Strip away the coverage and one fact decides everything: Maps ads are sold to businesses with physical locations in the US and Canada. If your product is an app, this is not a channel you can buy.
That should be the end of the article for most readers. It isn’t, for two reasons. First, a meaningful minority of app companies do have physical locations — restaurant chains, retail banks, gyms, grocers, carriers, healthcare networks — and for them the boundary between "app marketing" and "local marketing" is about to get expensive to get wrong. Second, and more importantly for everyone else: Maps is the clearest public statement Apple has made about how it intends to build advertising products. The targeting model, the ad density, and the category bans are all design decisions, and every one of them is a signal about the surfaces you do buy.
01What Apple has actually committed to
The announcement came on 24 March 2026, bundled with Apple Business — a consolidated business suite that TechCrunch reported would launch on 14 April 2026 across 200 countries, folding together employee email and calendar, directory services, free mobile device management for app distribution, "Blueprints" for setup, and AppleCare+ for Business from $6.99 per month. Maps advertising was the revenue-bearing component inside an otherwise free bundle, a framing The Register noted at the time.
The mechanics Apple described are conventional and worth stating precisely, because a lot of the commentary has drifted from them:
- Auction-based, pay-per-interaction. Advertisers pay only when a user interacts with the ad.
- Physical locations only. Eligibility runs through businesses that have a location and an existing Maps listing, claimed through Apple’s business tooling.
- US and Canada. Apple’s own Maps advertising page states that only advertisers with businesses located in the United States and Canada are eligible.
- Labelled, with a visual tell. Ads appear next to relevant search results and in Suggested Places, marked as ads and carrying a small blue halo around the map pin.
- Advanced controls for larger spenders. Scheduling and location targeting are described as available to larger advertisers.
Then, on 14 July 2026, the advertising policy took effect — the same fortnight Apple rewrote its Advertising Services terms to remove the word "Apple" from the description of where its ads may be delivered, a change we unpacked in the terms rewrite analysis. Those two events are close enough together to be worth holding in the same frame.
02Why it isn’t an install channel
Apple’s Maps advertising page describes the product in terms of local businesses reaching customers who are actively searching for businesses like theirs, at the moment they are deciding where to go. The worked example on the page is a restaurant. The onboarding instruction is to claim your business location and upload photos.
There is no app-install objective described anywhere in Apple’s public Maps advertising material. There is no keyword-level bidding, no product page destination, no conversion-value schema, no attribution surface. The unit being advertised is a place. The conversion being bought is a visit or a call, not a download.
This matters because "Apple Ads" is now an umbrella covering products with almost nothing in common operationally. Since the April 2025 rename from Apple Search Ads, the brand has stretched to cover App Store search and Today tab placements, Apple News formats, and now Maps. Treating them as one channel because they share a brand is the same error as treating Google Search and YouTube as one channel because both are Google. If you are budgeting, budget by surface and by objective, not by vendor name — a discipline we set out in the 2026 Apple Ads guide.
Maps sells proximity. The App Store sells intent. They are both Apple Ads, and they are not the same business.
03One ad per result: scarcity as design
The most instructive detail in the July policy reporting is the ad density. Per TechCrunch’s read of Apple’s documentation, Apple will show only a single ad per search result in Maps — an explicit divergence from Google’s local advertising model.
Set that beside what Apple did on the App Store in the same period. From 3 March 2026, App Store search results began carrying up to two ads per query, rolling out from the UK to Japan and then globally, with existing campaigns automatically eligible and placement decided by auction dynamics. One surface got denser; the other launched deliberately sparse.
The reasonable inference is that Apple is calibrating ad load per surface against how tolerant that surface’s users are, not applying a house standard. Maps is a utility people rely on under time pressure; the App Store is a storefront where commercial content is expected. If you are modelling future inventory on Apple properties, do not assume the App Store’s density curve generalises. And in the near term, the practical consequence of two slots rather than one is competitive: your ad now routinely appears beside a rival’s, which is a creative problem before it is a bidding problem. We worked through the structure for that in the multi-slot playbook.
04The targeting model, and what it signals
Apple’s Maps advertising page is unusually specific about what the system uses and what it refuses to use. Targeting relies on contextual signals only: the search term, approximate device location, and the area of the map being viewed. It explicitly does not use precise location history, prior user interactions, or demographic targeting by age or gender. Apple’s stated position, quoted at announcement, is that users’ personal data stays on the device, is not collected or stored by Apple, and is not shared with third parties.
Read that as a product philosophy rather than a privacy press release, because it has a direct commercial consequence: Apple is building ad products where the query carries the targeting. No audience graph, no retargeting pool, no lookalikes. Just the thing the user typed and where they are standing.
That is exactly the thesis behind buying App Store search in the first place. The reason Apple Ads reaches the highest-value iOS users is not clever audience modelling — it is that the App Store audience monetises far above the market, and search placement intercepts intent a user has already declared by typing it. We derived the underlying spend asymmetry in why Apple Ads buys the highest-ARPU iOS users, and the auction consequence in the expensive-user paradox. Maps is the same bet applied to physical destinations: no profile, just declared intent plus context.
→ WHAT THIS MEANS FOR YOUR ACCOUNT
Apple keeps rewarding the same thing
Across every surface Apple has shipped, the lever that pays is match quality between what the user asked for and what you offer — not audience construction. That is an argument for spending your effort on keyword-to-destination mapping and creative relevance, and very little of it on the audience-modelling instincts imported from social channels.
There is one measurement consequence worth stating. A contextual, on-device model with no cross-app identity is, by construction, hard to measure with the tools most UA teams own. Apple has published no attribution surface for Maps ads. If it ever becomes relevant to you, expect to evaluate it the way you would evaluate out-of-home or local radio: geo holdouts and store-level lift, not last-click.
