AI Associate AI Capabilities in CRM Practice Test 1 — 15 Questions

Question 1

An admin notices that Einstein Opportunity Scoring is not generating scores for new opportunities created in the past week. Which troubleshooting step should the admin take first?

Accepted Answer

Check that there are at least 50 won and 50 lost opportunities with populated fields. Einstein Opportunity Scoring requires a minimum of 50 won and 50 lost opportunities with populated fields to generate scores. Without this historical data, the model cannot learn patterns to score new opportunities. The admin should first verify this prerequisite before considering other steps.

Answer

Retrain the Opportunity Scoring model

Answer

Verify that users have the 'View Einstein Scores' permission

Answer

Wait 48 hours for the model to update

Question 2

Which THREE factors influence the prediction accuracy of Einstein Lead Scoring?

Accepted Answer

Historical conversion data of leads. Einstein Lead Scoring relies on the historical conversion data of leads (C) as its core training dataset — the model learns which lead attributes and behaviors preceded conversion, so the quantity and quality of past converted/unconverted leads directly drive prediction accuracy. It also analyzes conversion patterns across different lead sources (D), since the channel a lead originates from (web, referral, campaign, etc.) is a strong predictive signal the model weighs when scoring. Values in standard lead fields such as industry and company size (E) are among the key input features Einstein uses, so complete and consistent data in those fields improves the model's predictive power. Options A and B are not correct: custom formula fields are not used as scoring inputs (formula fields are calculated, not stored data Einstein trains on), and the number of times a lead is viewed by sales reps is not a tracked scoring factor in Einstein Lead Scoring.

Answer

Custom formula fields on the lead object

Answer

Number of times a lead is viewed by sales reps

Question 3

An admin reviews the Einstein service configuration JSON. Based on the exhibit, which statement is true?

Accepted Answer

Opportunities will not have Einstein scores. The JSON exhibit shows that the Einstein Opportunity Scoring feature is disabled (the 'status' field is set to 'Disabled'), meaning opportunities will not have Einstein scores. This is determined by the configuration flag that controls whether Einstein scoring is active for opportunities in the org.

Answer

Lead scoring is disabled for the org

Answer

The lead scoring model retrains daily

Answer

The admin has not configured any scoring fields for leads

Question 4

A mid-size company uses Sales Cloud with Einstein Lead Scoring and Einstein Activity Capture. The sales team reports that lead scores are not updating for leads that have been engaged via email and calendar events over the past two weeks. The admin checks the Einstein Lead Scoring model and finds that the model status is 'Active' and was retrained last month. The admin also verifies that Einstein Activity Capture is enabled and syncing data correctly. However, the lead scores remain unchanged. Upon further investigation, the admin discovers that the leads were created before the Einstein Lead Scoring model was activated, and the model's training data includes only leads created after activation. The company has over 10,000 leads, but only 200 were created after activation. Historical conversion data for leads created before activation is not being used. What should the admin do to ensure lead scores reflect recent engagement?

Accepted Answer

Retrain the Einstein Lead Scoring model using all historical lead data, including pre-activation leads. Einstein Lead Scoring builds its model from historical lead data, and if the training set only includes the 200 leads created after activation, the model has almost no conversion signal to learn from. Retraining the model with the full historical lead dataset (including pre-activation leads) gives the model enough positive/negative conversion examples to score engaged leads accurately. Once retrained on the broader dataset, recent email and calendar engagement captured by Einstein Activity Capture will be reflected in updated scores.

Answer

Map the email and event fields to the lead object so that the model can use them

Answer

Add the Activity Count field to the scoring fields list in the model configuration

Answer

Re-enable Einstein Activity Capture to resync all historical emails and events

Question 5

A Salesforce admin notices that Einstein Account Scoring is not generating scores for all accounts. Some accounts have no score even though they meet the data requirements. What is the most likely cause?

