Upload
scott-smith
View
103
Download
0
Tags:
Embed Size (px)
Citation preview
MMoobbiillee VViiddeeoo BBeenncchhmmaarrkk SSttuuddyy��2233 AApprriill 22001144 ��
M�A�M�
AAcckknnoowwlleeddggeemmeennttss��
�� MMA Member Data Contributors: �
�� Data Aggregation and Normalization – ImServices ��� Analysis – Gerard Broussard, Pre-Meditated Media �
3�
��
��
TTaabbllee ooff CCoonntteennttss��
�� Study Set-up and Overview: � � � � Slide 4-8 ��� Study Findings: � � � � � � � Slide 9 ��� Industry Benchmarks: � � � � � �Slide 10-11 ��� Ad Unit Format: � � � � � � �Slide 12-14 ��� Skippable Ad Drop-off: � � � � �Slide 15 ��� Ad Length: � � � � � � � �Slide 16-18 ��� Hour of Day: � � � � � � � �Slide 19-22 ��� Ad Frequency: � � � � � � �Slide 23-24 ��� Appendix: � � � � � � � � �Slide 25 �
4�
SSttuuddyy OOvveerrvviieeww��
Background/Purpose��� Mobile Marketing Association (MMA) sought to provide mobile video
performance insights to advertisers, agencies, publishers and 3rd party video servers�
�� MMA secured participation from six mobile publishers to provide one month’s (March, 2013) worth of mobile advertising data to report video performance benchmarks �
5�
Performance Measures -- the analysis focused on three key measures: �
1.� Completion Rates�2.� Click Rates (CTR)�3.� Non-Skippable vs. Skippable Ads �
Evaluation Criteria –– the three performance themes were evaluated through the lens of: �
1.� Ad Unit Length, Ad Format, Site Genre�
2.� Dayparts/Time of Day�3.� In-App vs. Mobile Web�4.� iOS vs. Android�5.� Phone vs. Tablet�
FFiivvee KKeeyy FFiinnddiinnggss��
��1.� #>18;4�,834>�03�4=6064<4=C�<4CA82B�?A>E834�JABC�<40=8=65D;�
benchmarks for the industry ��2.� Engagement for skippable video advertising is robust despite the
option for viewers to bypass ads �
3.� #>18;4�E834>�2><?;4C8>=�A0C4B�F4A4�K0C�02A>BB�03�;4=6C7��7>F4E4A��CTRs dropped for ads greater than 30 seconds �
4.� A high-engagement, late night window exists among a small group of consumers �
5.� Mobile video ad performance, like other forms of advertising, declines with excessive ad frequency �
6�
OOtthheerr FFiinnddiinnggss ooff NNoottee��
1.� In-app ad volume dominated; three in-app units ran for every one on the mobile web �
�2.� Linear video ads ran at two times the volume of Interstitial and Value
Exchange combined �
3.� Arts & Entertainment, Games, Hobbies and Sports represented the E0BC�<09>A8CH�>5�03�02C8E8CH��A4K42C8=6�2>=BD<4A�<>18;4�E834>�DB4��
�
7�
555599 MMiilllliioonn AAdd IImmpprreessssiioonnss ��SShhaarree ooff AAdd IImmpprreessssiioonn VVoolluummee��
8�
#4�%%���$#��=?E�
�$��!�������$#��@E� #4�%%����!�)�
<9E�
�$��!��������!�)�;E� �# #$.#�
:=E���%��� �
$#4( �%%��!��?AE�
� �%%��!��<;E�
���- ��������,�����������
����A;E�
�#�'$���:AE�
�����"����' !���
More than 75% of all advertising occurred in-app, the majority on phones . . .�
iOS activity accounted for more than 80% of ad volume . . . �
Skippable ads comprised a third of all ad impressions . . . �
555599 MMiilllliioonn AAdd IImmpprreessssiioonnss ��SShhaarree ooff AAdd IImmpprreessssiioonn VVoolluummee��
9�
��#��'�����$�?>E�
#)�'(**�!�:@E�
��!,���/���#���:>E� �# #$.#�
<E�
��������!�
3:>�$'�!�((�>?E�
3:?43<9(�<<E�
3<:F�AE��# #$.#�
<E�
������!��
�')(�6��#)�')��#"�#)�=@E�
��"�(�;=E�
�$����(�6�#)�'�()�:=E�
�%$')(�@E�
�)��'�BE���!��������
