MEASURE AND OPTIMIZE SOCIAL MEDIA MARKETING CAMPAIGNS
Facebook | Anke Audenaert
WEEK 1: EVALUATE YOUR MARKETING
RESULTS AGAINST GOALS
1.1 Reporting in Facebook Ads Manager
Tabs in Ads Manager Reports
When Ads Manager shows you your reports,
there are three tabs you can select: campaigns,
ad sets, and ads. Your dashboard will
automatically default to show the campaign tab.
Campaigns
In this tab, information in the dashboard is
organized by the ad campaign it’s in. The
dashboard will automatically show you a list of
your active and inactive campaigns, the
performance of each campaign, and more.
Ad Sets
In this tab, information in the dashboard is
organized by ad set. By selecting a campaign in
the campaign tab, you can narrow the list of ad
sets shown here to the selected campaign.
Ads
In this tab, information in the dashboard is
organized by ad. Information includes ad
relevance diagnostics.
Metrics in Ads Manager reports
Each column in an Ads Manager report provides
a different status or measurement that you can
use to analyze the impact of your campaign.
Now, we’ll go through the main columns that
appear in a default campaign report.
Status and Campaign Information
For the default dashboard settings, the first two
columns show Delivery and Bid Strategy.
Delivery
The delivery column lets you know whether your
ad campaign is running or not. Possible delivery
statuses include: off, active, completed, or in
review. Here, we can see that our example
campaign’s ad campaign is active.
Bid Strategy
This column shows the campaign bid strategy.
Bid strategy choice determines how much you
spend on bids in ad actions. Our example
company chose two different bid strategies,
which you can see reflected here.
Result Metrics
Results metrics for your campaign can give you
a sense how many people your ad campaign
has reached and the results your campaign
achieved. We’ll cover three types of result
metrics, although there are more.
Results
Results are the number of times your ad
achieved an outcome, based on the objective
and ad delivery optimization you selected. Since
our example company chose to optimize its
campaign for link clicks, their results are
reported as the number of link clicks the
campaign generated.
Reach
The number of people who saw your ads at least
once. Reach is different from impressions, which
may include multiple views of your ads by the
same people. For our example company, the
number reported here is the number of people
who saw an ad for each specific campaign.
Impressions
This metric measures how often your ads were
on screen for your target audience. For our
example company, the number reported here is
the number of times their campaign ads were
shown.
Cost Metrics
Cost metrics are metrics that cover specifically
how the budget is being spent. They’re the last
metrics in the row of a default Ads Manager
report.
Cost per result
This indicates how cost-efficiently you achieved
the objectives you set in your ad campaign. It is
the ratio of total amount spent to number of
results. Since this ad campaign objective is link
clicks, the company’s cost per result is reported
as cost per link click.
Amount spent
This is the estimated total you've spent on your
campaign. For example, this column may display
$0.00 for a campaign that's in review, or $6.57
for a campaign that was recently approved and
started running. This number should always be
less than your budget.
1.2 Understanding ROAS and ROI
Return on Ad Spend
ROAS is simply a way to find out how much
revenue you made on your advertising versus
what you spend on it. The calculation for ROAS
is Revenue made from the ads divided by
Advertising Costs of those ads.
It's a good indicator of knowing the effectiveness
of your campaign and not only if its profitable,
but how profitable. You never want your ROAS
to break even at one to one or be lower than
one, because that means you're losing money
on your campaign.
Page | 1
MEASURE AND OPTIMIZE SOCIAL MEDIA MARKETING CAMPAIGNS
Facebook | Anke Audenaert
Return on Investment
ROI looks at bigger picture spending and profit
and takes into account the total costs related to
advertising and sales, like how much the product
took to make.
 ROI is a metric that takes all the money
you spend into account. It doesn't only
look at the advertising dollars you spend
and the revenue you get in return.
Instead, it looks at all the money you
spend, and the net return or your profit.
 means our results were high.
TO IMPROVE ROAS
Because ROAS is based around just two
variables, there are only really two actions you
can take to improve the ratio.
 Either lower the cost of your ads to
get a better return or increase the
revenue you generate.
 Ideally, you could do both. You can do
this by looking to see if you're wasting
time targeting the wrong audience, or by
not using the right ad placements.
 You may also have a bidding strategy
that doesn't work for what you're trying
to accomplish.
 You can also look at ways to simplify the
buying experience once the customer
gets to your website.
1.3 Understanding Cost Per Result
Cost per result looks at ow much do you spend
per customer action, like a link click, app install,
or conversion.
