Trust & Predictive Technologies 2016

Trust & Predictive
Technologies 2016
A data driven study into privacy, prediction and personalisation by Edelman
& The University of Cambridge Psychometrics Centre

// 02
Contents
Executive Summary // 03
Introduction // 04
The Trust Divide - Business & The Public // 10
The Trust Divide - Marketers & The Public // 15
Emerging Best Practice // 17
Ambient Intelligence - the Next Level // 21
Conclusion // 22

// 03
E
delman, in conjunction with The University
of Cambridge Psychometrics Centre,
embarked on a research study to examine
and monitor public attitudes towards Big Data and
predictive technologies. Building a data-driven
platform that enabled individual engagement and
contextualised feedback, the study explored the
diverse opinions and psychological attributes of
both global consumers and key corporate decision
makers in this area, namely marketing and
communications professionals.
Consolidating the views of more than 34,000
individuals from around the world, the study
provides an up-to-date view of the pivotal role of
trust in the development and adoption of predictive
technologies. Its findings lend some objectivity
to several rapidly changing areas of technology,
business and society, with the goal of stimulating
further research and an informed dialogue on
what Big Data can do, and more importantly, what
it ought to be used for. Privacy concerns operate
across age, gender, country and personality: Psycho-
demographic variables explain less than two percent
of the variance between yes and no answers to
questions about uses of Big Data. Privacy is therefore
a universal concern, not an esoteric interest.
Executive Summary

// 05
Key Findings
Privacy concerns operate across age,
gender, country and personality
Psycho-demographic variables explain less
than two percent of the variance between yes and no
answers to questions about uses of Big Data. Privacy
is a universal concern, not an esoteric interest.
There is a deep lack of trust in
data-driven businesses and in
government
71 percent of people thought most companies with
access to their personal data did not use it ethically;
only 26 percent of people trust the government not
to sell their electoral roll and demographic data
without their consent.
Marketers are at risk
of overestimating
consumers’ willingness to
adopt predictive technologies
Across sectors, marketing professionals were
consistently more knowledgeable and more open to
sharing data for prediction than the general public.
This reveals a clear need for companies to better
communicate their data practices, or face potentially
dire consequences.
There is a clear gap between
consumer desire and the reality of
how predictive technology is used
84 percent of people thought predictive tech should
be used to improve the quality of healthcare and
47 percent of people thought it should be used to
determine the price of their car insurance.
People want personalisation
66 percent of people would prefer to see
personalised advertising, assuming they
have to see some advertising. Privacy, transparency
and relevance are the building blocks of effective Big
Data-based marketing.
‘Pay for Privacy’ is a real opportunity
for traditionally data-dependent
businesses
27 percent would pay $3 a month to use Facebook
without their behaviour being recorded. Offering
paid options helps remind consumers that their data
has value, and that even if they use a service for free,
they are still effectively paying for it.
Businesses are investing in smarter
Big Data
77 percent thought their organisation
ought to invest in predictive data and 94 percent
said it was important for them to understand the
psychological attributes of their customers.
Demand for secondary services from
IoT data is soaring
57 percent of people thought that e.g.
smart fridge data should be used to recommend
groceries to them when they go shopping; 58 percent
would like to be automatically warned of unhealthy
dietary habits.
There is demand for personalised
finance, yet distrust of actuarial
prediction
The majority think predictive tech should not be
used to assess mortgage eligibility or likelihood of
default (62 percent and 67 percent respectively), but
were open to its use for better account management
and advice.
1
2
3
4
5
6
7
8
9