Linear video was the predominant ad form followed by a split of Interstitial and Value Ex . . . �
More than half of ads were :15 or less . . . �
�>DA�3><8=0=C�64=A4�A4K42C8E4�>5�2>=BD<4A�E834>�02C8E8CH�8=�<>18;4�B?024�����
SSTTUUDDYY FFIINNDDIINNGGSS��
:�
11.. MMoobbiillee VViiddeeoo aadd eennggaaggeemmeenntt mmeettrriiccss pprroovviiddee JJrrsstt mmeeaanniinnggffuull bbeenncchhmmaarrkkss ffoorr tthhee iinndduussttrryy����
21�
CTR Benchmarks �Mobile Video Aggregated Findings �
11�
�� CTR rates for all ad formats serve as an initial proxy for ad engagement �
3+41=�
2+61=�
3+32=�
2+52=�
3+77=�
1+11=�
1+61=�
2+11=�
2+61=�
3+11=�
3+61=�
4+11=�
�����-����������������
������
������������������������
�����-����������
��!�� ""���������
��������������!�� ""���������
�����$���&�������������
��$���*��������������������������!$�')�3124�
22.. EEnnggaaggeemmeenntt ffoorr sskkiippppaabbllee vviiddeeoo aaddvveerrttiissiinngg iiss rroobbuusstt ddeessppiittee vviieewweerrss’’ ooppttiioonn ttoo bbyyppaassss aaddss ��
23�
AAdd UUnniitt FFoorrmmaatt��CCoommpplleettiioonn RRaattee –– PPhhoonneess//TTaabblleettss��
:4=� :3=�:8=�
9=�
34=�
1=�1=�
21=�
31=�
41=�
51=�
61=�
71=�
81=�
91=�
:1=�
211=�
������������ ��!�� ""��� ���$�.�&����������-���������� ����������
24�
�� Skippable completion rates are dramatically lower than non-skippable . . . �
Number of Impressions: �Non-Skippable 194.9mm � � 53.5mm 59.1mm � Skippable 142.9mm � 16.9mm N/A �
�$,'��3�� ��5-=5;9:<:;5��������5,%��)���
:=� 36=�
AAdd UUnniitt FFoorrmmaatt��CCTTRR -- PPhhoonneess//TTaabblleettss��
3+5=�3+3=�
3+5=�
2+6=� 2+5=�
1+1=�1=�
2=�
2=�
3=�
3=�
4=�
4=�
������������ ��!�� ""��� ���$���&�������
���-���������� ����������
25�
�� . . . however skippable CTR rates don’t follow too far behind �non-skippable�
Number of Impressions: �Non-Skippable 45.5mm � � 53.5mm 59.1mm � Skippable 142.9mm � 16.9mm N/A �
�$,'��3�� ��5-=5;9:<:;5��������5���5� ��5,%��)���
73=� 75=�
Implication: ads were more relevant for those who decided not to skip �
SSkkiippppaabbllee AAddss��OOvveerraallll CCoommpplleettiioonn RRaattee** ���� Ad skipping is a process that unfolds in stages of time �
–� More than half of skippers stick around until 25% of the video has played �–� Around 30% remain through a video ad’s halfway mark �
�$,'��3�� ��5-<5�������5����������
56%�
27%�
21%�18%�
63%�
32%�29%� 27%�
50%�
31%�26%� 24%�
0%�
10%�
20%�
30%�
40%�
50%�
60%�
70%�
25% of Ad Time� 50% of Ad Time� 75% of Ad Time� 100% of Ad Time�
iOS Phone� iOS Tablet� Android Phone�
* Includes only ads where quartile and skippable data were available�
Completion Rate by Quartile �
26�
33.. MMoobbiillee vviiddeeoo aadd ccoommpplleettiioonn rraatteess wweerree KKaatt bbyy aadd lleennggtthh,, hhoowweevveerr,, CCTTRRss ddrrooppppeedd ffoorr aaddss ggrreeaatteerr tthhaann 3300 sseeccoonnddss��
27�
AAdd LLeennggtthh ��CCoommpplleettiioonn RRaattee –– PPhhoonnee//TTaabblleett��
:3=� :5=� :7=�
:=�23=� 25=�
1=�
21=�
31=�
41=�
51=�
61=�
71=�
81=�
91=�
:1=�
211=�
� �!����*26� *27-*41� *42>����-���������� ����������
28�
�� Virtually no difference in non-skippable completion rates for ads of different length ��� Completion rates of skippable ranged between 9% and 14%�
Number of Impressions: �Non-Skippable 146.1mm � � � 120.1mm 40.1mm � Skippable 122.3mm � � � 33.9mm � � � � 2.6mm �
�$,'��3�� ��5-=5;9:<:;5��������5,%��)���
AAdd LLeennggtthh��CCTTRR,, PPhhoonnee//TTaabblleett��
