Ogilvy 360 Digital Influence Conversationimpact 2009
Ogilvy 360 Digital Influence Conversationimpact 2009
Ogilvy 360 Digital Influence Conversationimpact 2009
I n t e l l e c t u a l c a p i ta l f r o m o g i lv y
OCTO B ER 2 0 0 9
Nothing has been more important than developing a credible, workable measurement model for ou
social media-based programs at Ogilvy. Having participated in many of the forums where measurement geeks come together to test-drive their models, it became clear last year that we were all stil
the first third of the process. If you think of the inevitable arc of standards development, the first thir
is floating meaningful measures - no matter how convoluted the model - marketers want to know th
there are valuable metrics that even exist to justify spending in word of mouth-based social media.
into
all
Conversation Impact
Meanwhile, brands that explore the more complex and more meaningful measurement models inev
table backslide into KPIs. You can hear it now from your most senior brand marketers, "Can't we jus
reduce all this complex modeling to 2-3 numbers we care about most? Numbers that reflect how
much reach or how much time people are spending with the brand. What if we just count video view
and Tweets?"
This remains a glimmer in my eye. Once we have the solid and simple modeling out of the second
third, we need to ramp up quickly for the inevitable: complexity. The real power of social media mar
keting is integrated into all sorts of communications and marketing - ALL sorts. What is the impact
social media mentions on the performance of adjacent advertising - does it make the ads perform 3
O g i lv y I n s i g h t 1
Track the
Few Metrics
2 O g i lv y I n s i g h t
Measure
Cross Channel
O g i lv y I n s i g h t 3
Provide
Actionable Measurement
that focuses on key goals
There is rich data now available on a continuous basis, including daily data from the semantic
analysis of the millions of conversations in social media. We integrate these analytics into the
model to help drive optimization.
As an example, for a brand positioning campaign, we can evaluate preference and action in ways
that help us understand which social media influencers are adopting which types of messaging
and in which channels. We look at what people voluntarily say and do across the social Web. We
can use this data to help guide ongoing creative and spending decisions.2
97.2%
94.9%
n Adults 18+
93.6%
91.5%
4 O g i lv y I n s i g h t
We organized our model directly around three main goal categories that build off years of
marketing/communications funnel research into the best way to drive action:
Awareness
Consideration
Preference
Action
Loyalty
Awareness
Consideration
Preference
Action
Loyalty
Awareness
Consideration
Preference
Action
Loyalty
Preference
Sentiment index of online conversation (% positive% negative), #/pts change, people reached vs.
all, Cost Per Increase in Sentiment Index (CPISI)
Share of positive voice in category (= brand positive
mentions/category positive mentions), #/% change,
people reached vs. all, Cost per Increase in Share of
Positive Sentiment (CPISP) 3
Relative net promoter score (NPS) in category
(=brand NPS category NPS), #/% change, people
reached vs. all, Cost per Point Increase in NPS
(CPINP)
Action
Registration: RPA, CPA, $, #
Sale: RPA, CPA, $, #
Advocacy: RPA, CPA, $, #
Depending on volume and the prevalence of negative discussion, we may look at share of net positive voice,
where net positive voice = brand positive mentions brand negative mentions
O g i lv y I n s i g h t 5
Preference
Share of positive voice within category = volume of positive-sentiment mentions for brand/
total volume of positive-sentiment mentions in category
Can also be compared to/calculated as share of net positive voice, which = volume of (positive
negative) mentions for brand/total volume of (positive negative) mentions in category
Calculated via social media monitoring/listening software or through a direct consumer
survey
Cost per point increase in above metric
Action
Campaign- or influencer-attributable actions
Calculated using tracking analytics or through Ogilvys social media activation
platform technology
Note that the measures shown are representative, not comprehensivethe key focus of the
model is to use categories and metrics that provide simple, useful data in ways that provide
for ease of comparability and analysis.
The denominator may also be modified to include only the brand plus specific competitors
6 O g i lv y I n s i g h t
consumers,
influencers or both
We used the model to evaluate two example brand campaigns. We selected these two campaigns
to demonstrate the flexibility of the model in measuring impact on both consumers and
influencers. For the purposes of this discussion, consumer represents traditional end users/
target audience members. We define an influencersomeone who is effective at broadly
distributing a message or driving actionbased on a number of factors, including a persons
connectedness, reach and ability to engage and drive results around a specific target audience
and discussion topic.
We employed our social media listening post and survey-based data collection methods for
these campaigns.
O g i lv y I n s i g h t 7
n Post
67%
48%
48%
20%
6%
10%
Positive Opinion (9 or 10)
8 O g i lv y I n s i g h t
19.0%
18.2%
17.7%
18.0%
17.0%
17.1%
16.0%
16.2%
4-month Trend
15.0%
Dec-08
Pre-Launch
5
Jan-09
Campaign Launch
Feb-09
Mar-09
It should be noted that our teams approach is software-independent. We set up Social Media Listening Posts
with underlying software that is the most appropriate for the specific client engagement.
O g i lv y I n s i g h t 9
Today
In our early uses of the model, were finding that Preference and Positioning are two key areas
in which social media can deliver particularly efficientlyand were now implementing new
Action-based campaigns to further expand the range of social media goals.
We believe the model represents a useful step forward in social media impact measurement,
primarily due to its focus on tracking metrics with comparability across different types of
advertising and communications.
As social media increasingly becomes a standard component of both advertising and communications campaigns, this type of simple, cross-channel comparative framework will become more
useful in answering media allocation questions and helping marketers decide which social
media efforts are worthy of scaling further.
10 O g i lv y I n s i g h t
References
1 eMarketer Bridge Ratings and University of Massachusetts 2007.
2 RAMA/BIGresearch All About Moms, 2009.
3 Depending on volume and the prevalence of negative discussion, we may look at share of net
positive voice, where net positive voice = brand positive mentions brand negative mentions.
4 The denominator may also be modified to include only the brand plus specific competitors.
5 It should be noted that our teams approach is software-independent. We set up Social Media
Listening Posts with underlying software that is the most appropriate for the specific client
engagement.
The Nielsen Company, Global Faces and Networked Places, 2009.
O g i lv y I n s i g h t 11
John H. Bell
John heads up the 360 Digital Influence teamOgilvy PRs global, digital word
of mouth marketing practice designed to manage brands at a time when
anyone can be an influencer and we are all influenced in new ways. His team
has developed and executed social media strategy for clients as diverse as
TJMaxx, Lance Armstrong Foundation, Lenovo, Unilever and more. The teams
focus is on engaging through conversations, outreach to new influencers and
word of mouth marketing.
John is a Web 1.0 graduate. As Creative Director at Discovery Communications,
he transformed a single Web site into 14 Web communities and services from
DiscoveryKids.com to Animalplanet.com and more. In the early nineties, when
interactive television was imminent, John headed up the creative studio for the
joint ITV venture between Viacom and AT&T.
Currently, John serves as the President of the board of the Word of Mouth Marketing Association. He teaches
graduate studies in Digital Influence at Johns Hopkins University.
His blog is: The Digital Influence Mapping Projecthttp://johnbell.typepad.com
His twitter name is: jbell99
12 O g i lv y I n s i g h t
Ogilvy & Mather 2009. All rights reserved. Neither this publication nor any part of it may be reproduced, stored in a retrieval system or transmitted in any form or by
any means, whether electronic, mechanical, photocopied, recorded or otherwise, without the prior permission of OgilvyOne Worldwide.
14 O g i lv y I n s i g h t