Yuma 4×4

Media and Communications

8.4.3 R8. Google AdWords – Video 2: How Online Advertising Works

8.4.3 R8. Google AdWords – Video 2: How Online Advertising Works

The process by which
Google determines which ads to display
for which queries consists of three key steps. First, Google runs an auction
where advertisers place bids for different queries that they
want to display their ads on. Next, Google uses each
bid in a metric known as the Quality Score,
which basically measures how well a particular ad
fits a particular query to decide a quantity known
as the price-per-click. Google does this for each
advertiser and each query. Finally, and this is where
optimization plays a key role, Google decides how often to
display each ad for each query. This problem, as
we’ll see shortly, can be formulated as a
linear optimization model. Let’s begin by
thinking about the data that we need for this model. In particular, let’s think
about the price-per-click. So as we just discussed,
Google decides each advertiser’s
price-per-click. The price-per-click is
how much each advertiser pays Google when a user clicks
on the ad for that query. Each advertiser also
specifies a budget. This is the total
amount of money that the advertiser
has available to pay for all the clicks on their ad. Every time a user clicks
on the advertiser’s ad, the advertiser’s budget is
depleted by the price-per-click for that ad for
that user’s query. Let’s illustrate this
with a small example. So suppose that we are Google,
and three of the major wireless service providers in the United
States — AT&T, T-Mobile, and Verizon — come to us
wanting to place ads on three different search queries: query
1, which is “4G LTE”; query 2, which is the “largest
LTE”; and query 3, which is “best LTE network”. If you’re not familiar
with these terms, 4G and LTE basically refer
to different standards of high speed wireless
data communication. The table here shows
the price-per-click of each advertiser
in each query. So for example,
this 10 here means that T-Mobile will
pay Google $10 every time a user
searches for query 1 and clicks on T-Mobile’s
advertisement. In this example,
T-Mobile’s budget is $100. If T-Mobile begins
advertising and by some point five people have
clicked on T-Mobile’s ad when they search for
“4G LTE”, then T-Mobile will need to pay five times
$10, or a total of $50. If T-Mobile’s
budget is $100, this means that T-Mobile is
left with $100 minus $50, for a remaining budget of $50. Now, while the price-per-click
is important to know, we must remember that the
price-per-click is exactly that, the price that
the advertiser pays to Google for a single click of
a given ad, on a given query. This price is paid only if
the user clicks on the ad. But typically, the people who
use Google search engine, who are you and me, will
not click on every ad that is shown to them. Therefore, we need
a way to capture how often users click on ads. This is where the idea
of click-through-rate becomes useful. The click-through-rate
is the probability that a user clicks on an
advertiser’s ad for a given query. You can also think of this as
the average number of clicks that we expect per user. And this quantity is
defined, as we said, per advertiser and per query. So to illustrate this, for the
example that we just introduced a few slides ago,
suppose that we have the following
click-through-rates. The number 0.08 here means
that there is an 8% chance that a user who searches
for best LTE network will click on AT&T’s ad
if it is shown to them. In terms of the number of users
who click on an ad for a given query, this means
that for 50 users, if the click-through-rate
is 0.08, we expect to see 4 users
clicking on the ad. Similarly, if there
are a hundred users who view the ad and 8% of
them click on the ad, we expect to see
eight users clicking on AT&T’s ad for query 3. In the next video, we’ll
define additional data that we’ll need to
formulate the problem. In particular, we will see
how the click-through-rate and the price-per-click
can be combined together to obtain a new quantity called
the average price per display. This derived quantity
will form a crucial part of our linear
optimization model.

Leave comment

Your email address will not be published. Required fields are marked with *.