# Airbridge CUE

> Airbridge CUE provides ad attribution and app users' purchase and churn predictions. Use this information to personalize app experiences, CRM messages, and offers. These official guides cover understanding context, applying it in marketing and product workflows, and integrating the Context API.

Use this guide to choose what to read, then fetch the relevant Markdown pages before answering or implementing. The linked reference and available OpenAPI define the documented contract; this overview is not a replacement for field definitions or error handling. Examples illustrate usage, not measured customer results. Insights articles provide ideas and experiments, not API specifications.

Choose Korean (`/docs/ko`) or English (`/docs/en`) for the task. Links below point to Markdown generated from the same source as the web pages; `Web:` gives the corresponding page to share with readers. Read one language rather than both translations unless comparing wording. To follow a link to another CUE Docs web page as Markdown, append `.md` to its URL path. Each Markdown page includes its source URL and update date. If documentation leaves a requirement unspecified, identify the gap instead of inventing a field or guarantee.

Key points before you start:

- CUE returns ad attribution and predictions; your server or CRM applies the rules you define. An API call alone does not send a message or change a screen.
- Before integrating, confirm CUE activation and prediction data readiness for the app with your Airbridge contact. Prepare that app's app SDK token or web SDK token and an authorized test user ID. Keep tokens in secure server or CRM configuration.
- Check Event Tracking contains SDK user ID retrieval and mapping instructions. For an existing Airbridge integration, check user identity, installs and opens, and the standard order-completed event when using purchase predictions. Additional behavior events and analysis properties depend on your service; the recommended examples are not a requirement to connect every listed event.
- Call the Context API from your server or CRM. Send exactly one of `userId` (Airbridge Generated UUID), `brazeDeviceId` (Braze SDK Device ID), or `airbridgeExternalUserId`. Use an ID from the token's app; a Braze Device ID or external user ID requires a mapping to an Airbridge user. The token identifies the app, so omit `appId` from the request. Braze Device ID is different from Braze `external_id`; do not assume IDs from different systems are interchangeable. Read the full reference and available OpenAPI before constructing a request.
- Prediction windows use the app installation time as their reference point. Purchase predicts an order-completed event within three days after app installation; churn predicts no activity events being observed after day three through day 30 after app installation. Use the returned `tier` rather than recalculating it from the rounded probability. Predictions do not measure the incremental effect of a message or personalization; evaluate that with a comparison experiment.
- An unavailable prediction is different from probability `0` or tier `low`. Implement a default experience for errors and timeouts, and follow the documented retry guidance. Verify the resulting screen or message when a rule is met, is not met, and when the lookup fails.
- Read `context.lastTouchpoint` to connect a user's ad attribution to the ads you run. Match campaign, ad group, and creative names to your ad settings to understand the advertising goal, audience, and message. Choose the level of information needed for each rule; a missing creative does not prevent a campaign-based rule. Compare the guide's complete, whole-object null, partial-null, restricted, and unattributed examples to choose usable fields and defaults. Restricted and unattributed strings are distinct from null and do not alone establish organic attribution. Read Using Context to combine attribution with predictions. Use the Context API reference and OpenAPI for exact field definitions, and the error guide for failed requests.