// 06
Introduction
B
ig data and predictive technologies are
among the most powerful tools for positive
change in business and society. The term
‘predictive technologies’ is used in this report,
and was used in the consumer survey, to refer to
services, analytical techniques, machine-learning
algorithms and other tools capable of discovering
and analysing patterns in data to predict future
behaviour on the basis of past behaviour.
Each of us are generating more data than ever
before, with at least 2.5 quintillion bytes’ worth of
human behaviour being digitised every day. This
rapidly expanding and searchable landscape of
information is being utilised by organisations in
practically every sector. From FitBit in wearable
technology or Samsung in Internet of Things
(IoT) devices, to Barclays in consumer finance and
indeed the UK Government. Furthermore, every
organisation from Uber to Homeland Security are
getting better and better at using it.
But how good are we really at understanding the
individuals behind all this data? Does control
rest with the organisation, the government or the
citizen? And how can Big Data become personalised,
inclusive and relevant for everyone, rather than
merely being ‘big’?
The University of Cambridge Psychometrics Centre
(hereafter UCPC) is attempting to tackle these
questions head on. Through multidisciplinary
academic research and product development, the
UCPC hopes to make Big Data not just machine-
readable, but more human-interpretable. Its team
has created a trait prediction engine called Apply
Magic Sauce that translates digital footprints of
behaviour (such as the Pages you Like on Facebook),
into detailed psycho-demographic profiles. These
predictions enable citizens to see how others see
them, bringing to life the ways in which our digital
personas are monitored and perceived.
UCPC’s research is underpinned by the largest and
richest social science database in history, called
myPersonality, comprising of the psychometric test
scores and social media profiles of over six million
volunteers globally. This resource provides a deep
psychological perspective on the audience attributes
of over 200,000 brands. Edelman supports this
research and brings the impressive predictive
potential of the UCPC's tools to its clients.
With a sample of only 10 Likes, the Psychometrics
Centre’s platform can more accurately predict an
individual’s personality profile than their colleagues.
Not only does it surpass 360 degree feedback,
but given a sufficient tranche of digital behaviour
(around 300 Likes), the computer models can in
fact better detect personality traits than a person’s
husband or wife1
. Combined with predictions of
age, gender, relationship status, intelligence, life
satisfaction, leadership potential, political views and
a host of other variables, predictive technologies
such as these promise to bring greater control and
personalisation to all aspects of the internet.
From travel and retail to healthcare and education,
building psychological sensitivity into machine
intelligence is arguably the next frontier for Big
Data. But with that comes a pressing need for service
providers to evolve with consumers at front of mind.
The Psychometrics Centre’s research was recently
cited by the European Data Protection Supervisor in
Opinion 7/20152
to illustrate that ‘the more powerful
computers become, the more acute is the challenge
[of deciding] what is fair and lawful and what is not
when it comes to big data analytics’. This exercise
requires a comprehensive and multidisciplinary
approach to balance the legal, ethical, commercial
and personal interests at stake.
1 www.pnas.org/content/112/4/1036.abstract
2 https://secure.edps.europa.eu/EDPSWEB/webdav/site/mySite/shared/
Documents/Consultation/Opinions/2015/15-11-19_Big_Data_EN.pdf
1.1 The Power of Predictive Data

// 07
As Edelman’s most recent Trust Barometer revealed
- there is a significant divide between the beliefs
of business and government, and the beliefs of the
general public, with regard to how personal data
is being used. The optimism of the informed elite
contrasts with a deep mistrust from consumers in
many sectors. This report illuminates several aspects
of this conflict through the lens of the individual
traits and desires underlying the public’s beliefs. It
provides a summary of initial findings from what
promises to be an ongoing, knowledge-seeking
collaboration between Edelman and the University
of Cambridge Psychometrics Centre.
1.2 Research Objectives
The aim of this research project was to understand
and monitor the public’s attitude towards predictive
technologies. The assessment platform developed
for this purpose will continue operating at
www.predictivedataproject.com as a shared research
asset, enabling objective measurement of attitudes
towards the many possible applications of predictive
technologies, both now and in the future.
Another aim was to contrast the opinions of the
general public with that of data professionals,
uncovering the extent to which specialist knowledge
might engender different ideas on how predictive
technologies ought to be used. For this purpose,
part of the questionnaire targeted marketing
and communications professionals, whose rapid
adoption of digital solutions represents one
of the most advanced practical applications of
these technologies. Questions spanned the usage
domains of advertising, politics, quantified self and
wearables, data protection, cloud storage and IoT.
It is hoped that this research will help improve
the business world’s understanding of consumers’
views of predictive technologies, thus helping to
outline inclusive strategies and best practice for any
organisation facing challenges related to Big Data.