Accepted Answer

The org has fewer than 100 account records. Einstein Account Scoring requires a minimum of 100 account records to generate meaningful scores. If the org has fewer than 100 accounts, the model cannot establish a baseline for scoring, resulting in no scores for any account. This is a hard threshold enforced by the Einstein AI engine, not a configurable setting.

Answer

Einstein features are not enabled in the org

Answer

Users do not have the 'View Scores' permission

Answer

The accounts lack custom fields for scoring

Question 6

A company wants to use Einstein to predict which customers are likely to churn in the next 30 days. Which type of prediction should be created in Einstein Prediction Builder?

Accepted Answer

Binary classification. Einstein Prediction Builder is designed for binary outcomes, such as whether a customer will churn (yes/no) within a specified time frame. A binary classification model predicts one of two possible labels, making it the correct choice for this churn prediction use case.

Answer

Numeric prediction

Answer

Multi-class classification

Answer

Regression

Question 7

A Salesforce admin has enabled Einstein Opportunity Scoring but notices that some opportunities are not being scored. The admin checks the data and finds that the opportunities have all required fields. What could be another reason for missing scores?

Accepted Answer

The org has fewer than 100 opportunities with activities. Einstein Opportunity Scoring requires a minimum of 100 opportunities with activities (e.g., emails, events, tasks) in the last 90 days to generate scores. Even if all required fields are present, the model needs sufficient historical data to calculate a predictive score. Option C correctly identifies this data volume threshold as the likely cause.

Answer

Opportunity team members have not been assigned

Answer

The org is on a Professional Edition

Answer

The opportunities are missing custom fields

Question 8

A company uses Einstein Prediction Builder to predict whether a lead will convert. The model's confidence score is low, and the admin wants to improve accuracy. What is the most effective action?

Accepted Answer

Increase the amount of historical data for training. The correct option is C: Increase the amount of historical data for training. Einstein Prediction Builder learns from historical records, so a low confidence score often reflects insufficient training examples; providing more historical data gives the model a larger, more representative sample to learn patterns from, improving accuracy. Options A, B, and D do not fit: averaging multiple models is not a supported native approach in Prediction Builder, adding more custom fields can introduce noise or overfitting rather than improve accuracy, and retraining more frequently does not help if the underlying training dataset remains small or unrepresentative.

Answer

Create multiple prediction models and average their scores

Answer

Add more custom fields as predictor fields

Answer

Increase the frequency of model retraining

Question 9

Refer to the exhibit. An admin created a prediction using Einstein Prediction Builder. The prediction is configured to calculate a score on the Lead object. What does the JSON indicate about the model?

Accepted Answer

It predicts a binary outcome (e.g., convert or not). Einstein Prediction Builder for the Lead object, when configured to predict a binary outcome like 'convert or not,' outputs a JSON payload containing a 'probability' field (e.g., 0.85) and a 'predictedValue' field (e.g., 'Converted' or 'Not Converted'). The JSON shown indicates a classification model that assigns a probability to one of two discrete classes, which is the hallmark of binary classification. The presence of a 'predictedValue' field with a categorical label confirms it is not a regression (numeric) or multi-class text prediction.

Answer

It predicts a numeric value

Answer

It predicts the value of a text field

Answer

It is currently retraining the model

Question 10

Refer to the exhibit. A developer wrote a trigger to call an Einstein prediction API on lead insert. When new leads are created, the trigger fails with a 'Too many SOQL queries' error. What is the most likely cause?

Accepted Answer

The trigger is making API calls in a loop, causing governor limit issues. The trigger is making synchronous API calls to the Einstein prediction service inside a loop that iterates over each new lead. Each API call counts against the 'Number of callouts' governor limit (default 100 per transaction), and if more leads are inserted than the limit allows, the transaction fails with a 'Too many SOQL queries' error—though the error message is misleading, the root cause is exceeding the callout limit, not SOQL queries. The trigger should batch the API calls or use asynchronous processing to stay within limits.

