What impact are fake news and clickbait having on consumer trust in social media? We look to find out and assess the implications for marketers and their media buying.
1. A ‘CRISIS OF TRUST’ IN SOCIAL MEDIA?
Why does it matter and what are the potential
implications for marketers and the media.’
2. TODAYS DISCUSSION
The ‘crisis of trust’ – where has it come from?
•
Why does it matter?
•
How did we get to the current situation.
•
Where do we go from here?
3.
4. Edelman 2018 Trust Barometer
• Less than a quarter of the UK population now trust social media
• Edelman study finds 7 in 10 say social media firms aren’t doing enough to address extremism and
cyberbullying
• 64% back tougher regulation of social media companies
Ed Williams, CEO of Edelman UK
“After a flood of negative headlines in 2017, it’s time these companies sat up and listened. The public want
action on key issues related to online protection, and to see their concerns addressed through better
regulation. Failure on their part to act risks further erosion of trust and therefore public support.”
5. Why it is important? Huge increase in smartphone adoption
6. Although choice of screen depends on context and activity…most of online
time now on mobile and most of that time is on social media
6
7. With a profound impact on ‘other’ news and entertainment sources
8. To add insult to injury….news brands increasingly rely on the
‘duopoly’ for their digital traffic
9. We are at a tipping point: online is more than 50% of all advertising spend
10. However 90% of digital ad growth go to two platforms
Google and Facebook dominate due to scale, quality user IDs and ease of buying media for SMBs
Together the big two have 65+% of the total market, much higher share of mobile
12. Looking at digital growth in more detail - It’s direct response –
or activation – advertising that’s growing
0
2,000
4,000
6,000
8,000
10,000
12,000
2000 2015
Analogue Digital
Display ad expenditure (£m)
0
2,000
4,000
6,000
8,000
10,000
12,000
2000 2015
Analogue Digital
Direct response ad expenditure (£m)
13. We have gone from 40:60 activation/brand to 50:50
60%
1%0%1%
38%
Analogue brand display Digital brand display
Digital DR display Digital DR
Analogue DR [Source: Enders Analysis estimates based on Warc/AA]
Adspend by campaign goal, 2000
61% Brand display39% direct response
43%
9%9%
28%
12%
Analogue brand display Digital brand display
Digital DR display Digital DR
Analogue DR [Source: Enders Analysis estimates based on Warc/AA]
Adspend by campaign goal, 2016
52% Brand display
48% direct response
18. Where do we go from here?.....a crisis of trust amongst advertisers as well as consumers?
19. The supply chain crisis: publishers lose out; agency reputations tarnished
100
0
6
15
12
1
9
15
10
32
0
10
20
30
40
50
60
70
80
90
100
[Source: ANA/ACA/Ebiquity/AD/FIN, Enders Analysis]
TitleProgrammatic ad tech tax w/ATD model,
Canadian study from 2015-2017
100
0
6
4
2
26
62
0
10
20
30
40
50
60
70
80
90
100
Advertiser
spend
Agency of
record
Targeting
data
Verification (Publishers,
Instant
Articles)
Facebook
[Source: Enders Analysis illustration based on WFA/agency conversations]
Title
Programmatic duopoly, Facebook model
20. Disparate views on digital trading
Agencies Brand owners
Overall market perspective Good Poor
Market maturity Learning from mistakes Problems have gone nowhere
Viewability and fraud Teething problems Fundamental obstacles
Gap between “promise” and “practice” of
audience targeting
Good progress
Some teething problems
Fundamental obstacles
“Black box” of agency data Nature of the model Not trusted
Lack of financial transparency Great value for money Fear that they are overpaying
Key goals and metrics
Targeted consumers
Volumes
Context
Targeted consumers
Volumes
Context is key (but often little willingness to
pay for it!)
[Source: Enders Analysis study, Context is Queen]
21. Where do we go from here? (and if I really knew I’d be a very rich man…)
21
23. x3 effective for
mid-funnel metrics
(favourability, consideration)
+51% higher
effectiveness
“THE HALO EFFECT”
(2016)
24. 41% uplift on brand
KPI’s
(online magazines)
30% longer
interaction time
“Metrics that
Matter”
(2016)
25. 16% more personally
relevant than social
(Facebook) content
13% greater impact
on memory encoding
than Facebook
“Your Brain on
Premium Content”
(2017)
26. Ad viewing is 2.5x
more likely on a
newsbrand site
30%+ longer
dwell time
“Attention in
a quality
environment”
(2017)
27. Ads on publisher-branded
sites increase consideration
by 60% vs. sites without
publisher branding
For regular readers
consideration is 152% higher
on publisher-branded sites
“(In)famous by
association”
(2017)
28. Themes from these
and other recent studies
Editorial media delivers significant…
impact on
medium-to-
long term
brand health
positive
impact
on key
business
metrics
Significant
“media
multiplier”
effect
29. Project objectives
Neuroscience the obvious
approach to take
Better understand what’s happening in the brain
when consumers are exposed to ads in different
online environments
1.