// 08
1.3 Research Methodology and Sample
This research represents the largest audit of public
opinion on Big Data and predictive technologies
in history. It was conducted by Vesselin Popov,
Business Development Director for the University of
Cambridge Psychometrics Centre, and Sandra Matz,
a PhD candidate investigating psychological fit in
marketing, finance and other business contexts.
Delivered openly to the public via UCPC’s online
Predictive Data Project assessment platform, 34,267
people from around the world have taken part,
providing detailed opinions on Big Data and self-
assessing their personality traits.
Of this group, 4,454 (13 percent) were marketing
and communications professionals, with at least
1,868 of these respondents answering all seven
'CMO-focused' questions. We can therefore be
confident that the study is a statistically robust
indicator of the views of both C-suite and less senior
marketing staff.
The research also benefited from strong levels of
engagement in all other areas of the study. 10,411
people answered all 27 questions relating to
predictive technologies, and 8,877 people answered
all the survey questions in total, which included 20
questions focused purely on personality assessment.
Participants were not financially incentivised, but
were rewarded for their honesty and participation
by being offered instant feedback for every answer.
This was done using an iterative calculator of
cohort similarity, which displayed the percentage
similarity of an individual’s yes/no response
pattern to the response pattern of thousands of
previous test-takers. Participants could therefore
see instantly how similar their views were to others’.
Personality was also calculated in real-time to
provide a summary of the individuals’ openness,
conscientiousness, extraversion, agreeableness and
neuroticism, following the popular BIG5 model.
Aside from size, the sample was truly global in
nature. 43 percent of participants were from Europe;
27 percent were from North America, and 15
percent from South America. In national terms, the
countries with the highest proportion of respondents
were the USA, with 23 percent of participants, the
UK with eight percent, Brazil with seven percent,
and France with five percent.
Lastly, self-reported demographic data showed an
average age of 30 (with the distribution varying from
10 to 78 years old), and a 54:46 split between male
and female respondents. The sample was sufficiently
large and diverse to be representative of the general
public whilst permitting more focused group
comparisons in future.
Lastly, 16 percent of respondents worked in
technology, 11 percent in education, five percent in
advertising, and four percent in financial services.

// 10
The Trust Divide
Business & The Public
T
he 2016 Trust & Predictive Technologies
study found that only 29 percent of people
believe most companies with access to their
personal data use it ethically. Furthermore, only 22
percent of respondents thought the threat of bad
press was a sufficient deterrent to prevent businesses
from misusing personal data.
This is a severe indictment of the current levels of
consumer trust regarding corporate data practices,
and a strong indication that consumers are looking
for more regulation from government. 78 percent of
the public specifically thought that legal restrictions
were more important and effective than public
relations concerns at preventing
misuse of their personal data. Yet,
in many cases, it is likely to be the
PR and marketing department
rather than the legal team making
the final decision on whether a
technology, once approved, ought
in fact to be deployed. It is in this
grey area that considerations of
law and ethics interact with those
of revenue targets, competition,
media, brand strategy and customer service.
Where new regulation is implemented, its application
is often hindered by legal complexity on jurisdictional
issues, place of breach, place of business and other
technical aspects, such as the ubiquity of cloud
computing3
. All in all, the fast pace of technological
change in Big Data and prediction make the
regulatory landscape very difficult to navigate,
especially for smaller enterprises.
The study also found a low level of consumer trust
in government. While 55 percent of those surveyed
thought the government should use predictive
technologies when designing new policies, at least
in some areas, only 26 percent of respondents
3 Companies, digital transformation and information privacy: the next
steps, a 2016 report from The Economist Intelligence Unit http://
www.eiuperspectives.economist.com/sites/default/files/EIU_
Companiesdigitaltransformation_PDF_1.pdf
trusted the government not to sell their electoral
roll or demographic data to companies without their
consent. When it came to their private financial
information, however, a much higher 72 percent of
people said they trusted the Government not to sell
it on without consent.
Contrasting statistics such as these emphasise
the fact that context matters, and that one must
consider the specific data sharing being proposed
even when the parties are the same. There are
different expectations of privacy in every situation,
and different opportunities to miss the mark as a
business. Recognising the areas where consumer
desire does not align with current
practices should therefore
be a priority for data-driven,
consumer-oriented organisations
in future.
84 percent of people thought that
predictive technologies should
be used to improve the quality
of healthcare, for example by
helping doctors recommend
personalised nutrition plans.
But only 47 percent thought it should be used
to determine the price of their car insurance.
Contrasting the prevalence of the latter practice
with the slow adoption of the former raises an
important point: the rapid pursuit and adoption
of Big Data techniques in the last few years has
led to many assumptions being made about what
consumers really want. Another study by Pedraz et
al. (BMJ Qual Saf, 2015) found that 71 percent of
1,432 patients surveyed were happy to share social
media information with their doctor, providing
further evidence of demand and potential for Big
Data in health4
.
4 Linking social media and medical record data: a study of adults presenting
to an academic, urban emergency department, Padrez et al, British Medi-
cal Journal Quality & Safety 2015
http://qualitysafety.bmj.com/content/early/2015/10/09/bm-
jqs-2015-004489.abstract
2.1 The Public’s View
Rapid pursuit and
adoption of Big Data
techniques has led to
many assumptions
about what consumers
really want

