Latest / The Yield Doctor with James Deaker / Non Programmatic Pricing Attributes

Transcript

Non Programmatic Pricing Attributes · The Yield Doctor with James Deaker

0:00
hi today I'm going to be talking about
0:03
how to set prices in a non-programmatic
0:05
environment for digital media I'm going
0:09
to focus on the question of which
0:11
attributes are important this is almost
0:15
a forgotten art as more and more media
0:17
has moved programmatic but as we'll
0:19
discuss it's still very relevant in a
0:22
lot of different environments for a lot
0:23
of different products and a lot of
0:25
different markets I'm going to be
0:27
talking about the features that you need
0:29
to to think about to incorporate into
0:32
your pricing and some of the features
0:35
that honestly you shouldn't be wasting
0:37
your time on these are very hard earned
0:41
lessons from many years where I feel
0:43
like I was focusing on some of the wrong
0:46
attributes let's get into it let's talk
0:49
about the areas that still use
0:51
non-programmatic
0:52
advertising the first is areas where the
0:55
ad product is differentiated or unique
0:58
in some way so the programmatic
1:01
standards don't apply right now this
1:04
pertains to a lot of things like digital
1:06
out of home or maybe it's a new app
1:10
where there are new ad standards that
1:13
haven't yet been
1:15
established another example is when the
1:17
market is a lagged when it's
1:20
underdeveloped in some ways and these
1:22
tend to be actually the countries that
1:24
are developing countries India China
1:28
southeast Asia where the programmatic
1:30
standards aren't used as much and where
1:32
the buying and selling of AD inventory
1:34
still relies much more on
1:36
person-to-person
1:38
interaction the final example is when
1:40
the publisher believes that there is
1:42
value in the product that hasn't been
1:45
well represented by the uh product team
1:49
and isn't really well understood by the
1:52
market in this case the publisher may
1:54
choose to make the decision to invest in
1:56
a sales team to sell the product
1:59
directly to
2:03
advertisers let's think about what do
2:05
the advertisers care about when they're
2:07
buying ad inventory from a publisher and
2:10
again I'm using the publisher term very
2:12
broadly here basically anyone who's
2:14
creating ad inventory the first thing
2:17
they care about is the size location
2:21
duration of the ad basically how
2:24
prominent how impactful is it and when
2:27
we talk about location it could be the
2:29
physical location or it could be the
2:31
device location are we talking about a
2:33
TV PC phone Etc and then location could
2:38
also be the location on a page if a web
2:41
page loads and multiple ads are loading
2:44
is it the location that's prominent
2:46
takes over the whole screen or is it a
2:48
location over on the right rail or even
2:51
below the fold all these things are very
2:53
important when we think about the value
2:55
of the ad the second thing is the
2:58
environment and again
3:00
clearly closely related to the location
3:04
but when I think about the environment
3:05
and if I'm an Advertiser thinking about
3:07
the environment I'm thinking about is it
3:11
uh an environment that is well
3:14
understood clean not a lot of spam not a
3:17
lot of other advertisers being shown
3:20
again it's a proxy for how impactful
3:23
that ad is going to be and then we need
3:25
to talk about the audience how big is
3:28
the audience and who is the audience
3:29
audience the first is simply a matter of
3:31
numbers are we talking about tens of
3:34
people hundreds of people thousands or
3:36
millions of of people who might see this
3:38
ad but then the question also comes
3:42
up what is the audience composition are
3:45
these people who are likely to buy this
3:48
product or likely to be impacted and
3:50
swayed to buy this product and that is
3:52
often looked at through demographics
3:55
through behavioral targets through other
3:57
ways of talking about the audience
3:59
that's seeing that
4:01
product let's look at an example and for
4:04
the sake of this example I'm going to
4:05
keep the audience size the same at 10
4:08
million people let's talk about an ad
4:11
showing on a traditional TV screen it is
4:14
going to take over the full screen and
4:15
it's going to last for potentially 15
4:19
but in this case I'm going to say 30
4:20
seconds might may last for longer and
4:23
let's compare that against a different
4:25
ad
4:26
format let's talk about an online video
4:30
small player this is going to appear on
4:32
a PC or potentially on a tablet it's not
4:35
going to be full screen it's going to be
4:37
smaller it's going to be what's called a
4:38
300X 250 pixel ad you don't necessarily
4:42
need to know the size but be aware that
4:44