Focus on editorial contexts vs social media2.
30. AOP research: Headline results show that ads seen in a premium
context are:
•Viewed for longer (+17%)
•Create 29% higher engagement (personal relevance)
•Generate greater levels of left brain (+42%) and right brain (+9%) memory encoding
than ads on social media (Facebook and YouTube) *
•Ads seen within a premium context also elicit stronger, more positive emotional
responses
*Left brain is words and detail, right brain is emotional triggers
31. A reappraisal of ‘premium’ environments?
• Different from platforms, bloggers, social
• Brands are more important in the digital and mobile era than in the past….they are a more important
navigation post, and stamp of authority and quality, than ever before
• BUT every consumer touchpoint, every instance of the brand needs to embody those premium brand
values
• Ensure that everything is designed to reinforce the differentiated, premium quality of the brand?
32. Effectiveness v Efficiency? Is it wise to ignore the 60:40 rule?
• The purpose of brand marketing is to create sustainable premium prices
• By generating awareness and reinforcing emotional connections, brand advertising ensures consumers
select a brand rather than own-label – the point of brands is they generate and sustain a higher price
point
• Quality brands are in part a consequence of relevant association – no medium on the planet provides a
richer combination of consumer targeting when they are in the right frame of mind in a premium
environment than that provided by trusted media brands.
• The ‘Long and the Short’ of it is that the optimum balance for effective campaigns hasn’t changed….but
marketing strategies have.
33. Ignore analogue at your peril; We need to remember that the over-45s are
the true multimedia generation…
34. And over-45s are 60% of unequivocal income/spend
Note: Adult population. Split by age/age of head of household
[Source: Enders Analysis based on ONS Family Spending and Household Disposable Income and Inequality]
Disposable household income Consumer expenditure
35. Green shoots….We do seem to be seeing signs of an increasing willingness
to pay for digital content..
Notes de l'éditeur
As a result of their data advantages, these two companies captured 90% of UK digital ad growth in the last year, with the rest of the market left fighting over scraps.
The digital ad market is increasingly divided into Facebook and Google on the one hand, and everyone else on the other. The big two’s scale, depth of usage, and most significantly their ability to solve the cross-device measurement problem through encouraging identifiable, logged-in usage across channels and devices makes them much more attractive to advertisers than the messy, probabilistic world of the rest of the internet.
This advanced portion of the market has its own issues, which we’ll discuss, but first let’s take a look at the rest of the market and what options there are for it.
Uber – Fetch example
New CMO at Uber
Uber is seeking a minimum of $40m in damages after claiming Fetch ‘squandered tens of millions of dollars’ snapping up nonexistent, nonviewable or fraudulent ads which it then sought to cover up
To be clear from the front, our findings don’t highlight insuperable issues with machine-to-machine trading or with algorithmic marketing
Everyone thinks automated trading will grow and everyone thinks it will improve
But I think we do highlight some gaps in perception regarding how the marketplace is performing today, and how close it is to the “promise”
Additionally, brands want to know how much of their spend is going to publishers, they want more transparency – I will return to this question
Automated evangelists will sensibly and correctly argue that automated can address every one of these issues
Our point is different: as automated moves up the value chain, not everyone seems to believe the tools in the marketplace are fully optimised, or that they are even fully understood
And among some general concerns, one specific fear rattled around these discussions: that context -- the value of quality media for motivating consumer brand engagement -- is being under-rated as a measure, as an objective, as a relevant marketing indicator – for a growing amount of marketing expenditure
If this is right, or even a bit right, then we believe the timing of this discussion is critical
Let’s get some perspective on digital marketing spend
Summary:
comScore US wanted to understand whether its true that ads appearing on premium publishers performed better than on other inventory? And if it’s true that premium inventory performs better, how much more valuable is it?
They conducted research to examine the branding effectiveness of digital display and video ads appearing on Digital Content Next (DCN) member sites (seen as ‘premium’) compared to non-member sites. (FYI – DCN is the US equivalent of the AOP.)