// 11
2.2 Privacy Is Highly Valued
Low levels of consumer trust stem from wider privacy
concerns that touch all areas of life, not only those within
the remit of this study. But it is still remarkable, given
how much of our behaviours and social interactions are
digitally mediated, that the majority of respondents (58
percent) reported having not used a digital service due to
privacy concerns.
These are not inconsequential fears, but strong
emotional reactions that are undoubtedly driving
decision-making around which apps to download, which
email addresses to share, which social networking site
to log in with, and more. This is further supported by
26 percent of respondents in the IoTPI 20155
report
mentioning security or privacy of data collection as a
reason why they did not currently own a smart device.
While there are, without question, massively high levels
of mistrust around personal data usage, only 27 percent
would pay $3 a month to use Facebook without their
behaviour being recorded. Bearing in mind Facebook’s
average revenue per user is estimated to be around $1.33
per month, it is perhaps surprising that more would not
find this option appealing. Yet it would be foolish to think
of the 73 percent as being tight-fisted, or as only paying
lip service to the value of privacy. Many people consider
privacy a right rather than a service to be paid for, either
not appreciating Facebook’s business model or simply
not believing that a do-not-track option would ever be
feasible, let alone genuine.
On the other hand, 27 percent of the population,
translated to Facebook’s user base, would represent
a potential demand for anonymous usage from over
429 million users, a number that is surely sufficient
5 2015 US IoT Privacy Index from TRUSTe https://www.truste.com/resourc-
es/privacy-research/us-internet-of-things-index-2015/
to justify dedicated product lines. Aside from the
economic arguments for business model innovation
in how user data is treated and monetised, there are
other factors that could make ‘pay for privacy’ a good
idea for many businesses.
One is that it would help consumers realise that digital
services cost money to design and deliver, so if they are
not paying money to use them, then they must be paying
with their data. Even if the user still decides to continue
with the free option, at least they will appreciate they
are giving something up and are therefore more likely
to weigh up the benefits and disadvantages intelligently.
If the user makes an informed decision to share that
data, they are likely to be much more comfortable with
its use by the company, and the data is more likely to
be accurate, creating a mutual benefit. This creates a
beneficial data-for-insight partnership between the
user and platform that is the hallmark of a mature and
successful data-driven business.
In cases where people do choose to pay for a do-not-
track option, they can come to appreciate how sharing
parts of their data can actually improve their experience
of the service. For example, while a news site might be
ad-free upon subscription, recommendations for which
articles to read cannot be personalised without some
data being shared. However, the crucial difference in
this situation would be that the service provider might
generate revenue from increasing subscriptions for the
private, personalised version of the platform, rather than
by trying to record as much sensitive behavioural data as
possible. ‘Featurisation’ of data protection may therefore
become a more common feature of online platforms in
the near future.
The opportunity for industry to better align with
peoples' desires is now clearer than ever. Desire for
more control, customisability and service quality
is there and it is growing, but businesses cannot
assume that every sector or market will have
reached the same level. Clearly, with 16 percent of
people still believing that predictive technologies
should not be used in healthcare, providers cannot
assume all of their stakeholders to be equally
keen, and so should evaluate and tailor their
communications accordingly.
Overall, the paucity of consumer trust regarding
data is in keeping with the larger trends currently
at play, as revealed by the 2016 Edelman Trust
Barometer. Now in its 16th year, this independent
survey found that trust in business and government
is below 50 percent for the mass population - in over
sixty percent of the 28 countries surveyed. In other
words, the public’s trust in business and government
combined has barely increased since the days of the
Great Recession (2007-9).

