Your Apple Ads campaign says Running. Yesterday it spent money. Today the dashboard shows zero. Before raising a bid, pull the Apple Ads report for the exact app, campaign, storefront and time window. Find the first signal that changed: impressions, taps, spend or installs. If a campaign or ad group is on hold, resolve the displayed reason. If impressions continue, check taps before troubleshooting eligibility. If Apple Ads records spend and only your attribution dashboard goes quiet, investigate the measurement feed.
This guide covers post-launch Apple Ads Advanced campaigns in App Store search results. The same symptom can occur in Apple Ads Basic or another placement, where the controls and reporting differ. Use the Apple Ads placements comparison if the delivery loss is in a different placement.
Find the first signal that changed
The labels in this table describe observations within the same report slice. Compare the affected campaign and storefront with their own baseline. A total across the account can conceal a stopped market, and a blank keyword cell has a different meaning from a measured zero. Apple's campaign dashboard documentation describes the available dimensions and separates impressions, taps, spend and several install measures.
What you see First evidence to check Decision it supports --- --- --- The report is blank or recent rows look incomplete Compare the selected dates, account, filters and export timestamp with a known complete export Repair the view or allow the data to populate Impressions are zero and a status shows a reason Read the reason on the affected campaign, ad group or ad Correct the documented hold before testing auctions Impressions are zero in an apparently eligible slice Compare storefront availability, targeting, matching and relevant search demand with a healthy slice Test one coverage or bid constraint, or escalate with evidence Impressions are present but taps are absent or very low Segment by storefront, query family and ad variation where the data permits Examine whether the people seeing the ad have a reason to tap Apple Ads has spend and installs, but another report is flat Reconcile the event definition, date basis, filters and feed Route the incident to the measurement owner Spend is below budget, with qualified acquisition still occurring Check where additional relevant coverage exists and what it is worth Keep the current limits or fund a bounded test
Several of these observations can coexist. If the Canadian ad group shows a hold while the UK campaign has spend and no events in the mobile measurement partner, open two incident records. A bid edit in the UK cannot repair the Canadian hold. A measurement fix in Canada cannot establish that the ads were eligible.
Save a baseline before editing
Capture a dated export from the affected account and a comparable healthy period. For a campaign that has never served, use a healthy sibling only where the app, storefront, placement and strategy make the comparison meaningful. If none exists, record that limitation. An established brand campaign in the US is a weak reference for a first generic campaign in a new language.
Record the app and campaign IDs, placement, storefront, ad group, strategy, status, budget and applicable bid control. Add the date range and time basis, impressions, taps, spend, tap-through new downloads and redownloads. Record the currency when comparing accounts or reports. For search terms, note whether a row is absent, grouped into a low-volume bucket or present with a measured value. Apple's reporting definitions and Apple Ads glossary explain the reported measures and the low-volume grouping.
For an established campaign, look for the last interval with normal delivery and the first with missing delivery. Compare that boundary with Change History, App Store Connect changes and billing notifications. Include edits made outside the ad dashboard, such as a storefront availability change or a product page update. A timestamp match gives you a candidate cause. The affected scope and the system status still have to fit.
A new campaign needs a different baseline. Apple says it can take 24–48 hours for initial performance data to gather and recommends allowing several days before major changes. Treat that as general reporting guidance. An explicit hold, a future start date or an unavailable app can be investigated immediately. A mature campaign with an abrupt break and a healthy sibling also gives you evidence that a new campaign lacks.
Read the status from the account down to the ad
Start with the largest scope that fits. If several unrelated campaigns stopped together, inspect account-level billing and messages first. If one storefront stopped while another still serves, compare the affected storefront, schedule and ad groups. A running campaign can contain an ad group whose own status needs attention. Open the affected row and save the exact status and reason before handing it to the owner.
Apple's billing help explains that a declined charge can pause ads. The billing owner should resolve the issue through the existing account, then verify the new status. Apple's campaign API reference lists possible system reasons including missing payment, app eligibility, future start and exhausted budget. Those API names describe possible states. Use the reason your own account provides.
Check whether the app and the intended ad are eligible in the selected storefront. If the incident followed a change to a custom product page or ad variation, compare the variation with the campaign's default ad and check their actual statuses. Apple explains how the default ad and search results ad variations are sourced from App Store product pages. A variation problem only explains the observed loss if the affected configuration depends on it.
Then inspect audience settings. Apple documents that audience refinements can narrow reach and that some ad groups can be held below an audience minimum. Review the audience status and settings against your approved targeting. If a restriction is required for compliance, market fit or an existing test, keep it. The commercial conclusion may be that this particular segment is too narrow to use as planned.
Diagnose the strategy you are actually running
Manage Bids
Open the relevant keywords, negative keywords, Search Terms and the ad group's Search Match setting. Verify which keyword or discovery route was intended to enter each query family. One exact-match keyword provides little evidence about all relevant demand. A negative keyword that routes traffic away from one group can be intentional. Confirm that the destination group is active and eligible before removing the exclusion.
If there is relevant demand, check max cost per tap for the specific keyword and any CPA cap the account still exposes for the ad group. Apple warns that a low cap can reduce impressions and conversions. Review Apple's performance guidance and the keyword controls before changing either. Check the campaign budget and observed spend too. A campaign with zero spend throughout a complete period has a different symptom from one that spends early and then goes quiet.
Maximize Conversions
The first question is whether you are looking at the automated ad group or any separately configured manual keyword activity. Apple documents that the automatic ad group uses Search Match and the campaign uses a target CPA with automated bids. Check the campaign setup and Maximize Conversions guidance. The target is an install-cost objective over time. A daily result can vary from that objective, and a new target does not establish that the app is eligible or that suitable searches exist.
