How I: Use BigQuery to find my most valuable customers Video Lecture | "How I" by Google Developers - IT & Software

Hello.
My name's Marcel and I work
as a data analytics enthusiast
at Futurelytics startup.
Today, I will show
you how I use BigQuery
to find who are my most
valuable customers.
So the basic insight
of the Pareto rule
is it takes 20% of your most
valuable customers, which
brings you 80% of your revenue.
In this example, I
will show you how
to use BigQuery to show you
the most valuable contributors
on Wikipedia.
Data is taken from other
public data sets, which
is available on a
BigQuery console,
so you can play
with the basic SQL.
For the basic calculation of the
Pareto rule on a Wikipedia data
set, we should start with
a basic calculation, which
is the number of edited
pages for each contributor,
for content creator.
So we will run this
simple query and we
will be able to see the
Contributor ID and the number
of pages he edited.
This basic calculation is
used for advanced calculations
of the Pareto rule.
This one, this one,
this one, this one.
There is the basic
calculation we used first,
and we calculate the running
totals of each contributor
so we will be able to divide
each contributor in two groups,
if the contributor is
important or if it's
a small contributor with a
small amount of edited pages.
The results show us some
running totals numbers
and the total contribution
per contributor.
From these calculations,
we are able to divide
those contributors
in a specific group.
In this example, we have the
basic query, and a second query
which helps us to calculate
the running total metrics.
I chose the wrong query.
Should use this one.
Again, in this query,
we have the basic query,
which we use to
start, the second,
which helps us to calculate
the running totals.
And we are really interested in
how many important contributors
Wikipedia has, so we
will put this query
into another one,
which allows us
to count the number of
contributors in each group.
So the result is
that Wikipedia has
26,000 important contributors,
and 4,400,000 small ones.
From the calculation of Pareto,
we found that there is no 20%
of the most valuable
contributors.
There is only 8% of the most
valuable which brings you
80% of valuable
content on Wikipedia.
We used this basic
calculation when
we started building
Futurelytics.
For us, it was the
simplest way how
to find the most valuable
and less valuable customers.