nobody else to ask about the advert.

The five to measure instead

1. Clicks on your own landing page, by platform and country

The first outcome you own end to end. If your adverts point at your own link page rather than straight at a streaming service, you can see how many people arrived, where from, and which platform they chose.

That last part is worth the setup on its own. Platform choice at the link page is the closest thing to a preference signal you get, and it is frequently not what the team assumed. Backline reports it by source, platform, country and device on the same page as the spend that produced it.

2. Click-through from the landing page

Of the people who arrived, how many went on to a service? A low rate points at the page rather than the advert: wrong artwork, slow load, missing platform, a broken destination on mobile.

This is the cheapest fix in music advertising and the least examined.

3. Movement in streams during and after the window

Not attributed. Movement. Take the streaming series, mark the campaign start and end, and look at whether the level changed and whether it held after spend stopped.

The bit after spend stops is the interesting part. A campaign whose effect ends the day the money does bought attention. A campaign with a higher baseline three weeks later bought an audience. Because the spend and the streaming series share one date axis in Backline, that comparison is a glance rather than an export.

4. Follower and audience movement over the same window

Followers gained per hundred spent, roughly. It is a crude figure and it is comparable across campaigns, which is more than can be said for impressions.

If a campaign produced a lot of streams and almost no followers, you rented attention. That is sometimes the goal, for a chart week or a placement push, and it should be a decision rather than a surprise.

5. Ticket sales in the markets you targeted

For any project with live dates, this is the outcome that pays. Look at whether pacing changed in targeted cities against untargeted ones, which is a natural comparison the campaign itself created. Ticket pacing per show sits in the same project as the spend, so the targeted and untargeted cities are two rows of the same view.

Several accounts, several currencies

Two operational realities decide whether ad reporting is trustworthy, and both are worth insisting on from any tool you use.

Multiple ad accounts. Projects routinely have more than one: a legacy account, the label's account, a newer one set up for a campaign. Backline connects as many Meta and Google accounts to a project as it actually has, and reports them separately as well as together. Ask it a question that could mean any of them and it asks which accounts you want before giving you a number, rather than quietly summing a set nobody chose. Every figure states the accounts it covers.

Currencies. Ad accounts report in their own currency. Adding pounds to euros to dollars into one spend figure is wrong unless a rate and a date are attached to it. Backline reports each account in its own currency, so a total is never silently manufactured out of three of them.

Neither of these is a technicality. They are the difference between a spend figure you can put in front of a label and one that falls apart the first time somebody checks it.

Reporting it

Four lines and a chart:

1. Spend, by platform and account, in each account's own currency.
2. Landing page clicks and click-through, with the platform split.
3. Movement in streams and followers across the window, and the level three weeks after spend ended.
4. What you would do differently.

The chart is the campaign window marked on the streaming and follower series, so anybody reading can see the shape rather than take your word for it. Backline will produce that as a branded report you can download and send, built from the project's own connected data rather than from a spreadsheet somebody maintained by hand.

The comparison that matters

Not against a benchmark. Against your own last campaign, and against what the project was doing before the money started.

The second is the harder one and the more valuable. What was the baseline, and did the campaign window move the project beyond its own normal range? Sometimes the honest answer includes a placement that was running at the same time and cannot be separated out. Saying so is what makes the times you can say something definite worth listening to, and it is why every campaign you record properly makes the next one easier to judge.