Configure
Fund the wallet, choose CPI, platform, GEO, bid, budget and cap.
Cost per install
CPI campaigns combine attribution, MMP validation, deterministic fraud rules and idempotent billing.

Fund the wallet, choose CPI, platform, GEO, bid, budget and cap.
Verified publishers use unique links carrying source and sub IDs.
Installs pass attribution and fraud review before financial processing.
Cost-per-install planning begins with an agreed definition of the event. Identify the application, platform and eligible audience, then state how an install will be reported and reviewed. A visit to a store listing is not the same record as the event supplied by the configured measurement process. Keep that distinction in the campaign brief so advertisers and publishers understand what the campaign is designed to measure.
Check the app destination and the relevant public information before requesting traffic. If the business needs a later action instead, consider whether an install-only objective answers that need. The purpose of this page is to help organize a CPI campaign around a declared outcome, rather than treating every interaction with an app promotion as an interchangeable success.
A publisher's unique link provides the starting reference for the campaign journey. The configured attribution path needs to preserve the identifiers required to connect a later install event to that start. Test the link and event setup together, including any measurement partner callback and the expected campaign mapping. Use a controlled example rather than assuming that a provider template is correct because the address looks familiar.
Stable event identity matters when callbacks are retried, because one reported install should not become several financial records simply through repeated delivery. Keep test references available to the team handling the integration. A successful destination redirect, an accepted callback and a reviewed conversion each establish a different part of the process, and they should not be collapsed into one unexplained pass label.
Set the bid, total budget and supported caps using the actual campaign requirements. A minimum bid shown in the product is a configuration floor, not a promise about the number or quality of installs a campaign will receive. Explain the eligible countries, traffic types and application platform to publishers so they can assess fit before sending visitors. Keep creative claims consistent with the store listing and avoid suggesting an unavailable reward or feature.
An initial controlled period can help establish whether the measurement process and source expectations agree. Record the allocation and distribution settings at the start so the report can be interpreted in context. Changing several settings at once can make it difficult to understand which decision contributed to a different result in the next period.
Read the reviewed conversion counts together with their attribution and source information. A high raw event total can still include rejected or duplicate deliveries that should not be treated as additional valid outcomes. If a source produces a discrepancy, compare its links, event identifiers and review records before drawing a conclusion about the whole campaign. Keep later engagement or business value as a separate analysis unless that action is the declared campaign outcome.
ADJODO's workflow provides controls and records for a CPI setup, but no page can guarantee that an acquired user will remain active or become a customer. Use traceable evidence to decide whether to pause, refine or expand distribution. The install campaign becomes more useful when its event definition, financial processing and source review remain consistent throughout the test.