Please wait while the transcript is being prepared...
0:04
AfterWork is a training provider
for data courses and we
developed an eight-week
part-time Data Science Program
for working professionals.
When we started recruiting
the first three cohorts,
we didn't have any
guidelines to use
for optimizing the
recruitment pipeline.
Instead, we simply used data on
marketing channels to decide
the marketing strategies
to undertake.
While this approach was
somewhat satisfactory,
it didn't help us to
exhaustively solve the
business problem at hand.
As we'll get to see
when we conclude
this talk, upon the
adoption of Crisp DM,
the increase in revenue for
the successful cohorts increased
by at least 25 percent.
Our main goal was to
optimize the student
recruitment pipeline
and in the process,
maximize student engagement
and increase revenue through
a data analytics workflow
that comprises five stages:
that's business understanding,
data understanding,
data preparation, data analysis,
and data presentation.
1:13
Going through the first
stage of the data analytics
workflow derived from Crisp DM,
we were required to first
define the business problem,
which was to optimize the
student recruitment pipeline
and, in the process,
maximize student engagement
and increase revenue.
Secondly, we are required to
translate the business problem
into a data analytics problem,
allowing us to solve
the business problem
directly by solving the right
data analytics problem.
There are many data
analytics problems
that could have been
solved, however,
we opted to solve
a data analytics problem
that required us to
minimize operational costs
and increase student enrollment.
The other key consideration for
the business
understanding stage was
defining the metrics of success,
which we would use at the
end of the project to
evaluate whether we had
solved for our goal.
If we were able to determine
the optimal values
for the expense variables
using linear programming
and also determine their
relationships in students
demographics data
and curriculum data,
then it would mean
that we had solved for
the data analytics
problem and as a result,
so for the business problem.
The last consideration at this
stage was to come up with
a project plan in the form of
a work plan that
outlined the duration,
the resources and tools required
to complete the project.
The project plan
ensured that we did not
overlook or miss out on
key project requirements.