// 13
It is vital that organisations appreciate the
consistency with which many consumers mistrust
the use of their personal data. Around the world,
people are deeply concerned.
The findings show that personality differences
accounted for only two percent of the variance
between opinions on Big Data. This means
that privacy concerns are more universal than
previously thought, not being merely characteristic
of guarded or unadventurous users, but widespread
across the population. The age, gender and location
of participants were also not predictive of whether
they would accept or reject a proposed use of
predictive technologies.
The predictive data project is collecting longitudinal
data to investigate these effects in greater detail,
and it may be that the public’s understanding of the
technology is still too nascent to bring out strong
effects mediated by individual differences. For the
moment at least, it would appear that considerations
of trust and privacy are so pervasive in consumers’
minds that they cross international, cultural and
psychological boundaries. One possible advantage
for organisations given these findings may be that a
clear and consistent stance on privacy issues could
help to inform large segments of the public, without
necessarily alienating people of certain psychological
profiles. The flip side is that this universality
may not last for very long, and a broad brush
communications approach would still not recognise
the personal nature of sharing one’s unique digital
footprint with a third party.
Businesses will ultimately need to decide upon a
strategy that instils confidence in the public and
trust in the individual consumer. Recognition of
universal concerns must be balanced with the
desire for personalisation if businesses are to show
leadership in this area and remain sensitive towards
the consumer mind set. While it could be said that
‘the EU is the highest common denominator when
it comes to privacy issues’ (EIU Report 2016), there
is also a strong impetus for legislative reform in
the United States6
. Calls for advancement of the
Consumer Privacy Bill of Rights, establishment of a
6 e.g. Big Data: Seizing Opportunities, Preserving Values. Interim Progress
Report. February 2015. Executive Office of the President.’
www.whitehouse.gov/sites/default/files/docs/20150204_Big_Data_Seiz-
ing_Opportunities_Preserving_Values_Memo.pdf
national data breach standard or greater oversight
of data brokers are principles not far removed from
those underpinning the Digital Single Market.
The outcomes of these debates are global in scale
but individual in effect. It is therefore essential to
recognise the small but significant role of personality
in shaping our views on Big Data. This role was
elucidated by the Predictive Data Project in findings
that are largely in line with prior research in this
area7
; Korukonda (2007). For example, this study
found that people who are more accepting of
predictive technologies tended to also be more open-
minded, cooperative and extroverted than those who
are more reserved on the topic. Agreeable people
were more likely than average to share their data for
new services, but less likely to admit that those with
access to their data could learn something about
them. Liberal and artistic people were more trusting
of companies and the government in general, but
less comfortable with predictions for financial
purposes, including eligibility for car insurance,
default risk on loans and mortgage assessments.
Finally, extroverts had greater belief in the threat
of bad press as being a powerful deterrent against
corporate misuse of personal data, possibly as
a function of their greater engagement with the
outside world.
7 Zhou, T., & Lu, Y. (2011). The effects of personality traits on user accep-
tance of mobile commerce. Intl. Journal of Human–Computer Interaction,
27(6), 545-561.
Korukonda, A. R. (2007). Differences that do matter: A dialectic analysis of
individual characteristics and personality dimensions contributing to com-
puter anxiety. Computers in human behavior, 23(4), 1921-1942.
2.3 Big Data Issues Transcend Psycho-demographics
"Privacy concerns span the
full spectrum of personality,
but consumers still exercise
their rights as individuals.
Communications need to be
both transparent and tailored
to inspire long-term trust in
the business of Big Data"
Vesselin Popov, Business Development Director
University of Cambridge, The Psychometrics Centre