If the campaign can serve, compare the target and daily budget with the economics of the product and the demand you can observe. Check whether a recent target edit lines up with the affected slice. Avoid judging an established campaign's economics from a single partial day. If your business would reject the implied cost of a paying customer even if volume returned, preserve the limit and examine a different acquisition opportunity.
Use impression share to narrow the possibilities
In Apple Ads Insights, compare the same app, country or region, search term and day. Apple describes Impression Share reports as aggregate reports that may extend across campaigns. Check the report's grain before using it to explain a particular campaign. Search popularity and relative rank are contextual signals. They do not tell you how many incremental customers a higher bid would buy.
A relevant term with recurring demand, measured impressions and low share can justify a small competitiveness test. A term with little observable demand and no stable history gives you a weaker case for bidding up. An unavailable report row should stay labeled unavailable. Grouped low-volume terms and campaign-level totals can both exist when individual search terms are hard to inspect.
Review your App Store metadata if the app's fit for the query is in doubt. Apple says relevance matters for entering the search-results auction and identifies app metadata as one input in its troubleshooting guidance. That supports checking whether the listing accurately describes the product. It gives you no basis to infer the auction formula or to add terms unrelated to what the app does. If query coverage is the actual bottleneck, use the Apple Ads keyword strategy to plan a relevant discovery test.
Separate paid delivery from attribution and quality
If Apple Ads shows impressions but very few taps, look at the query mix and the proposition users see. Compare a brand segment with its own history before comparing it with broader discovery. If taps and spend are healthy while installs soften, check the product-page experience and report maturity. Changing the bid can alter who arrives, which makes it harder to tell whether a product-page change helped.
If Apple Ads reports installs and an MMP or in-app dashboard reports none, compare App Store downloads with what the downstream system counts, such as first open or activation. Align attribution windows, new versus returning users, campaign IDs and dates. Apple explicitly documents reporting differences with mobile measurement providers. If discrepancies appeared after an app release or reporting change, trace a known event through the affected feed and verify that it arrives under the expected campaign and definition. The Firebase versus MMP buying framework covers overlapping feeds in more detail.
Design a test with a ceiling and a decision
Write one hypothesis you can disprove. For example, “Our max CPT is limiting access to this storefront's relevant exact-match queries.” Record what would support it: those queries gain impressions and taps after a bounded bid change, with stable matching and targeting. Record what would weaken it: the same slice stays quiet despite the approved bid change and the comparison data remains inconclusive. Leave the rest of the account stable where practical.
Separate the spending ceiling from the unit-economic ceiling. The first is the maximum test spend the team can afford. The second is what you can pay for a tap or install given observed downstream behavior. For a hypothetical example, if you can pay $40 per activated user, 20% of tap-through new downloads activate, and 25% of taps yield a new download, an initial planning ceiling is $2 per tap ($40 × 0.20 × 0.25). Use recent, comparable cohorts and record uncertainty. That figure describes modeled economics. It does not estimate what you need to bid to enter or win an auction.
Set an owner, observation window, budget, decision date and rollback value. Decide in advance whether a lift in impressions alone buys further time. If the test produces enough taps to evaluate quality, look at new downloads and mature activation alongside cost. If the only extra volume comes from a weaker query family, isolate it before declaring that the constraint was solved. Apple describes the daily budget as an average pacing control. Apple also explains in its campaign setup guide that spend can exceed the daily budget on individual days. Monitor actual spend and intervene early if your internal ceiling is tighter.
A fictional incident with two different next moves
An iOS invoicing app sees Canadian and UK spend fall in the Monday report. Canada had steady delivery before the weekend. The UK campaign is new and was added alongside a tracking release. The shared dashboard shows the same orange warning for both. The growth lead exports each storefront from Apple Ads before making changes.
In Canada, the last complete interval shows no impressions and the active ad group has an audience hold after a Friday edit. The team asks whether the refinement was deliberate. It was required for a small pilot. They keep the pilot's restriction and restore the separately approved, previously working ad group after confirming there is no overlap they intended to exclude. They check the status and first full delivery interval before deciding whether the pilot merits another market. The incident owner records the hold and the approved configuration.
In the UK, Apple Ads shows impressions, taps, spend and new downloads. Only the downstream activation dashboard is empty. The team leaves bidding alone and investigates the release, event definitions, ingestion and campaign-ID joins. They also keep new and returning users separate during reconciliation. The observable delivery makes an auction test a poor first use of budget. The two storefronts share a dashboard symptom and require different owners.
Know when to wait, hold or escalate
Wait for comparable reporting if the only missing evidence is a recent row and the campaign has no explicit blocker. Hold the current configuration if no test fits the acquisition economics or if it would require removing a legitimate targeting constraint. Escalate to the account owner or Apple Ads support when a system reason persists after the documented correction or when a mature, apparently eligible campaign has a repeatable abrupt loss that you cannot isolate.
A useful support packet contains IDs, storefront, placement and strategy, the last healthy and first affected windows, status text, redacted exports, relevant change history, and what you have already ruled out. Ask a narrow question such as whether the specific ad group is eligible in that storefront. Keep payment details, credentials and user-level data out of the packet. This process helps make the uncertainty explicit. It cannot establish hidden auction behavior or guarantee recovery.
For a broader account redesign after delivery is understood, read the Apple Ads campaign structure playbook.
Scope an Apple Ads delivery review
If you lead acquisition for a live iOS app and can share a redacted Apple Ads export, change history and the exact storefront or campaign affected, talk to Sharply Labs. We can examine the delivery scope, the bid strategy and where reporting stops, then scope an evidence checklist and one bounded test or support handoff. The review carries no guarantee of restored delivery, lower CPA or performance. See our mobile app growth services and paid acquisition services for the work this diagnostic can lead into.