Measure the success of marketing campaigns

[Music]
welcome back let's talk about success
people love success because it makes
them feel good if you work on a
successful marketing campaign you'll
feel good about it but success is more
than a good feeling a common way to
measure a campaign success is by how
well potential customers moved forward
in the customer journeys
ultimate success is measured by the
number of completed sales
metrics can help determine the success
of marketing initiatives and campaigns
return on investment is one metric and
you'll probably hear it more often as an
acronym formed from each of the first
letters roi
for ad campaign specifically there's
return on ad spend which is often
referred to as another acronym roas
a business always has to make a profit
to stay in business that's why these
metrics on returns are so critical
you'll learn how to use these metrics to
measure marketing success to help
increase returns tests are sometimes
performed to determine which of two
variants generates better performing
results
these tests are called a b tests they're
also called split tests or bucket tests
they help marketers make decisions that
impact the success of a campaign
you'll learn more about these tests in
upcoming lessons and finally you'll
examine case studies of marketing
campaigns with varying levels of success
you can learn a lot from a case study
that describes how best practices were
applied you can also learn a lot from
case studies describing what to avoid or
what went wrong
success feels good so i wish you success
in this part of the course and for the
rest of the program
[Music]
do you measure marketing success you
were just introduced to the terms return
on investment or roi and return on ad
spend or row as marketers use these
calculations to help them determine the
success of marketing campaigns now it's
your opportunity to learn how these are
calculated let's start with roi roi is a
measure of the profit generated from a
marketing campaign more specifically
it's the ratio of net income or money
made to investment or money spent roi
can be calculated one of two ways
in the first method to calculate roi
subtract the marketing cost from the
total sales growth during the period the
campaign was run and then divide the
result by the marketing cost sales
growth is a positive increase in sales
when compared to prior sales
in other words it's the positive change
in sales or sales lift for example if
sales growth was 200 000
and the marketing cost was fifty five
thousand dollars the marketing roi is
two hundred thousand minus fifty five
thousand or a hundred and forty five
thousand divided by fifty five thousand
their roi is two point six
if you consider an roi of 1 as the
break-even point or when sales growth
equals marketing costs an roi greater
than 1 is on the upside the higher the
roi the greater the upside
a second method to measure roi uses
something called customer lifetime value
or ltb
customer lifetime value is the average
revenue generated by customers over a
certain period of time
software analytics tools can help with
ltb measurement google analytics 4
reports ltv for users acquired through
different channels over an acquisition
date range you specify up to 120 days
this allows you to monitor customer
behavior over several months to observe
the effect of a marketing campaign you
can find ltv in the acquisition overview
report an upward ltv trend indicates an
increasing roi
now let's shift from roi to return on ad
spend or row as
row as is a number calculated as the
revenue generated divided by the amount
spent on advertising performance goals
are created from and tie into higher
level marketing and business goals
row as targets for digital channels are
usually set relative to an overall
campaign level row as across a mix of
media a marketing goal of five to one
roast usually produces related roast
performance goals for each channel such
as three to one row as for search and
four to one row as for display the per
channel row as results contribute to the
overall row as across all media
if you're monitoring the analytics for a
campaign you might wonder when your
intervention is needed if the roas isn't
meeting your target what can you do this
isn't a complete list but here are a few
things you can consider doing
you can lengthen the campaign duration
measuring conversions is a fluctuating
or noisy process
roads can temporarily be lower than a
target due to fluctuations in the
marketplace these fluctuations may be
outside of your control so be patient
a good practice is to evaluate roast
after a minimum of 50 conversions have
been reported
you can set row as targets by product
groups instead of using an overall row
as
suppose a clothing store sells evening
gowns dresses tuxedos and suits and
wants a four to one overall roast but
isn't able to achieve that goal
initially
instead of sticking with the row as an
overall target they can split out the
targets by product groups like formal
and more casual wear
roast by product groups may provide a
more accurate measurement for the
campaign's success
review how the roast target was set
if the roast wasn't based on the roast
from previous campaigns or was based on
margins that were far too optimistic you
could reset the roes target to a more
realistic number based on the data you
now have
lastly if applicable you can adjust the
automated bidding strategy to see if
that helps you meet the target roas this
is done for automated campaigns only
as you have now learned roads is an
effective measurement for performance
measuring roads helps you evaluate the
performance and success of a campaign it
may take more practice but in time