We don't know whether our cost per result is
good or bad, which is why, as we've discussed
before, many of these measurements need to
be looked at in context for them to make
sense.
The first example is cost per impression. If
we're a new business and want to get our
advertising out to as many people as possible,
spending $200 and gaining 1,000 impressions in
return at $0.20 an impression may be a great
use of our budget.
 Our cost per result was low, which
 And while we may not yet know how
many impressions will convert to sales,
our goal at this point may be to just get
the word out.
Our cost per result is $50, which may sound
expensive, but knowing that each lead typically
results in a contract for $2,000 per year, $50 a
lead sounds relatively cheap.
Cost per result isn't a static number either and is
affected by several settings in Ads Manager
like:
 the auction bid
 the settings for the audience
 the optimization type chosen
 ad placements or where the ads were
placed
 whatever schedule was set
But this means that you can compare your cost
per result based on your chosen settings to find
more ways to improve your market efforts.
 For example, you may have a lower cost
per result with a certain demographic,
which means you may want to keep
marketing to them.
 You may see another demographic with
higher cost per result where you'd
reconsider your marketing efforts.
 You may see a lower cost per result
from certain ad placements over others
or uncertain schedules.
Page | 2
MEASURE AND OPTIMIZE SOCIAL MEDIA MARKETING CAMPAIGNS
Facebook | Anke Audenaert

acquiring that customer versus how

much they'll eventually spend with you.

Average Purchase Value or the average

amount a customer spends per purchase.

If Inu and Neko made 30,000 in sales and the

average cost per sale was $30, then they only
spent $6 to make $30 for a return of $24 which
is pretty good. But while that's the correct metric Average Purchase Frequency Rate, which is
for cost per result for the conversion campaign, how many purchases you typically get from one
that's not actually the correct number for their customer.
return, right?
 In addition to what Inu and Neko spent
on the conversion campaign, they also
have to factor in what they spent on the
awareness and the consideration
campaigns.
 Since the 1,000 people who purchased
products came out of the 1 million Average Customer Lifespan or how long a
impressions and the 8,000 website customer relationship typically lasts with your
visits. company. You may notice that from experience
 So Inu and Neko would have to use the or by looking at your churn rate, or if you're
entire budget spent or $19,000 to figure younger company by forecasting.
out their cost per sale. Which would be
$19,000 divided by the 1,000 purchases
for a result of $19 spent per sale.
 With a return now of only $11 per sale
Inu and Neko can decide whether
they're spending was worth the dollars
or not.

Note that this is also an example of customer Sample Problem:

acquisition costs (CAC), in that Inu and Neko
spent $19 to acquire a customer.
 If those customers only make one
purchase of $30 and never return, it
won't be a good investment.
 But if they become loyal customers and
go on to have lifetime values of a few
thousand dollars, then spending $19
now is a bargain.
LTV for Marketing Efforts:

1.4 Understanding Customer Lifetime Value
Customer Lifetime Value, or LTV, is a metric
that specifically takes into account what happens
after that first purchase. LTV is how much your
customer will spend with your business over the
span of their relationship with you.
 LTV puts more than a price tag on your
customers. It puts a value on them as
you can track their loyalty to your
business. A higher LTV will lead to a
customer who continues to come back.
 LTV can also help you see whether
you're spending the right amount on
 If one segment has a high LTV and
another has a low LTV, investigate the
group with lower LTV and compare them
to the higher LTV segment.
 Are those customers abandoning your
products? When are they abandoning
your products? Now you can look at
why.
 For segments with higher LTV and more
loyal customers use them to create a
new lookalike audience, so you can
target your advertising to new customers
who have the same characteristics as
those already giving you the most
business.

Page | 3
MEASURE AND OPTIMIZE SOCIAL MEDIA MARKETING CAMPAIGNS
Facebook | Anke Audenaert
 Because you're looking not just at
monthly or quarterly sales, but sales that
extend years or even decades, LTV
forces, big picture thinking. Now you're
looking at customers, not just as a one
of sale, but as a long-term relationship
to grow and cultivate.
 ne of the biggest things looking at LTV
can do is to help your business make
better decisions in regard to customer
relationships. If your LTV is low or you're
seeing shorter customer lifespans, then
you know you need to invest in retention
initiatives and customer support.
Increase LTV through:
 Make improvements to your
customer service. It's not unusual to
hear about customers abandoning
brands after even one poor interaction
so make sure the service and attitude
you're offering your customers is best in
class.
