Mara, a data analyst, is trying to harness an AI tool within Salesforce to predict the next quarter’s product returns. However, after deploying the model, she finds the predicted return rates substantially off from the actual figures.
What could be a major data quality issue contributing to this discrepancy?