Free CIM Level 6 Specialist Award in AI Marketing Practice Questions

Sample Practice Questions

Try these 10 sample questions from the CIM Level 6 Specialist Award in AI Marketing practice bank. Each includes the answer and explanation. Continue with the interactive practice above to review the full question set.

1. 1A churn model learns from customer records paired with known renewal outcomes. Which learning approach is being used?

   1. A.Supervised machine learningCorrect answer: A
   2. B.Unsupervised machine learning
   3. C.Reinforcement learning
   4. D.Unsupervised clustering

   Explanation: Google states: "Training a model from features and their corresponding labels." Known outcomes supply the labels. Unsupervised methods find patterns without those outcome labels; reinforcement learning uses rewards from actions.

   Reference: Google Machine Learning Glossary — supervised machine learning

2. 2Customers misunderstand an AI assistant's technical explanations. Which change most directly addresses this barrier to informed use?

   1. A.Match explanations to users' knowledge and rolesCorrect answer: A
   2. B.Use one unchanged explanation for all audiences
   3. C.Replace explanations with an aggregate accuracy score
   4. D.Provide customer explanations only to developers

   Explanation: The source states: "with descriptions tailored to individual differences such as the user’s role, knowledge, and skill level." Tailoring helps customers understand the explanation. A uniform explanation may retain the barrier; an aggregate score or developer-only explanation does not explain the system to these customers.

   Reference: NIST Artificial Intelligence Risk Management Framework — §3.5, p.16

3. 3A marketer needs to divide customers into audiences using their interests and behaviours. Which AI application fits?

   1. A.Audience segmentationCorrect answer: A
   2. B.Sentiment analysis
   3. C.Content translation
   4. D.Programmatic bidding

   Explanation: IBM states: "divide up their customers by various traits, interests and behaviors" Segmentation groups customers by attributes and behaviour; sentiment assesses attitudes, translation changes language, and programmatic tools purchase ad placements.

   Reference: IBM: A guide to AI in marketing — Audience segmentation

4. 4A marketer groups unlabelled customer profiles by similarity without specifying target categories. What approach fits this task?

   1. A.Supervised classification
   2. B.Unsupervised machine learningCorrect answer: B
   3. C.Reward-based reinforcement learning
   4. D.Supervised regression

   Explanation: Google states: "Training a model to find patterns in a dataset, typically an unlabeled dataset." The task discovers groups without labels. Classification and regression require target outcomes; reward-based learning requires actions and rewards.

   Reference: Google Machine Learning Glossary — unsupervised machine learning

5. 5A brand is testing an AI service with customers. Which evidence best checks assumptions about how customers will actually use it?

   1. A.The engineering team's preferred features
   2. B.Feedback from prospective users in the intended contextCorrect answer: B
   3. C.The number of hidden layers in the model
   4. D.The supplier's total number of employees

   Explanation: NIST states: "checking their assumptions about context of use;" User feedback checks real usage assumptions. Internal preferences, architecture size and supplier headcount do not directly establish the customer context.

   Reference: NIST Artificial Intelligence Risk Management Framework — §5.2, p.25

6. 6A brand wants new draft subject lines and website copy from a brief. Which AI capability directly produces those materials?

   1. A.Predictive lead scoring
   2. B.Content generationCorrect answer: B
   3. C.Audience clustering
   4. D.Sentiment classification

   Explanation: IBM states: "creating output such as blogs, marketing messages, copywriting materials, emails, subject lines" Generation produces new draft text. Scoring, clustering and sentiment classification analyse or categorise rather than draft copy.

   Reference: IBM: A guide to AI in marketing — Content generation

7. 7What makes a neural network a deep model?

   1. A.It contains many input records.
   2. B.It uses one output category.
   3. C.It contains multiple hidden layers.Correct answer: C
   4. D.It predicts several campaign metrics.

   Explanation: Google states: "A neural network containing more than one hidden layer." Depth refers to hidden layers, not record count, number of classes, or number of business metrics.

   Reference: Google Machine Learning Glossary — deep model

8. 8Customers need to understand where an assistant's answers may be unreliable. What information most directly supports appropriate expectations?

   1. A.The product's advertising budget
   2. B.The supplier's historical sales volume
   3. C.The assistant's knowledge limits and oversight arrangementsCorrect answer: C
   4. D.The team members' preferred campaign channels

   Explanation: NIST states: "Information about the AI system’s knowledge limits" Knowledge limits help users calibrate expectations. Budgets, sales volume and preferred channels do not describe where answers are reliable.

   Reference: NIST Artificial Intelligence Risk Management Framework — MAP 2.2, p.26

9. 9Which AI application most directly supports natural-language responses to routine customer service questions?

   1. A.Recommendation engine
   2. B.Advertising bid optimiser
   3. C.Customer service assistantCorrect answer: C
   4. D.Audience segmentation tool

   Explanation: IBM states: "generative AI-powered assistants can interact with customers in natural language" Assistants support conversational service. Recommendations select products, bid tools purchase placements, and segmentation groups audiences.

   Reference: IBM: A guide to AI in marketing — Customer service assistants

10. 10Which field enables computers to process customers' written messages as language?

    1. A.Computer vision
    2. B.Regression analysis
    3. C.Reinforcement learning
    4. D.Natural language processingCorrect answer: D

    Explanation: Google states: "The field of teaching computers to process what a user said or typed using linguistic rules." NLP concerns spoken or typed language. Vision handles images; regression predicts numeric outcomes; reinforcement learning learns actions through rewards.

    Reference: Google Machine Learning Glossary — natural language processing