SAS Visual Analytics Demo for Retail

http://www.sas.com/visualanalytics
Learn how to use SAS Visual Analytics to identify customer segments and do Market Basket Analysis Reporting.

SAS VISUAL ANALYTICS
Get fast answers to even the most complex questions using data of any size – including big data in Hadoop. Guided exploration makes it easy. In-memory processing makes it fast. Advanced data visualization tools make it clear. Scalability makes it the perfect fit. And the price makes it within your reach.

LEARN MORE ABOUT SAS VISUAL ANALYTICS
http://www.sas.com/software/visual-analytics/overview.html

TRY VISUAL ANALYTICS YOURSELF
Browse sample reports or explore on your own
with this cloud-based demo.

http://www.sas.com/software/visual-analytics/demos/all-demos.html


SUBSCRIBE TO THE SAS SOFTWARE YOUTUBE CHANNEL
http://www.youtube.com/subscription_center?add_user=sassoftware

ABOUT SAS
SAS is the leader in business analytics software and services, and the largest independent vendor in the business intelligence market. Through innovative solutions, SAS helps customers at more than 75,000 sites improve performance and deliver value by making better decisions faster. Since 1976 SAS has been giving customers around the world The Power to Know.®

VISIT SAS
http://www.sas.com

CONNECT WITH SAS
SAS ► http://www.sas.com
SAS Customer Support ► http://support.sas.com
SAS Communities ► http://communities.sas.com
Facebook ► https://www.facebook.com/SASsoftware
Twitter ► https://www.twitter.com/SASsoftware
LinkedIn ► http://www.linkedin.com/company/sas
Google+ ► https://plus.google.com/+sassoftware
Blogs ► http://blogs.sas.com
RSS ►http://www.sas.com/rss

---

Closed Caption:

[MUSIC PLAYING]

DRAGOS COLES: Hi.
My name is Dragos Coles.
And I'm a systems
engineer at SAS.
Today, we're going to use
Visual Analytics to look
at transactional retail data.
We're going to go through
a segmentation exercise.
And we'll do some market
basket analysis reporting.
The reason I want to
do all these analyses
was because I'm going to try to
optimize my marketing campaign.
And also I want to try to
personalize my campaign lists.
So again, as I
mentioned, we are working
with transactional retail data.
On the side here,
on my Data tab,
you'll notice I have attributes
about the transactions,
like a transaction timestamp,
transaction ID, a location.
I also have some attributes
about my customers
like customer ID and some
demographic information
as well.
What I want to focus on this
time though is my sales.
And I want to look at the
average transaction sale.
And let's just explore
this really fast.
Overtime, I can see that my
average transaction amount
kind of hovers around
upper 30s, 40s.
I have a couple of
spikes and variations,
but typically stays
fairly constant.
Since I'm looking at
a trend over time--
let's build a
forecast really fast.

Visual Analytics is
looking in the background
and forecasting a
model portfolio.
And it's just selecting
my champion forecast.
The idea here is
though we're looking
at this average
transaction amount.
And my goal with
this campaign is also
I'm going to try to see if
I can drive this transaction
amount to go up a little bit.
The forecasting capability
in Visual Analytics
is going allow me to do
a few things to do more
than just a simple forecast.
Let's bring in some
additional measures.
And what I'm going
to try to do is
I'm going to try to see
if I have any underlying
factors that could impact
this average sales amount.

And I see that I have
actually three underlying
factors in this case.
I have margin, 12-week
RFM, and marketing budget.
This could be intuitive
that this might be the case.
But since I have this
underlying factor,
the next thing I
might want to do
is I'm going to try
to run a scenario.
Now, let's see what happens
if I increase this marketing
budget by 40%.

You'll notice actually that
this average transaction amount
and did go up a little bit.
However, now the
question becomes
are we going to get an increase
of 40% in our marketing budget.
And the answer is probably not.
So let's take a
different approach.
Let's work with them the
decision tree capability
in Visual Analytics.
And we're going to try to
segment our customer base,
so that we can create a more
targeted marketing campaign.
So again, our target variable
is sales or average transaction
amount.
Then bring in some predictors.

In this case, keep in mind we're
not doing a predictive model.
We're not going to
get propensity scores.
What we're going to use the
decision tree capability
is to run a segmentation
and try to identify
some segments of our
customer population
to make sure that
maybe there's something
we can target in there
in our upcoming campaign.
As I take a look at
the decision tree,
a few things are happening here.
One, I can see some
of those segments,
by looking at the terminal
nodes, for examples,
within this tree.
Also, as I hover
over these nodes,
I can see what the
average transaction amount
is for this specific segments.
So in this case, this node
has a $67 average transaction.
This other node here
on this left side
is a $19 average transaction.
So right away, we can
identify certain segments.
The segments are defined by some
of the attributes do we dropped
onto our decision tree model.
And we can see we have
transaction day of week.
We have age bucket.
We have channel type defining
some of these segments.
But what we're really
trying to do here
is I want to get a
better visual of all
these segments and the
average transaction amounts.
And as I look through
this list here,
I can see that we're dealing
with three main categories.
I'm going to call this
low spenders, people
who are spending on average
under $20 per transaction.
Our average spenders,
they, again,
hover around the upper
$30 per transaction.
And finally, I'm going to call
these the high spenders, who
are above $60 per transaction.
And I want to focus
my marketing campaign
on this particular segment.
Since I've built
this decision tree
and I am on identifying the
segments, what I'm going to do
is I want to be able
to capture the segment.
And the way I'm going to
do that is simply just
create a calculated item.
We will call this calculated
item decision tree segments.

And as I create this
variable, the variable
is going to show
up in my Data tab.
I'm going to try to
capture this segment along
with the rest of my
transaction information.
So let's just capture,
maybe, transaction
ID, transaction date, time.
We'll get the segments
and our customers ID.

What I want to do
with this data set now
I want to be able to run
market basket analysis.
And I'm going to use Enterprise
Miner, SAS Enterprise
Miner, which is our data mining
solution to run data analysis.
Now simply all I have to
do is export this data
so I can pull it into
Enterprise Miner.

Once we're done with the
market basket analysis, what
I want to do is I want
to be able to report
on those results.
Market Basket produces quite
of bit of information for us.
And we're going to
use that information
to be able to fine tune, again,
the customer marketing list.
And this is the goal of
this particular report.
And it also gives us measures,
like a lift measure, which
will allow us to identify,
one, the customers
that we want to advertise
too, and, two, which
items do we want to advertise.
This gets to the personalisation
of that marketing message.
If you look into my
customer market list,
the rules are representative of
two main items-- the left side,
which tells me what items
exist in the customer's basket,
and the right side, which
tells me the items missing
from their baskets.
And these are the items I
might want to advertise.
Also, the way I've set
up this little report
is so that I have
some interaction.
And first thing that
I want to do here
is I want to select one of those
segments that were produced
with the decision tree.
And here I have
those high spenders.
As I make selections
on this report,
let's also go after
customers with a higher lift.

Again, I'm going
to make selections.
As I make those
selections, things
are going up to update and
refresh on my dashboard.
And this really gets to
sub-setting the customer base
from within this
particular segment
so that I can really optimize
and fine tune my marketing
list.
But at the same time, I'm also
personalizing the marketing
message exactly.
Which product do I
need to advertise next.
I hope you liked
what you saw today.
If you'd like to try Visual
Analytics please visit
SAS.com/VA.