it's going to take up a fraction of the
4:46
screen possibly it's going to run for
4:48
really any duration it would need to run
4:51
for at least 1 second to be deemed
4:53
viewable by most of the standards but
4:55
ideally it's going to run for 15 seconds
4:58
and then let's comp compare that against
5:01
a static display ad again seen by 10
5:04
million people running on a PC let's
5:07
assume that it's the same size as the
5:11
video ad that was running in the small
5:13
player 300X 250 pixels and in this case
5:17
how long it is viewed is unknown so it
5:20
may be on the screen it may be sitting
5:22
there not being watched as someone reads
5:25
an article but it's
5:27
unknown these are three similar but in
5:30
many respects very very different ad
5:33
experiences if we think about the one
5:35
that's clearly the least valuable which
5:37
is the static display ad appearing on a
5:39
web page and make it 1X then what we see
5:44
is that the uh online video small player
5:49
is going to be somewhere between two and
5:52
five times more valuable and I
5:54
appreciate that is huge range there are
5:57
so many factors that come into play here
5:59
and the traditional TV ad is going to be
6:03
that much more valuable again it could
6:05
be 30 to 50 times more valuable than the
6:09
static display ad and I'm just using
6:11
these as ballpark numbers to sort of
6:13
show that even though we've really only
6:15
changed two or three attributes the
6:18
device that it was loading on and how
6:20
long it was seeing we get very different
6:23
results in terms of the value of those
6:25
product so I'm going to call these the
6:27
tier one attributes that you need to
6:28
think about when you're pricing product
6:31
the the next tier that we need to think
6:33
about that I'm going to call potentially
6:34
important in things that often consumes
6:38
the time of a pricing yield team is they
6:42
basically have to look at them because
6:44
they you know often can't change the
6:46
tier one attributes if you're a website
6:48
owner you're probably not going to be
6:51
able to change those tier one attributes
6:53
very
6:54
much so one of the first things we think
6:57
about is viewability
6:59
many years ago the key Dynamic was is
7:02
the ad showing above the fold meaning as
7:05
the user comes onto a web page do they
7:08
see the ad immediately or do they need
7:10
to scroll down quote below the fold in
7:14
order to see the ad that definition has
7:16
somewhat gone away now and people just
7:18
talk about the overall viewability of
7:21
the product but the second thing is the
7:24
ad serving priority is it a guaranteed
7:27
ad unit or is it Remnant meaning that
7:29
it's not going to guaranteed to serve on
7:32
the page in any given time and so it's
7:35
lower priority and then the third thing
7:37
to think about is the audience type
7:40
demographic or behavioral so in other
7:42
words a demographic example might be is
7:45
the audience primarily targeted that men
7:48
18 to 34 or women 18 to 34 Etc there
7:52
would be a demographic example a
7:55
behavioral example could be that the
7:57
website has knowledge about this person
7:59
and knows that they're going to be you
8:01
know some series of um interest in a
8:05
given category maybe they uh shown all
8:07
the signs of being interested in buying
8:09
a car that would make them an auto
8:11
intender these types of things can be
8:14
very valuable and so some people might
8:17
take issue with the fact that I'm
8:19
calling this a tier 2 kind of uh
8:23
attribute rather than tier one because
8:25
the difference between a very valuable
8:28
audience and a not valuable targeted can
8:31
also be very significant but to come
8:34
back to the previous example we just
8:36
talked about you're going to get tv ads
8:38
that are untargeted that can be very
8:40
valuable if you get a very targeted ad
8:44
to a very specific audience but it's
8:46
showing up on a web page in a very small
8:49
ad format I'm going to argue it's not a
8:52
very valuable ad and so these would be
8:54
the secondary consideration things that
8:56
you should think about and you should
8:58
think about them this is on the list of
9:00
things that that should be thought about
9:02
let's move our attention though to
9:04
things that people do spend a lot of
9:06
time uh looking at but I'm going to
9:08
argue possibly don't need to be looked
9:11
at and one one or two of these are
9:13
controversial so one is the sell through
9:16
rate and here uh let's talk about what
9:20
sell through rate means first of all
9:22
sell through rate is the idea that if
9:24
I'm a publisher I could be showing ads
9:28
every time my page loads but if I don't
9:31
have advertisers who want to put them in
9:33
there I may end up having to show House
9:35
ads or Remnant ads or basically it's
