Coursera || Marketing Digital – Display Advertising at Grainger – Sem 1 – 10

[MUSIC] Hi and welcome to the session. Today we are going to talk
about display advertising and we have with us Beth,
who's the head of display at Granger. And we will be talking to her about
how she does her job at Granger. And in the process learning some of the real world skills that you will be
able to use in your Capstone project. Beth, thank you for
taking the time to chat with us. Could you introduce yourself? >> Yeah absolutely. Well hello, my name is Beth Distol and I am a digital marketing
manager here at Granger. Spent about three years with
the company and prior to that I had an academic career at another Big
Ten university in an advertising program. And spent a majority of my
early career working in the agency side of the business,
when the media world was much different. So I have my roots in traditional media. TV, radio, outdoor, newspaper, etc.

But then, as the media world changed, I started getting much more
entrenched in the digital space. And have spent the past six years
really doing marketing automation and behavioral targeting
through display media. >> Great.
So we're talking about display advertising today and display is sort of one of
each channels that Granger uses. But for our students,
tell us what display means for Granger. Well, it can mean a lot of things. I think it's most rudimentary definition would be
visual advertising on the internet. So that could be everything from
app that you see on your phone. It could be a video that
you see on YouTube. It could be a banner add
that you see on ESPN.com. >> And in terms of volume and
sort of importance of this display in the list of channels that Granger uses,
where would you place display? >> So display is the second highest funded
element in our digital marketing mix.

Behind paid search. There's a lot of synergy between
paid search and display obviously. But we spend many millions of dollars
on our investment in display. >> Great, great, great. Just take us through what is the workflow
when you decide on a campaign. And then take us through some of the steps
that you might follow in deciding on where to put the ads and how to target
them and perhaps some examples. >> Yep, absolutely. So, one thing that I would like to touch
upon first, before I get into the kind of the mechanics of the campaign, is how
display advertising is purchased today. So it's changed dramatically from
the early days when you would buy a banner on a specific web site to
reach a specific demographic. Most all of what we do now is automated,
so we do behavioral advertising based on signals we're able to capture
on people that are on the Internet.

And then we purchase all of our
inventory in what is now called a programmatic marketplace. So essentially,
all of the big websites with high traffic take a certain percentage of
their inventory and they put it on what they call the open ad exchanges,
which is much like the stock market. Where marketers decide how much they
want to bid for an impression on that particular website to reach that
particular customer or prospect. So everything that we
do is programmatic and I think it's important to understand that. It's a very efficient way of doing
business and the media industry is trying to get more digital media to be bought and
sold programmatically.

Eventually the hope is that
television will be bought that way. But it allows us to reach a very qualified prospect as opposed to
a broad demographic. So in regards to actual
mechanics of campaign setup, first we would need to agree what
the objectives of the campaign are. Are we trying to influence
someone's awareness or get in their consideration set? Which I know is one of the phases in
terms of a customer decision journey or a consumer decision journey. Once we decide what the particular
objective of the campaign is, then we identify who the best prospect is. So, I'll give a specific example here. If we're trying to sell more
lighting products at Granger, during the month of November,
we might use our data to identify people that are Granger customers
coming to our website. But aren't buying GE lighting products,
for instance. So we're able to reach that particular
customer and give them an incentive. Maybe it's a discount, 20% off. So that would dictate
what the messaging is. We might run that campaign for
about a month. We look at the metrics, so
I know the students are probably somewhat familiar with some of
the analytics that we would look at.

>> Yes.
>> So we look at conversion rate, repeat visits, time spent on the site,
engagement metrics like that, when they actually purchase,
what the average order is. We look at all those
things to dictate how we optimize our strategy when
the campaign is live. In most cases because our display
channel's a big part of our e-commerce efforts. We're looking at driving incremental
sales, so incremental return on ad spend. So essentially for
every one dollar we spent in advertising how many dollars in sales
are you generating? So typically that's how we
would optimize our campaigns. >> Very good. I know you said you've been in
the industry for quite some time. So comparing what you're doing now,
in terms of digital advice, to what you would have done in traditional
media, what are the main differences that make possibly ROIs so much you know first possible to show ROI
and then perhaps easier to track ROI.

Any thoughts and the differences between you know what
you use to do ten years back to today. >> Yeah absolutely,
well traditional media was bought based on trying to identify time of day, day of week, demographics,
the type of content which are all really just indicators of who might
be in the market to buy your product. So, for instance, you might run at
TV ad for Kraft macaroni and cheese at 7:00 in a prime time show to try to
influence a mother's purchasing decision. Whereas, for digital media today what
we do in the programmatic marketplace is identifying someone who's indicating,
who's showing signals of purchase intent, and then reaching them
wherever they happen to be, whenever they happen to be there. Because those signals
are most relevant to them. So rather than trying to
artificially engineer our targeting, it's really following that
particular individual.

So the hyper-targeting ability
of it is really very different. >> So just to clarify, perhaps for
me and also for the students, do the ad purchases and
these ad offerings, display offerings, how much of this can be
targeted to meet the user? For example, one of the frameworks
the students have learned about is catching sort of the consumers or
customers at the zero moment of truth. It's before they even realize
that they're doing an active search through a concentration
set of brands. I'm going to, may or may not be one of those even before
people land up with that, at that space. So do the campaigns that you run,
are they, can they be run completely automated based
on user profiles, and user behaviors? And how much of daily interactions
are involved to tweak the algorithm. >> So to most effectively
run a programmatic campaign, it's a little bit of art and science.

So it's machine and man. Artificial intelligence doesn't completely
rule out the need for human intelligence. We bring a lot of customer insights, and
we help determine what the strategies are. We keep a close eye on the algorithms,
they work very effectively for us, but will come in and make adjustments to them
based on how the campaigns are doing. >> Okay. And these algorithms are there for buying
inventory and also serving inventory? >> Both.
Yes.

But in terms of the zero moment of truth. We do know a great deal about
our customers and prospects. And so, we're able in our messaging
to customize their experience. So if they haven't been to
our website in a while, we might reach out to them with
a video that reminds them why Granger brings great business
solution to our customers. Once they've been to our website, we may serve them an ad for
a product that they just recently browsed. So, very personalized and
based on where they are on their journey. >> Okay, give us, I know, of course, you have a lot of data on what's
happening on granger.com itself.

Give us a sense of how display
advertising leverages data that's not on your platforms. What data do you leverage and
how do you leverage that? >> So there's a lot of third party data
as well as our own first party data. We have a digital management platform. A DMP which, if you follow the industry
that's becoming something more and more brand marketers
are really relying on is the strongest signals
that you can capture. So that could be data that we
collect from direct marketing.

It could be from our paid search,
our SEO programs, our social programs, our email programs. So even though it may not be
display data that we have, it's data that we do
have about our customers. There's also a wealth of
third party data out there. XL 8, Blue Kai, Bizo,
that companies collect that is not necessarily representative
of our customer base but may have a lot of signals
that are important to us. Someone who's a frequent traveler,
for instance, might be someone whose of
value to Granger to message. So first party data, third party data,
are very valuable signals. The industry has put a lot more
regulations on search data.

So search query data is something that we
used to be able to leverage a lot more. Google makes it a lot more
difficult to use that data now. So I would say of all the signals,
the first party data is the most valuable. >> Okay good, oh great,
that's great overview. I know we'll get into more detail on some
of these topics in future videos but thank you for your time today. >> Yeah absolutely thank you. >> Thank you for listening in and we'll give you more updates from
Granger in the next videos..

As found on YouTube