Brand Effect of TrueView vs.

Preroll
Meta analysis of 18 ad effectiveness experiments

Hamburg 2012

1
Management Summary

1 The TrueView format allows users to skip ads. Advertisers only pay
per completed view based on an auction system (preroll: fixed CPM).

2 Overall brand impact of TrueView is lower than for preroll because

78% of views are skipped by the users.

3 Skipped views (= unpaid impressions) however still have a

significant positive impact on brand metrics

4 This leads to TrueView being 2.8 times more effective than Preroll
regarding brand uplift per paid contact.

5 Up to a bid of 0.04 € per completed view TrueView is more efficient

than Preroll. After that Preroll becomes the more efficient format.

Google Confidential and Proprietary 2

How TrueView works
In contrast to Standard Preroll, TrueView gives users the choice to skip the ad after 5
sec. TrueView is bought via a CPV auction. Only completed views are charged.

Online video ad formats on YouTube

PREROLL TRUEVIEW
1
2
3
4
5

Google Confidential and Proprietary 3

Methodology
Brand impact of online video was analysed using randomized experiments.
Each test campaign was displayed for 10 days within a panel of n = 10.000

Experimental study design

PRE-ROLL TRUEVIEW

Each panelist is randomly assigned

to either see Prerolls or TrueView.

By suppressing the campaign for 50% of

TEST CONTROL TEST CONTROL
exposed unexposed exposed unexposed
the sample, control cells are generated.

This allows a direct comparison of KPI

BRAND METRICS SURVEY uplift between formats to understand
compare test vs. control & pre-roll vs. TrueView which format works best for the brand.

Google Confidential and Proprietary 4

The test campaigns

•  Test campaigns were served exclusively to the participants of

the GfK “connected life” panel

•  Campaigns were served with a maximum frequency to each

panelist

•  Campaign duration was 10 days

•  Surveys took place within 4 days after the campaign

Google Confidential and Proprietary 5

Consumer Insights from skipping
 On average 78% of TrueView ads are skipped
Skip rate remains stable across industries – no matter the product

Skip rate per vertical (base: TrueView ads served)

100%
90%
78% 78% 83% 80% 78%
80%
75% 76% 75%
70%
60%
50%
40%
30%
20%
10%
0%
Total Average Shampoo + Social Networks Travel Electronics + Automotive + Retail + Finance +
Detergent (e.g. G+) Companies Telecom Car Supplies Fashion Power Supplier

Source: GfK (randomised experiment with 18 brands and n = 3.184 respondents)

Connected Life Panel Germany 2012 (n = 10.000 panelists)
Google Confidential and Proprietary 7
People skip if they already know the ad

Reasons for skipping ads (stated)

65%
are skipping the video ad because
they have already seen it

Google Confidential and Proprietary 8

The first view is more likely to be completed

Skipping behaviour (subsample with up to 4 TrueView impressions)

skipped
all ads
58%
watched
(at least)
32%
1st ad

10%
skipped 1st,
watched later

Source: GfK (randomized experiment with 18 brands and n = 3.184 respondents)

Connected Life Panel Germany 2012 (n = 10.000 panelists)
Google Confidential and Proprietary 9
Brand impact of TrueView
 Impact of TrueView vs. Preroll
Based on all served impressions preroll ads have a greater impact than
TrueView ads.

(Headroom) Uplift on top of mind awareness

Preroll TrueView

15%
14%
13%
12% 11%
11%
10%
9%
8%
7%
7%
6%
5%
4%
3%
2%
1%
0%

Uplift on TOM Awareness

(all views)

Source: GfK (randomised experiment with 18 brands and n = 3.184 respondents)

Connected Life Panel Germany 2012 (n = 10.000 panelists)
Google Confidential and Proprietary 11
 Impact of TrueView vs. Preroll
Given the 78% skip rate for TrueView the average frequency from paid
contacts is much lower than for preroll.

Brand Impact vs. average frequency of paid ads

Preroll TrueView Preroll TrueView

15% 7.00
14% 5.88
13% 6.00
12% 11%
11% 5.00
10%
9% 4.00
8%
7%
7%
6% 3.00
5%
4% 2.00
1.21
3%
2% 1.00
1%
0% 0.00

Uplift on TOM Awareness Average frequency

(all views) from PAID views

Source: GfK (randomised experiment with 18 brands and n = 3.184 respondents)

Connected Life Panel Germany 2012 (n = 10.000 panelists)
Google Confidential and Proprietary 12
 Impact of TrueView vs. Preroll
Based on all paid impressions TrueView ads are 2.8x more effective than
pre-rolls.