The study found that this segment of publishers delivered significantly better branding effectiveness results across a number of measures. The primary driver of this increased effectiveness is the halo effect that comes from the value of the contextual environment in which these ads are seen.
Methodology:
comScore compared metrics of inventory quality and advertising effectiveness for DCN premium publishers vs. all other publishers. The study analysed the brand lift results for 15 large brand display and video ad campaigns across a number of advertiser categories, including consumer packaged goods, household goods, auto, consumer electronics, and telecom. The campaigns were selected from among previously run comScore Brand Survey Lift studies that were sufficiently large in impression volume, ran on both DCN and non-DCN publishers, had at least 400 total survey respondents, had at least 75 respondents in the DCN publisher exposed group, and had at least one brand metric showing statistically significant lift.
Year: 2016
Summary:
‘Metrics that Matter’ acknowledges that metrics such as relevance and quality are much harder to move than others. These metrics also have some of the most significant impacts on brand KPI’s. The deep and meaningful connections magazine readers have with their chosen magazine brands makes them better drivers of key KPI’s.
Methodology:
Magnetic teamed up with Carat Insight, making use of their Integrated Communications Evaluation (ICE) tool – gaining access to a back catalogue of print, online magazine environments, display, advertorial and native activity. This consists of 65 studies, dating 2010-2015. Carat looked at all of the campaigns from the last five years to make comparisons. For a more detailed magazine media analysis, they focused on 39 magazine media campaigns, amounting over £11m in spend with publishers. ICE is a three stage process whereby people are asked about their brand perceptions, experiences and whether they recognise marketing and media activity in-situ. Carat then performs factor analysis and structural equation modelling (SEM) to quantify themes and finally regression analysis to isolate and quantify how effective each touchpoint is.
Year: 2016
Summary:
Teads.tv partnered with Neuro-Insight in the US to quantify the value of premium content for video advertising. The findings show that premium editorial content outperforms Facebook in 3 ways:
- More personally relevant (16% more personally relevant than Facebook social content)
- Creates memorable advertising experiences
- Provides an opportunity for a broad range of creative strategies to be successful
Methodology:
Neuro-mapping approach measuring brain activity looking at specific regions of the brain.
- Environment: Two test groups of n=50. One was exposed to eight, 15-second video ads on mobile within premium articles; the other group was exposed to the same eight ads within their personal Facebook feed
- Four publishing brand environments: Time Inc., Conde Nast, Forbes, and The Atlantic
- Social Environment Tested: Facebook Social Content
- Advertiser Video Creatives: eight, 15-second mobile creative spots across CPG, QSR, hospitality, entertainment, electronics, retail, and Ecommerce
Summary:
Eye-tracking research to demonstrate the value of newsbrands’ print and digital offerings for brands. Shows the high levels of attention paid to ads on newsbrands, with uplifts in standout (whether an ad is seen) and dwell time (length of time ad is looked at) compared with non-newsbrand sites/publications.
Methodology:
Digital and print data collected through desktop eye-tracking devices.
Digital – Tested through Aimia Lumen 500 strong panel. Panellists were asked to read articles on a selection of newsbrand and other sites, browsing naturally
Print – Print tested via a hall test. Sample = 739 for Christmas element and 160 for Black Friday element. Respondents asked to read the first 19 pages of a pdf copy of either The Mirror, The Metro, The Daily Mail, The Times or The Sun.
Year: 2017
Summary:
Inskin Media worked with ResearchNow and Conquest Research to investigate reactions to online advertising in different scenarios. They wanted to find out what impact an ad’s environment has on its effectiveness? Does the tone and content of an article really affect perception of the brand advertising around it?
Methodology:
An online test with multiple brands, display formats and publishers.
From July – August 2017, a nationally representative sample of 4,364 UK respondents was surveyed using both explicit and implicit measurement methodologies in order to uncover both conscious and subconscious associations people held of the brands tested.
Year: 2017
The studies outline in the previous charts use different research methodologies and cover different countries/environments/metrics.
However, there are some very clear themes about how advertising within editorial/premium environments is performing vs. other types of digital context…
It seems logical that a key determinant of advertising performance is what is happening in the brain when people are exposed to communications.
For this study we decided to focus on editorial/premium environments (provide by AOP/Newsworks members) and compare brain activity with social media environments.
To best understand this we decided that implementing a neuroscience study would be appropriate.
They dominate wealth, income and consumer spending in the UK. Employment rates of their group are higher than they have ever been, and they own or are continuing to set up businesses that expand their wealth, which they already had lots of due to ownership of the property stock of the country.