// 15
The Trust Divide
Marketers & The Public
A
cross the board, the Predictive Data
Project found that marketing and
communications professionals are more
accepting of the use of predictive technologies - and
view their adoption as more desirable - than people
who do not work in marketing. This supports
our hypothesis that the opinion of marketing
professionals towards predictive technologies is
running significantly ahead of the general public,
to the point where businesses are at risk of gravely
overestimating consumer willingness to adopt new
predictive technologies.
For example, 50 percent of CMOs are aware that our
future behaviour can be accurately predicted from
our personal data, compared to 36 percent of non-
CMOs. Also, while 62 percent of CMOs have heard
of “the right to be forgotten”, that figure drops to less
than half (47 percent) for non-CMOs.
Additionally, 94 percent of marketing professionals
said predictive technology is important to help them
understand the psychological attributes of their
customers, and 78 percent said their organisation
needs to invest in predictive technologies. This
highlights the cross-sector momentum in marketing
departments to utilise the latest technology and
integrate analytics into sales, media planning and
new product development.
A key point here is that issues such as the right to
be forgotten are no longer just part of in-house
professional conversations, but are present in the
public discourse as well. The legal complexities of
putting prediction into practice are to be elucidated,
not shielded from consumers, and further education
is needed on both sides of the ‘share’ button. This
study argues that businesses ought to be more
proactive in explaining changes in the consumer
data landscape to the public. Information should
not be withheld in order to preserve speculative
knowledge advantages over competitors, but open-
sourced and shared for discussion.
Too often a change in rules manifests in a new
clause buried in privacy policies or compliance
procedures, rather than an opportunity for
business-to-consumer dialogue, wherein lies the
greater competitive advantage. Consumers are both
more concerned and more curious than ever, and
businesses can foster greater respect by educating
them and shaping simple conversations about
privacy through their products and services. This is
the view adopted by the European Data Protection
Supervisor in arguing that “as a society, we must be
able to look into the ‘black box’ of big data analytics’
and that privacy policies should ‘genuinely serve to
safeguard the interests of the individual…not merely
to shield the controller from legal liability’.
Without serious efforts to develop, adopt and
popularise ethically sound data practices,
consumers’ lack of trust could become a real
problem for organisations that strive to be data-
driven. The complexity of the subject matter means
that those marketers who get ahead of the curve in
communicating with consumers will succeed. Those
who waiver risk user withdrawal and irreparable
damage to their brand.
3.1 The Marketing Professional’s View
94% of CMOs said
predictive technology
is important to help
them understand the
psychological attributes
of their customers

// 16
Personalisation is arguably the most immediate
benefit of predictive technologies, a realm where
the use of personal data can deliver a more personal
experience instantly. One might therefore expect
the views of marketers and of the public to be more
closely aligned. The Predictive Data Project confirms
this, but with some caveats.
64 percent of CMOs thought the brand, product and
or services they offered were highly personalised to
individual customers, while 55 percent of non-CMOs
thought this was the case – nine percent less. As was
the case with the relatively high use of prediction in
insurance compared to healthcare, such gaps should
be viewed as major opportunities for improvement.
With a reported 144 million people8
now using ad
blocking software when browsing the web, and
Apple’s recent addition of ad blocking to iPhone
and iPad software, the future of advertising is far
from certain. The study found that 66 percent
of people prefer to see personalised advertising,
on the assumption that they do have to see some
advertising, showing a clear demand for more
relevance in future. 56 percent of respondents
also agreed with the more general suggestion
that predictive technologies should be used to
personalise their online and mobile experiences.
8 PageFair, 2014
3.2 Personal, Just Not Personal Enough

// 17
Emerging Best Practice
C
ompanies should bear in mind that business
is increasingly expected to take on the
responsibility of solving the challenges
facing society. The 2016 Edelman Trust Barometer
found that 80 percent of people (up from 74 percent
in 2015), think that ‘a company can take specific
actions that both increase profits and improve the
economic and social conditions in the community
where it operates.’
This shows that the vast majority of the public is
not anti-profit, or indeed anti-progress – quite
the opposite in fact. There could therefore be an
opportunity for companies to harness greater support
for predictive technologies, if they can demonstrate
how it helps address wider challenges too.
4.1 Business Must Lead
As Jonathan Hargreaves, Vice Chair of Edelman’s Global Technology
Practice says: “The ability to see around the corner is a super power
that these new predictive technologies enable. Everyone in a leadership
position has a responsibility to ensure this new power is used for the
right purpose and good outcomes.”