(Headroom) Uplift on top of mind awareness

Preroll TrueView Preroll TrueView Preroll TrueView

15% 7.00 15%

14% 5.88 14%
13% 6.00 13%
12% 11% 12%
11% 5.00 11%
10% 10%
9% 4.00 9%
8% 7% 8%
7%
6% 3.00
7%
6%
5%
5% 5%
4% 2.00
1.21 4%
2%
3% 3%
2% 1.00 2%
1% 1%
0% 0.00 0%

Uplift on TOM Awareness Average frequency Uplift on TOM Awareness

(all views) from PAID views (per paid view)

Source: GfK (randomised experiment with 18 brands and n = 3.184 respondents)

Connected Life Panel Germany 2012 (n = 10.000 panelists)
Google Confidential and Proprietary 13
 TrueView impact: free vs. paid impressions
Total TrueView impact is at 7% uplift. A significant impact share comes from
skipped views which are not paid for by the advertiser.

(Headroom) Uplift on top of mind awareness

TrueView total (all views) TrueView skipped views TrueView completed views
10%
9%
8%
8%
7%
7%
6% 5%
5%
4%
3%
2%
1%
0%

Uplift on
T.O.M. Awareness

Source: GfK (randomised experiment with 18 brands and n = 3.184 respondents)

Connected Life Panel Germany 2012 (n = 10.000 panelists)
Google Confidential and Proprietary 14
Efficiency comparison
 TrueView efficiency analysis
Preroll CPM is at 15 € (= CPV of 0.015 €). TrueView has a 2.8 times higher
impact, thus becomes equally efficient to Preroll at an auction price of 0.04 €.

Preroll CPM vs. TrueView CPV at equal efficiency

Preroll TrueView Preroll (15 € CPM)

TrueView (auctionized CPV)
10% 0.08 €
9% 0.07 €
8%
0.06 €
7%
6% 2.8x 0.05 € 2.8x
5%
4% 5% = 0.04 €

0.03 €
0.042 €
3%
0.02 €
2%
1% 2% 0.01 €
0.015 €
0% 0.00 €

Uplift on TOM Awareness Cost per view at which

(per paid view) TrueView is equally efficient to preroll

Source: GfK (randomised experiment with 18 brands and n = 3.184 respondents)

Connected Life Panel Germany 2012 (n = 10.000 panelists)
Google Confidential and Proprietary 16
Two efficiency scenarios

From an advertiser’s perspective there are two possible

scenarios:
•  maximise campaign efficiency at given budget
“Up to which auction bid do I get greater brand impact from TrueView than
from preroll in return for my budget?”

•  reach the maximum possible brand impact at flexible budget

(→ assumption: standard preroll is the reference point for maximum impact)
“How much budget (relative to the preroll scenario) will I need for TrueView
to maximize campaign impact?”

Google Confidential and Proprietary 17

 Scenario 1: fixed budget
TrueView delivers greater impact than preroll up to 0.04 € CPV. TrueView is less
efficient if the auction rises to 0.05 € or more.

TrueView impact index to preroll at equal levels of media budget

IMPACT ON TOP OF MIND AWARENESS (indexed to preroll)

450% 428%

400%

350%

300%

250% 214%
200%
143%
150%
100% 107%
86% 71%
100% 61% 54% 48% 43% 39% 36% 33% 31% 29%
50%

0%
Preroll 0.01 € 0.02 € 0.03 € 0.04 € 0.05 € 0.06 € 0.07 € 0.08 € 0.09 € 0.10 € 0.11 € 0.12 € 0.13 € 0.14 € 0.15 €

TRUEVIEW IMPACT INDEX AT … AUCTION BID

Source: GfK (randomised experiment with 18 brands and n = 3.184 respondents)

Connected Life Panel Germany 2012 (n = 10.000 panelists)
Google Confidential and Proprietary 18
 Scenario 2: maximum impact
TrueView delivers the same impact as preroll at lower budget up to 0.04 € CPV. If the
auction rises to 0.05+ € a higher budget for TrueView than for preroll is required for the
same level of brand impact.

TrueView budget index to preroll at equal level of brand impact

TRUEVIEW BUDGET TO IMPACT TOP OF MIND AWARENESS (indexed to preroll)

400%
347%
350% 324%
301%
300% 278%
255%
250% 232%
208%
200%
185%
162%
139%
150% 116%
100% 93%
100% 69%
46%
50% 23%
0%
Preroll 0.01 € 0.02 € 0.03 € 0.04 € 0.05 € 0.06 € 0.07 € 0.08 € 0.09 € 0.10 € 0.11 € 0.12 € 0.13 € 0.14 € 0.15 €

TRUEVIEW TOTAL BUDGET INDEX AT … AUCTION BID

Source: GfK (randomised experiment with 18 brands and n = 3.184 respondents)

Connected Life Panel Germany 2012 (n = 10.000 panelists)
Google Confidential and Proprietary 19
thank you!
Google Confidential and Proprietary 20