Einstein Lead Scoring Data Refresh and Training
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?
Quick Answer
The two-week-old 'Pending' data refresh status is the clue that points straight at data freshness as the root cause, not the age of the model itself. Einstein Lead Scoring only reflects reality when the records it trains on are current and complete, so when the refresh job stalls and incoming leads arrive with gaps like missing company size, the model keeps scoring against a picture of the business that no longer matches how leads are actually behaving, which is exactly why active leads can score low while stale ones score high. Retraining on the same schedule without first fixing the underlying pipeline would only bake the same stale, incomplete data into a new model, so the transform to cleanse and refresh the data has to happen before training resumes; skipping straight to retraining treats the symptom rather than the cause. This is also why the slow list view performance matters as a supporting detail rather than a separate problem, since a backlog of unrefreshed data and pending jobs can degrade both scoring accuracy and page performance at once. Whenever a scenario shows a scoring or prediction feature behaving erratically alongside a stalled or pending data status, treat the data pipeline as the first thing to fix, since machine learning outputs are only ever as trustworthy as the inputs behind them.
⚠ Common exam trap
The trap here is that candidates may focus on model retraining (Option C or D) without realizing that the core issue is a stalled data refresh, not the model configuration or frequency.
Answer choices
Why each option matters
Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.
Correct answer & explanation
✓
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 analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✓
Schedule a data transform to cleanse and refresh lead data before model training
Why this is correct
Ensures the model uses accurate, up-to-date data.
- ✗
Increase data retention period for lead records
Why it's wrong here
Does not address data incompleteness or staleness.
- ✗
Rebuild the Einstein Lead Scoring model using different fields
Why it's wrong here
May help but root cause is data quality, not field selection.
- ✗
Increase the model training frequency to weekly
Why it's wrong here
More frequent training on dirty data worsens issues.
Go deeper
Related to this question
About these practice questions
One of 753 original AI Associate practice questions on Courseiva, each with a full explanation and wrong-answer analysis — not exam dumps or protected exam content. Learn why practice questions differ from exam dumps →
Same concept, more angles
2 more ways this is tested on AI Associate
These questions test the same concept from different angles. Work through them to make sure you can recognise it however the exam phrases it.
Variation 1. 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?
medium- A.Map the email and event fields to the lead object so that the model can use them
- B.Add the Activity Count field to the scoring fields list in the model configuration
- C.Re-enable Einstein Activity Capture to resync all historical emails and events
- ✓ D.Retrain the Einstein Lead Scoring model using all historical lead data, including pre-activation leads
Why D: Retraining the model with all historical lead data (including pre-activation leads) will include conversion patterns from a larger dataset, improving accuracy and enabling scores for older leads. Option A is wrong because field mapping alone does not cause scoring to update. Option B is wrong because Einstein Activity Capture is already syncing; the issue is with the model. Option C is wrong because the scoring fields are separate from activity tracking.
Variation 2. A company uses Einstein Lead Scoring and finds that leads with high scores are not converting. What should the admin do to improve prediction accuracy?
medium- A.Increase the scoring model's maximum score
- ✓ B.Retrain the model with more recent conversion data
- C.Disable field-level security for scoring fields
- D.Lower the lead conversion threshold
Why B: Retraining the model with more recent conversion data allows the Einstein Lead Scoring model to adapt to changing patterns in lead behavior and conversion criteria. When high-scoring leads fail to convert, it indicates that the historical data used to train the model no longer reflects current conversion dynamics, so refreshing the training dataset improves prediction accuracy by aligning the model with recent outcomes.
JA
Written by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
This AI Associate practice question is part of Courseiva's free Salesforce certification practice question bank. Courseiva provides original exam-style practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics to help learners prepare for the AI Associate exam.