Answer

The trigger is executed before the lead is saved

Answer

The trigger is querying a related object without an index

Answer

The Score__c field is not writeable

Question 11

A marketing team wants to predict which segment of customers will likely purchase a new product. Which Einstein feature is most appropriate?

Accepted Answer

Einstein Segment Prediction. Einstein Segment Prediction uses predictive modeling and machine learning to analyze historical customer data and identify which segments are most likely to purchase a new product. It is specifically designed for predictive segmentation based on behavioral and demographic patterns, making it the most appropriate choice for this use case.

Answer

Custom Reports

Answer

List Views

Answer

Campaigns

Question 12

Refer to the exhibit. An admin runs two queries on the Lead object. Both include the custom field Score__c used by Einstein Lead Scoring. The second query is significantly slower. What is the most likely cause?

Accepted Answer

The query is not using a selective filter on Score__c. Einstein Lead Scoring uses the Score__c field to store lead scores, and the field is not indexed by default. When a query filters on Score__c without an index, it forces a full table scan on the Lead object, which becomes significantly slower as the number of leads grows. The second query likely includes a non-selective filter on Score__c (e.g., a range or inequality), which cannot leverage any existing index and thus degrades performance.

Answer

The query uses a date function that is not selective

Answer

The Score__c field is indexed

Answer

The Einstein Lead Scoring job is running simultaneously

Question 13

Universal Containers (UC) uses Einstein Lead Scoring to prioritize leads. They have 500,000 leads in the system. Recently, the model scores have been inconsistent: some leads with low activity receive high scores, while active leads score low. The model was trained 3 months ago. UC updates lead records daily via an external system, but the data is often incomplete (e.g., missing company size). Support has reported slow performance on lead views. The admin notices that the 'Data Refresh Status' for Einstein Lead Scoring shows 'Pending' for 2 weeks. UC wants to improve model accuracy and performance.

Which action should the admin take first?

Accepted Answer

Schedule a data transform to cleanse and refresh lead data before model training. The 'Data Refresh Status' has been 'Pending' for two weeks, indicating that the model is not receiving updated lead data. Scheduling a data transform to cleanse and refresh lead data before model training directly addresses the root cause: incomplete and stale data (e.g., missing company size) leads to inconsistent scores. This action ensures the model trains on clean, current data, improving both accuracy and performance.

Answer

Increase data retention period for lead records

Answer

Rebuild the Einstein Lead Scoring model using different fields

Answer

Increase the model training frequency to weekly

Question 14

A manufacturing company uses Salesforce Sales Cloud with Einstein Activity Capture to log emails and events automatically. The sales team has noticed that some emails from Microsoft Outlook are not appearing in the Einstein Activity Capture dashboard, even though the users have configured the connection correctly. The administrator checks the activity capture settings and sees that the data source is enabled and users have the correct permissions. The missing emails are from a specific domain, @supplier.com, which is a known vendor. What should the administrator do to ensure these emails are captured?

Accepted Answer

Add the @supplier.com domain to the allowlist in Einstein Activity Capture settings.. Einstein Activity Capture uses allowlists and blocklists to control which emails are captured. By default, emails from external domains may be excluded to reduce noise. Adding @supplier.com to the allowlist ensures that emails from that domain are explicitly included in the capture process, overriding any default filtering.

Answer

Increase the sync frequency for Einstein Activity Capture to every 5 minutes.

Answer

Instruct users to manually log the supplier emails as activities in Salesforce.

Answer

Assign the 'Einstein Activity Capture User' permission set to the supplier's email domain.

Question 15

A marketing team wants to use AI to personalize email content for each subscriber based on their past interactions. Which Salesforce feature should they use?

Accepted Answer

Einstein Recommendations. Einstein Recommendations is designed to deliver personalized product or content recommendations in emails based on each subscriber's past behavior. It dynamically inserts recommended items, achieving content personalization. The other features either analyze copy, optimize send time, or select content blocks, but none provide individualized content recommendations based on interactions.

Answer

Einstein Copy Insights

Answer

Einstein Send Time Optimization

Answer

Einstein Content Selection