3+3=�
3+6=�
2+4=�2+5=�
2+9=�
1+9=�
1=�
2=�
2=�
3=�
3=�
4=�
4=�
� �!����*26� *27-*41� *42>����-���������� ����������
29�
�� Click rates for ads longer than 30 seconds tended to be lower than shorter-length units ��
�$,'��3�� ��5-=5;9:<:;5��������5���5� ��5,%��)���
Number of Impressions: �Non-Skippable 87.9mm � � � 66.9mm 2.3mm � Skippable 122.3mm � � � 33.9mm � � � � 2.6mm �
44.. AA hhiigghh--eennggaaggeemmeenntt,, llaattee nniigghhtt wwiinnddooww eexxiissttss aammoonngg aa ssmmaallll ggrroouupp ooff ccoonnssuummeerrss��
2:�
IImmpprreessssiioonn SSuummmmaarryy ��HHoouurr ooff DDaayy –– DDeevviiccee��
31�
9�
;29992999�
=29992999�
?29992999�
A29992999�
:929992999�
:;29992999�
:=29992999�
:?29992999�
:A29992999�
;929992999�
��$#��
9�
;29992999�
=29992999�
?29992999�
A29992999�
:929992999�
:;29992999�
:=29992999�
:?29992999�
:A29992999�
;929992999�
���!�)�
Number of Impressions: �
237.1mm � 144.2mm �
�� For both phones and tablets, impression volume drops sharply over night then shoots upward during the course of the day�
��'"������ ��������
:�-5 � 9-22 �
��'"������ ��������
:�-5 � 9-22 �
HHoouurr ooff DDaayy��CCoommpplleettiioonn RRaattee -- DDeevviiccee��
1+1=�
21+1=�
31+1=�
41+1=�
51+1=�
61+1=�
71+1=�
81+1=�
91+1=�
:1+1=�
��$#��
32�
1+1=�
21+1=�
31+1=�
41+1=�
51+1=�
61+1=�
71+1=�
81+1=�
91+1=�
:1+1=�
���!�)�
Number of Impressions: �
237.1mm� 144.2mm �
�� �><?;4C8>=�A0C4B�?40:�3DA8=6�;0C4�4E4=8=6�40A;H�<>A=8=6�7>DAB�0=3�0A4�A4;0C8E4;H�K0C�throughout the remainder of the day �
��'"������ ��������
:�-5 � 9-22 �
��'"������ ��������
:�-5 � 9-22 �
HHoouurr ooff DDaayy��CCTTRR bbyy DDeevviiccee ��
33�
1+1=�
1+6=�
2+1=�
2+6=�
3+1=�
3+6=�
4+1=�
��$#��
1+1=�
1+6=�
2+1=�
2+6=�
3+1=�
3+6=�
4+1=�
���!�)�
Number of Impressions: �
169.8mm � 72.5mm �
�� Phone CTR rates are highest during late evening/early morning hours suggesting that engagement is much deeper for people active during this time period �
�� Tablets CTRs did not follow the same pattern as phone �
��'"������ ��������
:�-5 � 9-22 �
��'"������ ��������
:�-5 � 9-22 �
55.. MMoobbiillee vviiddeeoo aadd ppeerrffoorrmmaannccee,, lliikkee ootthheerr ffoorrmmss ooff aaddvveerrttiissiinngg,, ddeecclliinneess wwiitthh eexxcceessssiivvee aadd ffrreeqquueennccyy��
34�
AAdd FFrreeqquueennccyy ��IImmppaacctt oonn CCTTRR**��
35�
�� Excessive ad frequency has clear downward impact on CTR rates �
1+:=�
2+6=�
2+3=�
1+8=�1+7=�
1+6=�
1+1=�
1+3=�
1+5=�
1+7=�
1+9=�
2+1=�
2+3=�
2+5=�
2+7=�
2-6� 7-21� 22-26� 27-31� 32-36� �%���36�
�����0�����'�&,�#�0�
Ad Frequency�
AAPPPPEENNDDIIXX����
36�
DDeeJJnniittiioonnss��11.. PPRREE//MMIIDD//PPOOSSTT--RROOLLLL LLIINNEEAARR VVIIDDEEOO AADD UUNNIITT:: ��
�4J=8C8>=��a mobile video ad that runs before, during or after video content �
22.. IINNTTEERRSSTTIITTIIAALL VVIIDDEEOO AADD UUNNIITT:: ���4J=8C8>=��a mobile video ad that runs between non-video content and occupies a majority of the device screen�
33.. VVAALLUUEE EEXXCCHHAANNGGEE VVIIDDEEOO AADD UUNNIITT::���4J=8C8>=��a mobile video ad that runs on user initiation and is viewed to completion in exchange for a digital reward such as points, goods, coins, coupons or paywall gated content that would otherwise be earned or purchased�
37�
38�
IIAABB TTaaxxoonnoommyy CCaatteeggoorriieess��
39�
IIAABB TTaaxxoonnoommyy CCaatteeggoorriieess��
39�
3:�
NNiieellsseenn TTVV DDaayyppaarrttss��