While configuring predictive lead scoring, a consultant discovers that leads generated through a partner referral program follow a noticeably different qualification pattern than direct inbound leads, and mixing them together is dragging down overall model accuracy. What should the consultant recommend?
Select an answer to reveal the explanation.
Short Explanation
When one type of lead behaves in a fundamentally different way than the rest, throwing them all into the same training pot just confuses the model, the same way mixing two different recipes' ingredients confuses a chef trying to learn one dish. The fix is not to shut the whole kitchen down, and it is not to just cross out the confusing dish's name on the menu and pretend it does not exist, and it is definitely not to pretend the two recipes use the same ingredients when they clearly do not. The sensible move is to keep the two populations separate in how the model is scoped and trained, so each gets judged against patterns that actually apply to it, and the main group's accuracy stops getting dragged down by a mismatched segment.
Full Explanation
The correct answer is A. Predictive lead scoring configuration supports filtering which records the model is trained on and scores, so when one segment behaves distinctly enough to distort accuracy for the rest, the right move is to scope the model to exclude or separately manage that segment rather than letting it dilute the patterns learned from the more consistent population. Option B is incorrect because disabling scoring organization-wide sacrifices a working, valuable signal for the majority of leads just to work around one segment's mismatch, when a targeted configuration change solves the problem without that cost. Option C is incorrect because manually zeroing out scores does not fix the underlying training data issue, it just hides the symptom for sellers while the model continues learning from an inconsistent mix, and it discards any real signal partner leads might still carry. Option D is incorrect because forcing dissimilar records to share identical attributes falsifies the data rather than resolving the pattern mismatch, and would further corrupt what the model learns instead of improving it.