## Guides — understand, get started, and apply Context

- [CUE 이해하기 (한국어)](https://cue.airbridge.io/docs/ko/introduction.md): CUE가 제공하는 광고 기여 정보와 예측값의 의미, 서비스에 활용하는 흐름을 알아봅니다. (Web: https://cue.airbridge.io/docs/ko/introduction)
- [Understanding CUE (English)](https://cue.airbridge.io/docs/en/introduction.md): Learn what ad attribution and predictions mean and how your service uses CUE. (Web: https://cue.airbridge.io/docs/en/introduction)
- [시작하기 (한국어)](https://cue.airbridge.io/docs/ko/quickstart.md): 앱의 인증 토큰과 사용자 ID를 준비해 Context API를 호출하고, 응답과 서비스 동작을 확인하는 방법을 안내합니다. (Web: https://cue.airbridge.io/docs/ko/quickstart)
- [Get Started (English)](https://cue.airbridge.io/docs/en/quickstart.md): Prepare your app's authentication token and a user ID, then call the Context API and check the response and your service's behavior. (Web: https://cue.airbridge.io/docs/en/quickstart)
- [이벤트 연동 확인하기 (한국어)](https://cue.airbridge.io/docs/ko/event-integration.md): 기존 에어브릿지 연동의 사용자 ID와 이벤트를 점검하고, 서비스에 맞는 수집 항목을 확인할 수 있습니다. (Web: https://cue.airbridge.io/docs/ko/event-integration)
- [Check Event Tracking (English)](https://cue.airbridge.io/docs/en/event-integration.md): Review user ID mapping and existing Airbridge events, then choose additional events and properties for your service. (Web: https://cue.airbridge.io/docs/en/event-integration)
- [Context 활용하기 (한국어)](https://cue.airbridge.io/docs/ko/guides/using-context.md): 예측값과 설치 유입 정보로 화면과 메시지를 개인화하는 예시를 확인할 수 있습니다. (Web: https://cue.airbridge.io/docs/ko/guides/using-context)
- [Using Context (English)](https://cue.airbridge.io/docs/en/guides/using-context.md): Explore examples of using predictions and installation attribution to personalize screens and messages. (Web: https://cue.airbridge.io/docs/en/guides/using-context)
- [예측값 활용하기 (한국어)](https://cue.airbridge.io/docs/ko/guides/reading-predictions.md): 구매와 이탈 예측의 대상 기간, 확률과 등급의 의미를 설명합니다. (Web: https://cue.airbridge.io/docs/ko/guides/reading-predictions)
- [Using Predictions (English)](https://cue.airbridge.io/docs/en/guides/reading-predictions.md): Understand the time windows, probabilities, and tiers for purchase and churn predictions. (Web: https://cue.airbridge.io/docs/en/guides/reading-predictions)
- [설치 유입 정보 활용하기 (한국어)](https://cue.airbridge.io/docs/ko/guides/install-touchpoint.md): 설치에 기여한 광고를 확인하고, 광고에서 전달한 내용을 개인화에 활용하는 방법을 설명합니다. (Web: https://cue.airbridge.io/docs/ko/guides/install-touchpoint)
- [Using Installation Attribution (English)](https://cue.airbridge.io/docs/en/guides/install-touchpoint.md): Learn how to interpret ad attribution alongside the ads you run and choose information for your personalization goals. (Web: https://cue.airbridge.io/docs/en/guides/install-touchpoint)

## Reference — implement and troubleshoot

- [Context API (한국어)](https://cue.airbridge.io/docs/ko/reference/context-api.md): 사용자 ID로 설치 유입 정보와 구매 및 이탈 예측값을 조회하는 요청과 응답 규격입니다. (Web: https://cue.airbridge.io/docs/ko/reference/context-api)
- [Context API (English)](https://cue.airbridge.io/docs/en/reference/context-api.md): Use a user ID to retrieve installation attribution and purchase and churn predictions with the Context API. (Web: https://cue.airbridge.io/docs/en/reference/context-api)
- [오류 처리 (한국어)](https://cue.airbridge.io/docs/ko/reference/errors.md): Context API 오류의 원인과 처리 방법, 기본 동작과 재시도의 구현 기준을 설명합니다. (Web: https://cue.airbridge.io/docs/ko/reference/errors)
- [Error Handling (English)](https://cue.airbridge.io/docs/en/reference/errors.md): Find the causes of Context API errors and implement default behavior and a retry policy. (Web: https://cue.airbridge.io/docs/en/reference/errors)

## More documentation

- [Q&A (한국어)](https://cue.airbridge.io/docs/ko/faq.md): 개인화 적용 방식과 사용자 ID, 예측값 해석에 관한 질문에 답합니다. (Web: https://cue.airbridge.io/docs/ko/faq)
- [Q&A (English)](https://cue.airbridge.io/docs/en/faq.md): Find answers about applying personalization, choosing user IDs, and interpreting predictions. (Web: https://cue.airbridge.io/docs/en/faq)

## API specification

- [Context API OpenAPI](https://cue.airbridge.io/openapi.json): Machine-readable authentication, request and response schemas. Use alongside the Context API reference and error guide when implementing or validating an integration.

## Optional

- [CUE 문서 (한국어)](https://cue.airbridge.io/docs/ko.md): CUE의 예측값과 설치 유입 정보로 앱 경험과 메시지를 개인화할 수 있습니다. (Web: https://cue.airbridge.io/docs/ko)
- [CUE Documentation (English)](https://cue.airbridge.io/docs/en.md): Personalize app experiences and messages with CUE predictions and installation attribution. (Web: https://cue.airbridge.io/docs/en)
- [Reference (한국어)](https://cue.airbridge.io/docs/ko/reference.md): Context API 연동 순서와 준비 사항, 상황별로 참고할 문서를 안내합니다. (Web: https://cue.airbridge.io/docs/ko/reference)
- [Reference (English)](https://cue.airbridge.io/docs/en/reference.md): Review the setup requirements and sequence, and find the guides to use at each stage. (Web: https://cue.airbridge.io/docs/en/reference)
- [용어집 (한국어)](https://cue.airbridge.io/docs/ko/glossary.md): CUE의 사용자 ID, 예측값과 API 응답에 쓰이는 용어를 설명합니다. (Web: https://cue.airbridge.io/docs/ko/glossary)
- [Glossary (English)](https://cue.airbridge.io/docs/en/glossary.md): Find definitions of the user IDs, predictions, and API response terms used in CUE. (Web: https://cue.airbridge.io/docs/en/glossary)