05The category bans, and what they reveal
Apple’s Maps advertising policy, effective 14 July 2026, prohibits a specific and revealing set of categories. Home services are out wholesale — plumbing, electrical, locksmith, HVAC, pest control, roofing and general contracting. Cryptocurrency ATMs and bail bonds providers are also barred. Medical services are evaluated case by case. On top of that sit the standard prohibitions: deceptive, profane and political ads, and content featuring weapons, violence, controlled substances or defamatory material.
The home services exclusion is the striking one, because that category is among the most lucrative in local search — it is the core of Google’s Local Services Ads business. Apple walking away from it is a deliberate choice, and the most plausible reading is fraud and quality control: locksmith and emergency-services listings are historically the most abused surfaces in local search, and Apple is not staffing a verification apparatus to police them.
The pattern generalises to something app marketers should recognise. Apple consistently prefers a smaller, cleaner inventory over a larger, messier one — and it applies that preference to creative as well as categories. The clearest current example is the 4+ content bar attached to the new App Store creative assets format shipping this fall, which we broke down in the 4+ creative policy piece. Same instinct, different surface.
→ FREE 10-POINT AUDIT
Make sure the surface you can buy is working first
Before you spend a minute on a channel that isn’t open to you, we’ll open your Apple Ads account and show you where App Store search spend is leaking. One hour. Senior strategist. No pitch.
Book my free audit →
06Summer is nearly over
Apple said summer 2026, in the US and Canada. On the Q3 FY26 earnings call, with Services at roughly $30.7 billion and up about 12% year over year, CFO Kevan Parekh cited June-quarter records in advertising among other Services lines and repeated that Maps ads would run in the US and Canada this summer.
As of mid-August 2026, Apple’s own Maps advertising page still reads "coming soon" and offers a sign-up form for notification when advertising becomes available. That is not a scandal — a summer window has weeks left in it, and Apple rarely pre-announces a date. But it does mean something concrete for planning: there is no launch date, no rate card, no published auction mechanics, and no benchmark data of any kind.
Treat any article you encounter offering Apple Maps ads CPCs, bid guidance or performance benchmarks as fabricated. There is no dataset to draw them from. This is the same discipline we apply to Apple Ads cost benchmarks generally, where even for the mature App Store surface the published vendor figures disagree materially — Adapty has reported average US cost per tap around $1.58, AppTweak around $1.91, and SplitMetrics a range of roughly $1.79 to $2.25 — because each vendor is describing a different panel of clients, categories and match types.
07Who this actually touches
Three groups, in descending order of urgency.
Group 1 · DirectApp companies with physical footprints
Restaurant groups, retail banks, gyms and studios, grocery and convenience chains, carriers, healthcare networks, hotel and travel brands with properties. For you, Maps ads will be buyable, and the awkward part is organisational rather than tactical: the budget will sit with a local or field marketing team while the app install goal sits with UA, and the same user will be counted differently by each. Decide now which team owns the Maps line and what it is accountable for, before an auction decides it for you.
Group 2 · PreparatoryAnyone whose competitors have footprints
If you are a delivery, booking or fintech app competing against businesses that have physical locations, those competitors are about to get a labelled, halo-marked placement at the moment of decision on a surface you cannot bid into. The counter is not to chase Maps — it is to be unmistakable on the surface where declared intent is yours to intercept. That means the App Store search result, and it means creative and destination discipline on it.
Group 3 · ObservationalEveryone else
For the majority of app advertisers, the correct action on Maps ads this quarter is to read the policy once and do nothing. The opportunity cost of a distracted growth team is real, and the surfaces that already pay you deserve the attention — starting with the weekly discipline in the weekly ROAS review and the signal quality covered in high-ARPU signals.
08What nobody knows yet
Four gaps, named rather than smoothed over.
No launch date and no pricing. Apple’s page says businesses can advertise "at a cost you control, with the flexibility to start and stop campaigns anytime," and discloses nothing about rates or auction mechanics. Everything beyond "auction-based, pay per interaction" is inference.
Market share estimates are shaky and dated. The most-cited figure in Maps ads coverage — RioSEO data via eMarketer, putting frequent use of Google Maps for local business information at 75% of US consumers against 31% for Apple Maps — has no clear collection date attached in the coverage we can verify. Multi-select survey questions of this kind also let a single respondent count in both numbers. Treat it as a rough directional read on a duopoly with an unequal split, not as a share model. Apple does not disclose Maps usage figures, and it discloses no advertising revenue line item at all.
The relationship to the July terms rewrite is unresolved. Apple removed "Apple" from where ads may be delivered and brought web pages and links into the definition of ad content in the same July window that the Maps ad policy took effect. Whether those changes are related, and whether Maps inventory or Maps advertiser demand is implicated in any off-platform expansion, is not addressed in anything Apple has published.
Nobody knows how users will react. Commentary has run from "Apple’s riskiest bet yet" to routine monetisation of an obvious surface, and the honest answer is that the tolerance of a navigation app’s users for commercial content is untested at scale. Apple’s one-ad-per-result design suggests Apple shares that uncertainty.
The through-line for an app marketer is simple enough. Apple is expanding the number of places it sells advertising while narrowing what any single placement is allowed to do — one ad in Maps, contextual signals only, a 4+ ceiling on creative, categories refused outright. That is a company optimising for a clean surface rather than a large one. Which is precisely why the audience on the surface you can buy stays worth what it is worth, and why the discipline that wins there is still match quality rather than volume — the argument we made against buying on cost in the case against CPI.
Watch what Apple refuses to sell. It tells you more about the value of what it does sell than any benchmark will.