Викторина, связанная с руководством · Средний Уровень
Платформы данных клиентов и викторина по искусственному интеллекту
Тестирование концепций CDP и искусственного интеллекта: унифицированные профили, разрешение идентификационных данных, качество данных, жизненный цикл конфиденциальности, доступ к модели, активация, хранение и исправление.
Вопрос 1 из 8
How does the CDP Institute define a customer data platform?
Предварительный просмотр каждого вопроса8 вопросы
Ответы и пояснения появляются по мере прохождения теста.
- How does the CDP Institute define a customer data platform?
- A model that guarantees every customer identity is correct
- A database used only for email campaigns
- Software that creates and maintains a persistent, unified customer record accessible to other systems
- A public directory of customer information
- Why should probabilistic identity matches remain distinguishable from verified matches?
- They carry different confidence and can incorrectly merge people or records
- Probabilistic matches are always exact
- Verified IDs cannot be corrected
- Identity confidence affects only storage cost
- A churn model needs only recent purchase frequency. What data practice is appropriate?
- Give it the full profile and all customer-service notes
- Retain every event permanently
- Share the profile with every activation partner
- Provide only the attributes needed for that use case
- What can go wrong if a customer’s corrected preference does not propagate?
- The model becomes more explainable
- Downstream segments or recommendations may continue using stale data
- The customer record is deleted automatically
- Identity confidence improves
- What kind of privacy risk does NIST’s Privacy Framework help organizations manage?
- Only risks from email delivery
- Only model accuracy errors
- Risks arising through data collection, use, sharing, and disposal
- Whether a campaign has a high click rate
- A CDP profile includes a model score. What should a reviewer know before activation?
- Only the campaign color scheme
- The number of rows in the database
- Whether the model generated text quickly
- What fields and identity links produced the score and where the segment will be used
- Why is profile unification not proof of data accuracy or consent?
- Combining records does not validate the source facts or permitted purpose
- Unified records always contain verified data
- Consent is inherited across all systems
- AI corrects every stale field
- Which control helps reconstruct an AI-driven campaign later?
- Save only the final click rate
- Keep the profile inputs, model and segment versions, recipients, and activation logs
- Delete the model configuration after launch
- Store the ad creative without audience rules