measuring both marketing roi and roas
will become a natural part of your job
they're both key metrics and marketing
analytics
[Music]
in marketing there are a lot of choices
about content and strategy choices
supported by the data to prove it are
the best kinds of decisions an a b test
can help you choose the best content or
marketing strategy for an online
business
an a b test also known as a split test
or bucket test is an online experiment
with two variants and a random fifty
percent split of users between the
variants to determine the better
performing option
almost all a b tests randomly send fifty
percent of users to one variant and
fifty percent of users to the other
variant
for example an a b test can randomly
direct web traffic to two different
versions of the same web page
responses from those pages are monitored
and the version that achieves higher
performance based on the chosen metrics
wins
an a b test is typically performed on
live web pages it's possible to perform
an a b test on a website that isn't live
as long as the number of users tested is
large enough to produce a statistically
meaningful result let's say you want to
test two versions of a direct response
ad to join a loyalty program one version
appears alongside a celebrity
endorsement and the other appears with a
savings offer which version will result
in more conversions or people signing up
for the program
the a b test will tell you which version
is the better one using a sample
population tested during a short period
of time
in this particular test it appears that
a greater number of conversions occurred
with the savings offer an a b test
relies on statistical tests to determine
which of the two options being tested is
more effective the statistical tests
used during an a b test depend on
whether discrete or continuous metrics
are used for comparison
discrete metrics have specific values
can be counted or are binary like on off
or true false settings examples of
specific values are click-through rate
or ctr conversion rate and bounce rate
examples of counts are new and returning
user counts and finally a binary
response could be if a user clicked or
didn't click on a direct response
continuous metrics are measured and
change over time
examples of continuous metrics are
revenue per user average session
duration and average order value the
data is continuous because the
measurement changes with each additional
session or order the key takeaway is
that statistical tests are critical for
a b testing the a b testing software you
use will perform the statistical
analysis on a large enough sample of
users to be informative
when you conduct an a b test you're
usually testing if a new version will
improve a metric compared to the
original version
when you plan for an a b test it's
helpful to document past performance
desired improvement and the performance
metric you'll use for the test
your organization might even have a
template to help record this information
for each test
there are many software tools available
for aap tests
you can conduct a b tests on ads with
google ads performance marketers also
rely on a b testing to identify pages
with the best performance
google optimize lets you run a b tests
on your website content to understand
what works best for your visitors
hubspot performs a b testing on email
messages and landing pages optimizely
offers a b testing for a variety of
touch points in the customer journey
intellimize helps personalize web pages
with machine learning and what could be
considered as continuous a b testing
these tools aren't being recommended
above others but are mentioned to point
out that a wide variety of solutions are
available from different vendors
whichever tools you use if you conduct a
b tests you'll have greater confidence
that you'll meet performance goals
increase the number of conversions and
benefit from improved customer
experiences through redesigned web pages
that have been tested your choices do
make a difference
[Music]
door number one or door number two
sometimes surprises are fun
but if you're running a campaign you
want to minimize surprises when changing
an ad
you learn that an a b test typically
tests whether a new version of an ad or
web page will improve metrics compared
to the original version
this video will demonstrate how to set
up an a b test in google ads to compare
the performance of two ad variations
an ad variation can test changes to ads
like a change to a url headline or call
to action
for example you can test how changing a
headline from act now while supplies
last
to huge savings limited time offer
impact sales across ads in multiple
campaigns even a small change to a call
to action from buy today
to buy now can be tested but not all
changes require a test
it's certainly reasonable to make minor
text changes to ads without fully
testing them first
using the first example of alternate
headlines here is how you would set up
an ad variation test a type of a b test
in google ads to test them
assume that the current or existing
headline is act now while supplies last
in the campaigns page in google ads
click all campaigns
then click on experiments in the
navigation panel
then
create a new ad variation by selecting
the plus icon
select whether this ad variation applies
to all campaigns or select which
campaigns this variation affects
assume this variation applies to all
campaigns
next select that you're updating text
and that the edit is for headlines
finally enter huge savings limited time
offer as the text for the headline in
the variation
let's name this variation huge savings
use the default start and end dates for
a 30-day run and keep the default 50
split for the test