 Re-evaluate the content you're
distributing, increase its value to
your customers so they stay
engaged. Offer resources like articles
and blog posts and use social media to
interact with customers or to distribute
useful informative content.
 Improve your listening. Follow
conversations about your brand and
products on social media. Gather
feedback from your customers.
 Build relationships with them and
solve pain points. Feeling hurt will go a
long way towards cultivating and
keeping those long-term relationships.
 In terms of products, you could
change your subscription model to a
longer-term horizon like offering
annual contracts instead of monthly to
commit your customers longer. Or offer
discounts, rewards or loyalty programs
to keep customers engaged and feeling
like they're getting value from your
business.
WEEK 2: MEASURE YOUR ADVERTISING
EFFECTIVENESS
2.1 What Are Experiments?
You can use experiments to measure the effects
of your ads on the results you're hoping to
achieve.
 For example, if one group sees your ads
and another group doesn't see your ads,
you can measure how many results or
sales you achieve from the group that
saw your ads and compare that to the
number of results of sales you see from
the group that didn't see your ads.
 The difference in results or sales can
then be considered the result of your
advertising.
Experiments are carefully planned ahead of
time. That's because they involve creating
groups of subjects that receive different
treatments. If you want to run an experiment to
test your advertising effectiveness, you have to
plan it before you create and run your
advertising campaign.
2.2 Using Experiments to Measure
Advertising Effectiveness
To compare the effect of changing one variable
between two groups of people, it's important that
two groups of people be a similar as possible.
To create two similar groups, you can use
randomization.
Randomization is the process of signing
people to the different groups, by chance or
randomly. By doing that, you distribute people
with certain characteristics equally over the
groups, and that way you create groups that are
as similar as possible.
 Experiments in advertising use
randomization to assign people to a
control group or a treatment or test
group. Studies that use this type of
experiment are referred to as
Randomized Controlled Trials or
RCTs.
EXAMPLE
Imra, the owner plans a campaign to drive
online sales of the new fall bookcase she's
created. She's interested in knowing whether it's
worth it to spend money on advertising, so she
plans to run an experiment. Here's what would
need to happen for the experiment to really
prove to her what the effect of the campaign
was.
 First, you'd identify a group of people
you'd like to show the campaign to, in
other words, the target audience.
 Out of this target audience, people will
be randomly assigned to either the test
group or the control group.
 People who are assigned to the test
group may be exposed to a treatment. In
other words, they may see Imra's ad.
People in the control group don't receive
that treatment. They won't see the ad.
 After the campaign runs. Imra can
compare the results between the control
and the test group. In this case, she'd
look at the difference in conversions or
sales of the fall bouquets in the test
group-- the group that saw the ads,
compared to the control group-- the
group that didn't see the ad.
 If there are significantly more
conversions or sales from the test group
Page | 4
MEASURE AND OPTIMIZE SOCIAL MEDIA MARKETING CAMPAIGNS
Facebook | Anke Audenaert
than the control group, we can
confidently say that the advertising
campaign worked and had a real impact.
Well-designed tests use intention to treat or
ITT. ITT keeps all experiment participants in the
group they were originally assigned to,
regardless of what they see or do. This
minimizes the effect of potential bias.
 Whether or not you actually receive the
treatment, you will remain in the test
group, and your result will be part of the
test group.
 When you use intention to treat, it
reduces the effect of the treatment,
which makes it a bit harder to achieve
significant results. But it keeps bias out.
But if you plan to use experiments in your role as
an advertiser, make sure that these two
conditions are met: the experiment should use a
randomized controlled trial design and adhere to
ITT or intention to treat for the test group.
2.3 Experiments to Measure Ad Effectiveness
on Facebook
Facebook offers two different randomized
controlled trial tools: holdout tests and brand
survey tests. Both are tests that use
experimental designs, but they differ in the
outcomes they measure.
 Holdout tests measure the effect of ads
on conversions, which could be sales,
add to cart, leads, and so on. These
tests are sometimes also referred to as
conversion lift tests.
 Brand survey tests measure the effect
of ads on changes in brand recall and
other brand-related measures. You'll
also hear people call these tests brand
lift tests.
 In other words, holdout tests are all
about action, whereas the brand survey
test is all about attitudes.
HOLDOUT TEST
A Facebook holdout test is used to measure
conversion lift, which means that they measure
whether your campaign caused an increase in
conversions.
 A holdout test begins by randomly
assigning people from a campaign's
target audience to either a test or control
group.
 The test is referred to as a holdout test
since the control group is basically held
out from being exposed to your ad.