9:40
unsold there is a very different Dynamic
9:42
within a publisher when they are sold
9:45
through and they are you know have 80 90
9:48
100% sell through on a page then when
9:51
they have low sell through 20 30
9:54
40% and they people will be running
9:57
around scared within a publisher if the
9:59
sell through gets too low from an
10:01
Advertiser perspective though they
10:04
really don't care unless they actually
10:07
can't get the product they want there
10:09
are specific examples when you might
10:12
have a homepage that's heavily sold
10:14
through and advertisers really want to
10:16
be on it that may influence the pricing
10:19
but in general this is something that I
10:22
see a lot of people spending a lot of
10:23
time on I have spent a lot of time on
10:26
looking at how demand is going to come
10:28
in and in hindsight is probably
10:31
something that we should have really
10:33
deprioritized it's something that causes
10:35
a lot of internal churn within the
10:37
publisher but really probably shouldn't
10:40
be moving the value or the price that
10:42
it's represented in the Market at very
10:44
much at all the second thing is a
10:47
frequency cap it's the idea of how often
10:50
do you want to show a given ad to a
10:53
given user and this may be kept to say
10:56
okay I'm only going to allow a given
10:57
user to see that ad three times per day
11:01
it's potentially valuable it potentially
11:04
moves the value but again I'm going to
11:06
put it in that category of things where
11:08
probably too much time and attention
11:11
over the years has been put on this
11:13
attribute in very specific cases it can
11:16
it can change the value of the ad
11:18
product but it's not where I'd be
11:20
starting and then another one is time of
11:23
week or time of year and I actually did
11:25
put an asterisk on this the asteris
11:28
relates to the fact that it is known
11:30
within digital media that prices go up
11:33
as you get closer to the holiday season
11:35
and in Q4 in particular you may see the
11:37
price of inventory go up anything 10 20
11:41
30% and so during those peak times yes
11:45
you need to potentially move your prices
11:47
up but for the vast majority of the year
11:51
you probably actually don't need to be
11:53
moving your prices around very much at
11:55
all time of week is also a similar
11:58
dynamic where weekend traffic tends to
12:01
be lower and therefore you could argue
12:04
pricing often needs to be lower but this
12:06
is also in the category of kind of set
12:08
it and forget it if you've established
12:10
that your weekdays are more valuable
12:12
maybe you set up a different price on
12:14
your rate card but I wouldn't be
12:15
spending a ton of time digging into uh
12:18
time of weak effects and now we get to
12:22
the category that I'm calling don't
12:24
waste your time and one example is cross
12:27
elesy I've talked about about this in
12:29
other in other videos but the example
12:33
here is should you be worrying about the
12:36
price of one add product uh in
12:38
conjunction to another so for example on
12:41
your homepage if you are raising the
12:44
price of the banner that appears at the
12:46
top of the page but it starts to sell
12:48
out and you change the price should you
12:50
be worrying about the price of the other
12:52
ads on the page or on other pages that
12:55
you control I have spent a lot of time
12:57
on this and wasted a lot of time don't
13:00
worry about it there may be some Corner
13:02
cases where it is useful but it pales in
13:05
significance to the self elasticity
13:08
which is to say the price demand
13:10
response for an individual product in
13:13
that same category and again this is a
13:15
reminder that this is for
13:17
non-programmatic I'm going to put time
13:19
of day which is an interesting one
13:21
because in this particular case time of
13:24
day means there we know that there are
13:27
times that are more valuable around
13:29
lunchtime if you've got an advertis at a
13:31
fast food restaurant they may want to
13:33
get access to certain inventory at
13:35
certain times of the day but the cave
13:38
art here is it's really hard to sell to
13:41
those individual advertisers in a
13:44
non-programmatic fashion if when I do
13:47
one of these on programmatic pricing
13:49
time of day may be one of those features
13:51
where I say yeah put it in the hopper of
13:53
things that need to be thought about
13:55
when we're moving around pricing floors
13:57
but for non-programmatic pricing don't
14:00
waste your time so with that and
14:03
particularly the things not to waste
14:04
your time on I hope that this is
14:06
valuable to all the pricing and yield
14:08
teams out there in digital media thanks
14:10
for taking the time to review this
14:11
material with me