to create the variation click create
variation
after the a b test for this ad variation
runs you'll see the results in the add
variations table
here is an example of what the results
would look like in the table
you'll still see metrics like clicks and
impressions moving horizontally across
the table
each metric will show a positive or
negative percent which tells you the
amount of change there was between the
two variations tested
you'll want to pay attention to the
percentages with blue stars next to them
these stars indicate that there was a
statistically significant amount of
change between the two variations tested
if many metrics are displayed with stars
you might consider making the new
variation a permanent ad change
this was an introduction to a b testing
using an ad variation test in google ads
you can also use experiments in google
ads to test campaign bidding strategies
refer to the google ads help or a
reading for more information about
testing bidding strategies
finally as mentioned earlier remember
that you don't need to test every change
or variation
because a b tests take time and budget
it's important to be strategic and
selective about the ads that you and
your team decide to test these decisions
are unique for every team
[Music]
many metrics can be monitored during a
marketing campaign
so how can you conclude whether a
campaign was successful or not the
insights you use to evaluate the success
of a marketing campaign depend on the
marketing goal or goals and what the
campaign was trying to address
specifically but the main success factor
is going to be whether you met the
performance goals that were set
let's use two examples to understand
success indicators
in the first example an overall
marketing goal was to increase the
number of leads leads were measured as
micro conversions a micro conversion is
a completed response that indicates that
a potential customer is moving towards a
completed purchase transaction
a micro conversion performance goal of
increasing email signups by 20 was set
from past experience people who signed
up to be on an email list had a fifty
percent chance of completing a purchase
therefore a macro conversion performance
goal was set for sixty percent of people
signing up to complete purchases a 10
improvement next similar targets for
chat bot conversations and blog page
visits were set the chat bot was new so
the goal was for at least 20
conversations to occur during the
campaign blog page visits spiked by 40
recently so a performance goal was set
to increase new visitors to the blog
page by 20
macro conversion performance goals were
set
the goals were that 30 of chat bot users
would complete purchases and 10 of new
blog visitors would complete purchases
for email the results show that the
campaign increased email signups by 21
but it seemed to have no effect on
completed purchases which remained flat
for chatbot conversations both micro and
macro conversion goals were met or
exceeded
for the blog page both micro and macro
conversion goals were missed by quite a
lot
so was the campaign a success
based on the metrics there were mixed
results some parts of the campaign were
underperforming but other parts of the
campaign were very successful
several things did become clear for
future work
chatbot engagement should continue
because of the conversion pull through
email messages should be reviewed and
modified to possibly provide a better
lift in conversions and finally not too
much time or energy should be spent on
blog campaigns
in the second example a marketing goal
was to increase online sales by doubling
the average order value in other words
the average order value was 50 dollars
prior to the marketing initiative and
the goal was to increase it to 100
this also served as the performance goal
the digital marketing campaign included
direct responses so customers could
click to view additional products before
and during their checkout there was also
promotional ad for customers to spend a
hundred dollars and receive free
shipping
metrics monitored were the online sales
revenue number of orders average order
value and the cost per sale to keep
return on investment or roi in check
secondary metrics of interest were
orders by geographic region
top selling products contributing to
larger orders and the ratio of orders
from new versus returning customers
although the average order value
increased by 45 percent to 95
just under the performance goal to
double it the overall trend showed a
very successful campaign online sales
revenue went up by 15
the number of orders increased by 170
the cost per sale went up by 3 percent
orders by geography didn't change the
top selling products that contributed to
the increase in order value were small
kitchen appliances and electronics and
finally the ratio of orders from new
versus returning customers went up from
1 to 1.4
to summarize the two examples in this
video show that metrics of interest are
always dictated by the goals of the
marketing campaign the metrics used to
evaluate success were completely
different the benefit of using marketing
analytics tools is the ability to view
metrics and gather insights and these
insights help you to evaluate and define
the success of any campaign campaign
success will vary but the insights
gained always have current or future
value
congratulations on finishing this video
from the google digital marketing and
e-commerce certificate
access the full experience on coursera
including job search help and start
earning the official certificate by
clicking the icon or the link in the
description
watch the next video in the course by
clicking here and subscribe to our
channel for more lessons from google
career certificate