For Facebook to detect whether the people
tested took any of these actions, the advertiser
needs to pass some data back to Facebook.
 If the action is expected to take place on
the advertiser's website, advertisers can
pass data through the Facebook pixel.
 If the action monitored is app-related,
the data needs to be passed back to
Facebook via the Facebook SDK.
 f the action happens offline, like a sale in
a store, for instance, data needs to be
provided through a Conversions API. A
conversions API involves connecting
data from a system that monitors offline
sales to Facebook's system.
BRAND SURVEY TESTS
A brand survey test is used for campaigns that
have a goal related to a brand outcome, for
instance increasing brand awareness.
 All the people that were part of your test
group and the people in your control
group will be served one or a few polling
questions.
 The main question people will be asked
is: do you recall seeing an ad for this
brand online or on a mobile device in
the last two days?
 This is how you measure ad recall lift,
which is an important metric in
evaluating a brand survey test.
Next, you'll measure the difference between the
ad recall in the test group versus the control
group. The effect you're seeing on brand recall
is caused by the ad.
If you're focused on increasing brand
awareness (awareness and consideration on
the purchase funnel), you'll opt for a brand
survey test.
If you're focused on sales or other
conversions (conversions on the purchase
funnel), a holdout test would be the right test
for you.
THERE ARE TWO MAIN REASONS WHY YOU
SHOULD CONDUCT THESE TYPES OF
TESTS:
1. Prove advertising effectiveness. As a
marketer, you'll often be asked to prove
that your ads had an effect, so you can
use these tests to answer that question.
2. Optimize your advertising spending.
Once you know what the effect is of your
ad campaign, you will have more
information to make future spend
decisions. You can, for instance, decide
to shift more of the budget to the
campaigns that prove effective, or you
can create similar campaigns in the
future.
2.4 Creating a Brand Survey Test and a
Holdout Test
After Imra creates her campaign, she can set up
the brand survey using the Facebook
Experiment tool. This is part of the business
tools in Facebook, which you can find by
navigating from your Ads Manager here.
Page | 5
MEASURE AND OPTIMIZE SOCIAL MEDIA MARKETING CAMPAIGNS
Facebook | Anke Audenaert
***It's important to remember you must run an
experiment after you create a campaign, but
before you start to run the campaign. Otherwise,
you can't create a proper control group.
To be eligible:
 You have to have a certain level of
money spent in your account.
 Facebook also suggests that in order to
run a successful test, you should at least
spend about $30,000, on advertising in
the campaigns you’re testing.
The first section of the brand survey test focuses
on the test details. You start by specifying what
you're testing. You can test the effect of all the
campaigns that are running in your account
together. You can also test the effect of a group
of campaigns, or you can test a specific
campaign.
2.5 Interpreting the Results of Brand Surveys
and Holdout Tests
In this case, there's a 13.9 increase in brand
lift. In other words, the percentage of people in
the test group who recalled your ad was 13.9
points higher than the percentage of people that
recalled the ad in the control group. This number
is calculated by subtracting the percentage of
desired responses observed between the test
group and the control group. In this case, that's
28.8 minus 7.9 percent.
 You'll also get a cost per brand lift,
which reflects how much it costs you in
advertising spent for each additional
person who gives a desired response.
 You'll also see the total brand lift or the
estimated number of people that are
likely to respond favorably to your brand
as a result of your advertising.
Sales lift tells you how much value was
generated from those additional conversions.
Then you also get an evaluation of the results
compared to the costs.
 First, you see the cost of each additional
conversion cost by your ads, and next to
it, you get an evaluation of your return
on ads spent or ROAS.
 In this case, you achieved a ROAS of
10.7, which means that for each dollar
you spend, you'll generate $10.7 in
revenue.
WEEK 3: OPTIMIZE YOUR AD CAMPAIGNS
3.1 Optimize Your Ad Campaigns with A/B
Testing Overview
A/B testing is the process of comparing two
variants of an ad against each other to
evaluate which one performs best. A/B testing
can help reveal ways you can make your
campaign better while it is still running.
 We always found small changes to the
user experience could have a very big
impact on the revenue generated from
search ads on those search results
pages.
When you create your A and B version pf
what you're about to test vary only one
variable.
I can run an A/B tests by considering this ad to
be my A version and creating a new ad that will
be my B version. The best practice in A/B
testing is to make the B version differed in only
one variable from the A version.
 In this case, I'm going to vary only the
image. Instead of a blue stone ring, I'm
going to show a ring with three red
stones.
