3. Kanui
Comparable cost per install and
2.8x
better ROI
with placement optimization
(vs. Facebook and Instagram
separately)
3
4. Teachers (TV Land)
16%
increase in
impressions
by adding Instagram as a placement
21%
increase in
brand awareness
across Facebook and Instagram
4
2,200 likes
teachersseries Reading, writing, a righteous
hangover. Catch up on Teachers on demand or
with the app!
view all 711 comments
teachersseries
5. Additional case studies
Advertiser Country Vertical Objective
Placement
s
Results GMS Hub Link
TV Land US Entertainme
nt
Brand
Awareness
Facebook
Instagram
• 16% increase in
impressions
• 24% increase in ad
recall
• 21% increase in
brand awareness
https://our.intern.facebook.com/intern/ma
rketingcontent/gmshubv2/content/?id=10
93786793998298
World Golf
Tour
US Gaming Mobile App
Installs
Facebook
Instagram
• 44% lower CPI https://our.intern.facebook.com/intern/ma
rketingcontent/gmshubv2/content/?id=18
2732855452510
TM Lewin UK eCommerce Website
Conversions
Facebook
Instagram
• 9x decrease in cost
per acquisition
https://our.intern.facebook.com/intern/ma
rketingcontent/gmshubv2/content/?id=94
3187659129003
MVMT US eCommerce Website
Conversions
Facebook
Instagram
• 3x lower cost per
acquisition
https://our.intern.facebook.com/intern/ma
rketingcontent/gmshubv2/content/?id=45
1317691691464
5
7. 7
Breaking down the efficiency of running across the family of apps
Outcomes
Price per Outcome
Bid needed to deliver
campaign
Average cost per
outcome by placement
Bid
Price per Outcome
Outcomes
$3$2 $4
Average cost per outcome per
placement
$4.5
0
8. 8
Outcomes
Price per Outcome
Bid needed to deliver
campaign
Average cost per
outcome by placement
Bid
Price per Outcome
Outcomes
$3$2 $4
Average cost per outcome per
placement
$4.5
0
Instagram has the
highest cost per
outcome. Should
we remove it?
Breaking down the efficiency of running across the family of apps
9. 9
Let’s see what happens…
Outcomes
Price per Outcome
Average cost per
outcome by placement
Bid
Price per Outcome
Outcomes
$3$2 $4.5
0
$6$3.50
We had to increase our
bid to drive the same
amount of outcomesWe lost cheaper
Instagram outcomes
We had to buy more
expensive Facebook
outcomes
1
3
2
4 Our average price per
outcome on Facebook
rose
Editor's Notes
[PUBLIC]
MVMT [pronounced ”movement”] (eCommerce, US) watches saw a 20% lower cost per click (CPC) and cost per action (CPA) when opting in to placement optimization across Facebook and Instagram (vs. running on the two platforms separately).
Client Quote: “MVMT Watches has been extremely happy with not only Instagram ad performance in general, but with how easy it was to build Instagram into our overall strategy by simply allowing us to buy Instagram as another placement, instead of having to manage another separate platform.”
[PUBLIC]
UGO (part of TD Bank, Canada) saw a 45% lower cost per install by moving to performance optimization across Facebook and Instagram for their mobile app install campaigns (vs. Facebook mobile news feed-only campaigns).
[PUBLIC]
Kanui (ecommerce, Brazil) ran a mobile app install campaign initially targeting men 18+ on Facebook.
When they added Instagram to their Facebook ad set, they saw a similar cost per install and generated a 2.8x higher return on ad spend.
“We were so pleased by the outcome of our ads on Instagram and Facebook that we’re changing all of our ad sets to run across both platforms going forward,” says Guilherme Becker, Performance Director at Kanui.
[PUBLIC]
[INTERNAL ONLY]
You may have questions about creative.
In particular, are you required to run the same creative across every placement?
Since Facebook creative automatically extends to Audience Network, we’ll focus on Instagram today.
Many times, running an ad set across Facebook, Instagram and the Audience Network will result in reporting that varies by placement. In this example, Audience Network provides the least amount of outcomes, but produces the cheapeast cost per outcome ($2). Facebook provides the most outcomes and the median cost per outcome ($3). Instagram provides more outcomes than the Audience Network, but is the most expensive on a cost per outcome basis ($4). For this campaign, it was determined we needed to bid a maximum of $4.50 in order to pace the campaign and drive the desired number of outcomes.
Many times, running an ad set across Facebook, Instagram and the Audience Network will result in reporting that varies by placement. In this example, Audience Network provides the least amount of outcomes, but produces the cheapeast cost per outcome ($2). Facebook provides the most outcomes and the median cost per outcome ($3). Instagram provides more outcomes than the Audience Network, but is the most expensive on a cost per outcome basis ($4). For this campaign, it was determined we needed to bid a maximum of $4.50 in order to pace the campaign and drive the desired number of outcomes.
When you remove Instagram from the ad set, you remove all of the outcomes Instagram could have driven. Our auction now has to find outcomes to replace all the outcomes we lost when we removed Instagram. In order to do that, we have to look to the more expensive end of the Facebook curve.
We have to increase the bid to $6 in order to drive the same amount of outcomes we drove before. We have to increase the bid because we are replacing chearper Instagram outcomes with more expensive Facebook outcomes.
We lost the cheaper outcomes we could have driven with Instagram
To make up for the lost Instagram outcomes, we have to buy more expensive outcomes on Facebook.
The average price per outcome rises to $3.50 on Facebook after we add the more expensive Facebook outcomes we bought over the life of the campaign.
After removing Instagram, our cost per outcome on AN remains the same and rises on FB. The new price per outcome average on FB is not as high as IG’s was prior to removing. Was this a successful optimization? No. You traded cheaper outcomes for more expensive outcomes and made the entire campaign more expensive.