Apple Maps campaign data is now available — five new report types covering campaigns that promote a business location rather than an app.

Apple Maps campaigns — the ones that promote a business location rather than an app — can now be reported on in Improvado. Until now they could not be extracted at all: the previous Apple API never exposed the Apple Maps ad account as a selectable account, so the data was out of reach. The new Apple Ads Platform API exposes it, and five report levels are available on it.

What the reports carry

Alongside the familiar impressions, taps and spend, these reports carry the Apple Maps engagement actions that have no equivalent in App Store reporting:

• Get Directions — taps that opened directions to the business location.
• Calls — taps that placed a call.
• Website taps — taps that opened the business website.
• Shares — taps that shared the place card.
• Get the App — taps that went to the app.
• Gallery engagements — interactions with the photo gallery.

Dimensions and granularity

• Dimensions: ad placement, device class and business location, each selectable per report. Ad placement is not offered on the keyword and search term levels — Apple does not report Apple Maps data broken down that way.
• Granularity: daily, weekly and monthly on all five reports, plus hourly on the campaign, ad group and keyword levels. Apple does not report Apple Maps ads or search terms by hour.

Two things worth knowing before you build on these

• Most search terms arrive without their text. Apple releases the literal query only above a privacy threshold, so on the Search Term report most rows carry the keyword that matched but an empty Search Term Text. In one measured two-week window, 4 distinct queries were disclosed across 81 term-and-dimension combinations. The column is there for the queries Apple does release; it is not a complete list of what people typed.
• A report that grows past 5000 rows stops with an error instead of loading. Apple caps a single Apple Maps report request at 5000 rows, and its paging over these reports returns some rows twice while omitting others. Rather than load numbers we know to be wrong, the extraction stops and names the dimension to drop. Removing Location ID first usually brings a report back under the limit — it multiplies the row count the most.