 I will keep the ad copy the same, and I
will also keep the Call-to-Action Button
to say, Learn More.
 Since everything else in the ad was the
same. I know the image was the reason
my results were better. Now I know that
under these circumstances this image is
best, so I can keep this as my optimal
image.
 I may decide that I want to test some or
in fact, maybe I could improve the copy
so I can conduct another A/B test.
 Finally, I can test the Call-to-Action
Button. I start with the image and the
copy that I know is optimal, and I keep
Learn More as the call to action in my A
version of the ad. For my B version, I
keep everything constant, but I changed
the Call-to-Action Button to Shop Now.
To get the most out of your A/B tests, test
only one variable at a time.
3.2 Set Up an A/B Test
After Imra set up her campaign, she can indicate
that she'd liked to run an A\B test by selecting
the campaign in her ads Manager Dashboard
and clicking A\B Testing. This will make the
selected campaign the A version in the test.
 It will then prompt you to change one
variable in your campaign to create the
B version.
 Facebook will automatically show the
variables you can change based on the
specifics of your campaign.
 In our example, Imra would choose to
change the image, all the other aspects
of her campaign will remain the same.
 Imra is then prompted to upload the new
image for Version B. In this case, that's
the close-up flower picture.
 You then must give your test a name
and you can adjust the budget.
Page | 6
MEASURE AND OPTIMIZE SOCIAL MEDIA MARKETING CAMPAIGNS
Facebook | Anke Audenaert
**Note that the budget will be split 50-50 over
the two ad sets. You should make sure that the
budget is big enough to see meaningful results.
You then also tell Facebook how you will
determine the winner. This will depend on your
goal and how you're tracking success.
 For Imra, the campaign is focused on
generating traffic to her website.
 She's optimizing her campaign for link
clicks. She decides that the ultimate
test for her should help her understand
which image generates her most results
for the amount she's spending.
 So cost per result is the right metric to
help her determined the winner.
**Note that there's an indication here of the
estimated power of the test. This is the likelihood
that this test can detect a difference in your ads
if there is a difference to detect. To have a good
test, the power should be at least 80 percent.
You can increase the power of your test by
increasing the budget or the time you're
running the ads. Both will increase the number
of ad impressions, which will give you more
actions to compare in each group.
BEST PRACTICES FOR A/B TESTS:
 It's best practice to only vary one
variable at a time.
 You should make sure that your test has
enough power so you will get a reliable
result.
 To achieve that, you should make sure
that you have enough budget for the
campaign and that you run the
campaign long enough ideally, two
weeks or longer.
 Use the winner of your test as the new
baseline and test against another
version to see whether you could further
optimize your campaign.
3.3 Interpret Results of an A/B Test
Two Main Parts of Your Results
1. You'll learn who the winner of your test
is. In advertising, the winner is the ad
that led to the lowest cost per result.
2. You'll get information on the power of
your test. In other words, how good your
test was and how confident you can be
in the results.
Next Steps:
 Shift budget to the winner. Facebook
will prompt you and ask you whether
you would like to shift all your budget for
the campaign you ran the test for to the
winning ad set. Choosing to do this
makes the losing ad set stop running.
Only the winning version of the ad is
then served.
 Continue testing. Once you have a
winner from an A/B test, it's still possible
that there may be an even better
execution for your ad campaign, so why
not try another A/B test? This is an
especially good idea if you have a
strong hypothesis that another version
of the ad might perform better.
 Use what you learned in future
campaigns.
3.4 Measuring Across Channels with Media
Mix Models
Media Mix Modeling
Marketing mix modeling can help you analyze
last year's data to understand what worked,
predict what will work in the future, and plan how
to best spend your money next year to improve
the results.
 You would first collect data related to
any previous marketing action. More
specifically, you would look at data on
how much you spent and what results
you got.
 For instance, how much money did you
spend on TV ads and how many people
saw your ad as a result? How much was
spent on promotions and how many
purchases resulted? How much money
did you invest in Facebook ads and how
many conversions did you get, and so
on.
 Note that the results data you get from
different platforms are going to be
different. In some cases, you'll collect
several views, like in TV ads for
instance, in other cases, a number of
purchases and so on.
 All that data will then be cleaned,
categorized, and analyzed using a
marketing mix model. The result of this
analysis will show you the ROI of each
channel or the percentage contribution
of each channel or marketing activity on
your result.
A marketing mix model is a mathematical
equation that shows the statistical
relationship between variables and a KPI. In
other words, it's an analysis that shows the
relationship between the investments you made
and the final outcome.
