Scribd
Upload
3 views

Assignment Subjective Questions

The document outlines the top three variables influencing lead conversion, which are the total number of visits, total time spent on the website, and the current population as working professionals. It also identifies key categorical variables to focus on, including current occupation as working professionals, lead origin, and lead source. Additionally, it suggests strategies for aggressive lead conversion during intern hiring periods and for minimizing unnecessary calls when sales targets are met, emphasizing the importance of adjusting lead prediction cutoffs to optimize recall and precision.

Uploaded by

Dairymilk silk
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
3 views

Assignment Subjective Questions

The document outlines the top three variables influencing lead conversion, which are the total number of visits, total time spent on the website, and the current population as working professionals. It also identifies key categorical variables to focus on, including current occupation as working professionals, lead origin, and lead source. Additionally, it suggests strategies for aggressive lead conversion during intern hiring periods and for minimizing unnecessary calls when sales targets are met, emphasizing the importance of adjusting lead prediction cutoffs to optimize recall and precision.

Uploaded by

Dairymilk silk
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

1.

Which are the top three variables in your model which contribute most towards the
probability of a lead getting converted?

Answer:
The top three variables in the model which contribute most towards the lead getting converted
include:
• Total number of visits
Higher the time spent greater are the chances of conversion

• Total time spent on the website


Higher the time spent greater are the chances of conversion of lead
• Current population as working professional

2. What are the top 3 categorical/dummy variables in the model which should be focused
the most on in order to increase the probability of lead conversion?
Answer:
The top three categorical variables in the model which should be focused are:
• current occupation_working professional
• Lead Origin_lead add form
• Lead Source_welingak website

3. X Education has a period of 2 months every year during which they hire some interns.
The sales team, in particular, has around 10 interns allotted to them. So during this phase,
they wish to make the lead conversion more aggressive. So they want almost all of the
potential leads (i.e. the customers who have been predicted as 1 by the model) to be
converted and hence, want to make phone calls to as much of such people as possible.
Suggest a good strategy they should employ at this stage.
Answer:
During this period they can reduce the cut off for lead identification as the aim is to identify all
the potential leads and at the same time, resource is not a challenge. This shall help to increase
the Recall which measures the proportion of actual positives (converted leads) that the model
correctly identifies. Also, for aggressive campaign, it is important to identify as many potential
leads as possible, even if it means some false positives (non-converted leads are incorrectly
classified as potential).
Hence, focusing on recall ensures that you do not miss out on leads who could potentially
convert.

4. Similarly, at times, the company reaches its target for a quarter before the deadline.
During this time, the company wants the sales team to focus on some new work as well.
So during this time, the company’s aim is to not make phone calls unless it’s extremely
necessary, i.e. they want to minimize the rate of useless phone calls. Suggest a strategy
they should employ at this stage.
Answer:
In this case as the company wants to be selective in making the call, they can increase the
threshold for lead prediction cutoff. This shall increase the Precision of the model which is
correctly predicted positive leads (True Positives, TP) out of all predicted positive leads (TP +
FP).This shall help focusing on “hot” leads or leads with high probability of conversion. This
shall thus help in minimizing useless phone calls.

You might also like