 Most marketers will rely on statisticians
or data analysts to help them with this.
Very often, they'll enlist the help of an
agency or a research firm to do the
analysis for them.
The research agency analyzes the data and
comes back to you with a report that tells you
what the return on investment was for each one
of the channels you used, and how much each
channel contributed to your final results.
Page | 7
MEASURE AND OPTIMIZE SOCIAL MEDIA MARKETING CAMPAIGNS
Facebook | Anke Audenaert
 Based on this insight, you then adjust
your budget proposal for next year. You
might suggest putting a smaller portion
of the budget into TV and a larger chunk
into ads on Facebook.
Remember:
 Because media mix modeling uses a
variety of datasets from different
channels where information may not all
be immediately available in contrast to
online media, the information collection
can be long and complex.
 Moreover, the modeling approach is
typically iterative and building a model
takes time. As a result, marketing mix
modeling is a method that takes a long
time to execute.
3.5 Measuring Across Channels with
Attribution
As Kathy moved through her customer journey.
She met many different ads, and most likely all
these ads had some effect on Kathy's decision.
Attribution is the process of assigning credit
to touch points along a customer’s
conversion path.
 This insight can help you understand
whether you invested in the right
advertising channels and by
understanding which of the channels
worked best.
 Attribution is mostly used to evaluate the
impact of online media channels on a
purchase.
 Of course, you can also use it across
online and offline channels like TV, for
instance, but since the data is not
always easy to get by, the result from
this type of cross channel analysis is not
as immediate.
With Attribution modeling, you will decide on
a model in which you're going to credit the
different marketing channels before you start
analyzing.
 You will think about a customer journey
and make some assumptions about
which parts of your marketing you
believe have the biggest effect on the
final sale.
Elements of Attribution
 Data. Information about the ads a
person saw and whether they clicked on
them or took action.
 Attribution model. A model that is used
to assign value to the different
touchpoints. The output is a value for
each touchpoint that reflects the impact
it had on the conversion.
The accuracy of Attribution results depends
on the quality of the data input, but also the
accuracy of the model.
Most Attribution analysis will use what we call
rules-based models. In rules-based models
you select the rule that determines how
conversions should be allocated to add
touch points.
Last Click Attribution Model
A very commonly used model is the last click
Attribution model. It gives 100% credit to the
last click from a person before a conversion took
place.
 Even if a person saw and clicked on
many ads before that. To model still
assumes that the last ad the person
clicked on is solely responsible for the
sale.
 While this model is popular, marketers
agree that it's not a very accurate
representation of reality.
For most conversions, it takes more than one
touch point for people to convert, hence models
that give credit to more than one touch point are
considered to be more accurate.
Last Touch Model
Last touch models like last click models also
consider only one touch point, but in this
case there's no need for a click. This model
gives 100% credit. The last ad a person
interacted with before a conversion, whether the
person just saw the add or clicked on it.
Even Credit
In even credit models, equal credit is given to
the touchpoints, impressions, and clicks that led
to a conversion.
Time Decay
The time decay model gives credit to
touchpoints, clicks, or impressions based on
how recently they happened. This model gives
more credit to touch points that happened more
recently, so closer to the conversion.
Positional Model
In this model, first and last touchpoints, whether
they're clicks or impressions, are given more
credit. While the credit for the remaining
touchpoints in the middle is divided equally.
First Click
The first click Attribution model, which gives
100% credit for a conversion to the first click on
the conversion path.
First Touch
It gives 100% credit for a conversion to the first
click that happens in the conversion path, but if
there is no click then it will credit to conversion to
the first impression.
Statistical or Math-based Models
Page | 8
MEASURE AND OPTIMIZE SOCIAL MEDIA MARKETING CAMPAIGNS
Facebook | Anke Audenaert
In statistical or math-based models, you let the
data speak and you don't decide on a model
ahead of time.
 Statistical models use all the available
data to determine which ads impact
results.
 This data driven attribution model then
assigns fractional credit for a conversion
to touch points based on their estimated
incremental impact.
 Instead of using your intuition to choose
a rule, a statistical model uses the
available data to figure out which
touchpoints have most influence.
3.6 Challenges When Measuring Across
Channels
Measure across platforms and devices
For instance, if you sell a product on a mobile
app on the Web and in your store, you want to
make sure you have a way to capture all the
sales.
If you're trying to evaluate the effect of your
advertising on all sales, you'll have to make sure
that all sales are indeed included, which also
means including the offline sales.
Connecting touchpoints to one person
When a person visits your website and visits
your App along their journey to purchase. You'd
want to understand that that was the same
person, without that insight.
It's difficult to understand the contribution of
different touch points to your results. Historically,
advertisers have relied on cookies to connect a
sequence of touch points to a person.
With that same person’s interactions on
computers, platforms like Facebook leverage, a
persistent log in and a unique identifier
associated with a person's Facebook account,
when a person sees Facebook ads on mobile or
on the Web.
 Facebook can connect the dots,
together with the Facebook pixel, the
SDK, and the conversions API.
 This approach allows you to tie actions
directly to real people, using this
approach.
WEEK 4: COMMUNICATE YOUR MARKETING
RESULTS
4.1 Presenting Your Campaign Results
Your presentation should:
 Tell a story and give context. Create a
presentation that's going to tell the story
of your campaign from start to finish, not
just the results, because they would be
meaningless outside of the context of
the campaign.
 Show what the data tell you. You want
to tell the narrative of what you did and
show a through line from goals, to
campaign, to results, to implications for
the future.
 State conclusions and insights.
Explain how you got those conclusions
and show how those conclusions will
affect decisions going forward. Ten
slides of data to senior leadership won't
be helpful, but 10 slides telling them how
your actions affected their bottom line
and increased sales will.
Your story:
Think of your presentation as following a story,
or as loosely following the scientific method,
where you come up with a question or a goal
around a set of actions, you execute the actions,
gather data and metrics around the actions, and
then create new questions and goals.
 In our story, you start with a Smart Goal.
You create a Campaign, you evaluate
results of the campaign, and you
formulate Insights. These insights can
be the inspiration for a new campaign to
achieve a new goal.
Campaign Setup
First, you always want to start with the goals you
determined for the campaign, as this will set up
the premise of the story moving forward.
 Starting with your goals and why you set
them, will give you a baseline to build
from as you tell your campaigns story.
Next, explain how those goals turned into
campaign objectives, which will move the story
forward towards what actions you took in your
campaigns.
 If you were looking to increase brand
awareness, you planned for an
awareness campaign. If you wanted to
gain 25 percent more in sales, you
planned for a conversion campaign.
Additionally, now would be the time to speak
about the KPIs you are going to track to
measure the success of your campaign.
 For brand awareness, you'll track ad
recall lift. For sales, you'll track specific
conversion events. For more visitors,
you'll track link clicks, landing page
views, or reach.
Campaign Specifics
1. Show the creative you used
2. An overview of the characteristics of
the audience you delivered it to,
3. Which platforms it ran,
4. How long it lasted (duration)
5. How much you spent on it (budget)
Campaign Results
Page | 9
MEASURE AND OPTIMIZE SOCIAL MEDIA MARKETING CAMPAIGNS
Facebook | Anke Audenaert
 Data sources. It's always useful to go
over where you gathered your data
from. Ads Manager, Google Analytics,
etc. If you ran any experiments like A/B
or hold out tests for instance, or extra
analysis like an attribution study, you
would want to let them know that too.
 Metrics overview.
1. Results (KPIs)
2. Costs
3. ROAS (and ROI)
4. Effectiveness
 Insights.
1. Evaluate your goal. Did I reach
my goal? Did I do so for a
reasonable cost?
2. Breakdown by segment. Ask
yourself whether there's any
audience or any part of your
campaign that seems to stand
out in terms of results. Maybe
you did not reach your overall
goal, but you reached it among
a certain group, maybe a certain
ad worked better with some
people than with others.
3. Apply insights to next steps.
Example:
Imagine you find that you didn't meet your brand
awareness goal, you can show how while your
awareness campaign had a large reach, recall
was low in general, but when you break down
your results a bit more, you find that for one
demographic, recall was a lot higher.
Now you can find out why the brand resonated
with them and not with others and whether that
segment could be your new target audience.
When you tell your data story or convey your
insights, it's important that you look beyond
the average data points and try to assess
different segments or breakdowns of your
data.
4.2 Telling a Clear Story with Data
With most database or data analysis software,
you will get options to summarize your data in a
chart.
 A bar or a column chart, is a great way
to represent data that falls into different
categories, like preferences. For
instance, you'd use a bar chart to show
the number of people that like Version A
of an ad versus the number of people
that like version B of an ad. Bar charts
also make comparisons easy, like
comparing data for different years or
comparing data against the goal.
 A pie chart is a great way to show how
a whole is divided into parts. It could be
a great way to show demographic
distributions. For instance, if you have to
break down an audience by age, the
whole pie is your complete audience,
and different portions represent different
age groups.
 A line graph works well if you want to
show how numbers change over time, it
also makes it easy to compare trends,
too.
Other tips:
1. Don’t clutter the chart. As you're
adding your charts and graphs, you
should first think about how you can tell
the data story visually. In other words,
you may not need numbers and labels
to communicate what you want to say.
In fact, they might just get in the way.
2. Use color to help the audience focus
on what matters. Your color choice will
help communicate your story as well. As
you put together your visualizations,
think about using gradients of the same
color, a branded collar palette or even
color against grayscale to highlight
different metrics.
3. Make comparisons easy. It's a good
idea to make comparisons between two
data points so the audience can easily
see what's going on.
4. Order your data when possible. Also,
make the data easier to read by ordering
it, either by size or incrementally, so that
it makes sense to look at.
5. Pay attention to the scale. Make sure
each point on your access is evenly
spaced out, and with the right increment
like month to month, or year to year.
6. Call out important points. Let the data
tell the story, but don't force the data to
fit the story you would like to see. Not all
of your data has to be in a chart or a
graph. If you want to highlight specific
data points, you can do that with a call
out.
4.3 Communicating in a Presentation
 Think through who your audience is,
and what you want them to take away
from your presentation. Think about
what data you want to present to them,
how to tell your data story, and how to
visualize it in a unique way.
 Write out your talking points and
create your presentation in Google
Slides or PowerPoint or other
Page | 10
MEASURE AND OPTIMIZE SOCIAL MEDIA MARKETING CAMPAIGNS
Facebook | Anke Audenaert

presentation software. Essentially, you
want your slides to simply highlight what
you're speaking about. Not to carry the
weight of the presentation itself. You will
be the storyteller and your slides are
there to support you, and to stress the
main points of your story.
 Use only keywords or short bullet
points on your slides to do a simple
visualization of what you're saying.
 Keep your font size to 28 or bigger.
Anything that's smaller can get hard to
read on a slide, and that way you will
stick to the main points.
 Declutter them and use them to
highlight what you're saying or else your
audience will be distracted trying to
decipher your graphs.
 Don't read from your slides. Use the
presenter notes feature to see your
notes during a presentation, or have
summaries or bullet points on note
cards, or get to the point where you
know what you want to say well enough
to let your slides be a memory tool for
you.
 Catchy opening. A rhetorical question
related to your findings, or a story or an
interesting fact. Grab the audience
straight away and start them on the
journey with you. Don't spend too much
time on ramping up or introductions to
the topic as your audience may drift off.
 Use a story framework. Start with a
problem you were trying to solve or what
goal you want to hit. Then tell what
actions you took. Explain what you
found after you took those actions and
finish up telling how those findings will
impact future decisions.
 Make your conclusion as impactful as
your opening. In fact, circle back to
what you mentioned in your opening by
either answering the questions, finishing
the story, or adding an interesting fact to
presentation showed.
 Always tell your audience, you'll take
questions at the end and then do so.
Nothing derails a presentation faster
than stopping to answer questions in the
middle that will probably be explained by
your story.
grip the podium, but also don't wander
and pace as you talk or talk with your
hands, which may distract your
audience. Some public speakers build
deliberate movements like a step
forward into their presentation to
highlight a point.
 Slow down and breathe. Allow for
pauses to not only let the information
sink in, but to calm yourself as you go.
Slowing down and being deliberate in
what you can say also eliminates
distracting mmhs and uhs as well.
 Eye contact. Having a friend in the
audience or someone who is attentive
and nodding along, it's always helpful to
concentrate on, but pick a few people in
different spots of the room and look
back and forth between them or look just
over their heads.
 Watch the clock. Always have clock
inside so that you can adjust as needed,
and so that you can leave time for Q and
A.
 Tech and set-up issues.
 Set the room beforehand. Chat with
someone who's familiar with the room
where you will be presenting, so you can
prepare yourself and then walk through
the presentation situation in your head
and use this checklist.
Once your presentation is complete, offer to
send out a slide deck for anyone who wants to
know more. Then plan to treat yourself to
something nice for your accomplishment.

Mitigate Your Nerves

 Anchor. It might be pausing to take a
deep breath or deliberately taking a sip
of water before you speak or seeking
out a friend in the audience.
 Practice run time. You'll gain familiarity
and confidence in what you'll say the
day of. You can also practice intonation
in various spots when you want to hit
your point home or see if anything just
doesn't work or it's too confusing when
spoken aloud.
 Practice movements. If you're giving
your presentation while standing, don't

Page | 11