// 18
4.3 A Two-way Conversation
As the research findings show, creating a two-way
conversation with consumers is of the utmost
importance. Organisations have a responsibility to
empower and engage people.
Predictive technology, and the issues
surrounding it, are complex, as is the public’s view -
so only a sustained dialogue will prove effective.
For example, Edelman’s 2015 Earned Brand research
found that 66 percent of consumers believe brands
are innovating simply to make more money for the
company, rather than e.g. having a positive societal
impact or solving bigger environmental problem.
The key insight here, is that while the promise of
innovation can inspire people – they first need to be
reassured. Specifically - consumers are twice as likely
to want to be reassured than inspired by a brand.
Organisations should also realise that while the
data belongs to the customer – the power often lies
with them, as it is businesses who own, operate and
understand the predictive technology, as well as the
benefits it makes possible.
With this power comes the responsibility to help the
general public understand too.
4.2 How Media Providers are Getting It Right
Three words to remember: Privacy, Transparency & Relevance.
These are at the core of making Big Data
work for everyone. This means focusing on
offering accurate, detailed information about how
customers’ information is being captured, stored,
used and shared.
It is also important to celebrate
and raise awareness of the
effective and ethical uses of
data that benefit consumers.
The average customer is not an
expert in predictive technologies,
so business has a responsibility
to educate them and to lead best
practice.
Some of the best current
examples of predictive
technology are theory-driven
recommendation systems based on psychological
insights about the customer – such as Netflix’s
showcasing of content it predicts to be of most
interest, based on the individual user profile and
previous viewing behaviour.
Not only does this enable the business to better
understand its customers, it enables it to prove
relevance, and explain – in individual terms – why
a particular recommendation has been made. The
benefit is clear, and accepted, when strategy follows
this format: “We recommend X,
because you watched Y.”
As well as Netflix, many other
media organisations are among
the most focused on privacy,
transparency and relevance. The
Financial Times for example,
gathers data in return for
access to the site – making the
connection clear between insight
gathering and reward – as long
as the questions are deemed
reasonable of course.
Again, each sector will have its own context - the
questions people would be willing to answer on a
site selling consumer dental equipment could be
very different to what they would be willing to tell
a newspaper.
Some of the best current
examples of predictive
technology are theory-
driven recommendation
systems based on
psychological insights
Edelman’s 2015 Earned
Brand research found
that 66 percent of
consumers believe
brands are innovating
simply to make more
money for the company

// 21
F
or those service providers using predictive
technologies, who are already following
best practice with regard to privacy and
transparency, and demonstrating its relevance
to the consumer – the next step is what we call
Ambient Intelligence.
Ambient Intelligence is what emerges when
organisations are using customer data and
predictive technology in a way that is both
immediate and ubiquitous.
Immediate in that there is a speedy delivery of
a relevant consumer benefit following the rapid
collection and predictive analysis of the data. And
ubiquitous in that the technology is deployed and
benefits delivered in both static environments, such
as the home and office, and mobile environments,
such as via a smartphone or tablet.
To be clear, this is an opportunity that exists right
now.
The research found that 52 percent of people
thought the benefits of home smart appliances - such
as a fridge that automatically pre-orders groceries
or a thermostat that pre-adjusts room temperature
- outweighed the privacy risks related to the
generating such data about the patterns of their
home life. This is a staggeringly high percentage,
considering how recently the Internet of Things has
become part of the ordinary citizen’s knowledge and
vocabulary.
Furthermore, 57 percent of people would like such
smart fridge data to be used to recommend groceries
to them when they go out shopping.
So, in spite of consumers’ concerns, if executed
properly, the market demand for Ambient
Intelligence is already there.
To truly take advantage of the opportunity
Ambient Intelligence offers however, the data has
to be sound, the recommendations correct, and
consumers’ consent explicit.
Fortunately, the research finds that some of these
are areas where CMOs are in a good position
already. 88 percent of CMOs said it was important
to always back up their decisions with statistically
valid data.
However, 32 percent of CMOs admitted their
company stores data on sensitive personal
attributes, such as ethnicity, sexual orientation and
intelligence. Explicit consent is required under the
Data Protection Act 1998 for this data to be collected
and even then it must be for a limited purpose and
amount of time.
This leaves a possible 68 percent of companies that
are storing email addresses, census information,
social media data or other digital records of
behaviour (which 48 percent of respondents
explicitly said they do).
If, as psychological research has shown, it is possible
to predict sensitive attributes from this data as well,
then companies could be unwittingly storing data
‘relating to sensitive attributes’, and so be liable
under the Data Protection Act 1998. Thus both
rigour and caution are needed, for businesses to
ensure they remain compliant.
Organisations should also be aware of what is
driving the emergence of Ambient Intelligence.
Broadly, technological progress is driving changes
in people’s behaviour, which is changing how
businesses are operating.
5.1 A Call to Action
5.2 Taking Advantage of Ambient Intelligence
Ambient Intelligence
The Next Level

// 22
Conclusion
As the below diagram shows, getting the approach
to data and predictive technologies right means
organisations can initiate a virtuous circle of
improvement. An ethical approach delivers more
consistent data and over a longer period of time.
Better data enables a company to enhance its
offering, building a stronger business, which leads to
a stronger brand from an ethical perspective too.
Of course, should best practice not be followed,
the reverse cycle is equally possible. The stakes are
high because the opportunities are so great, and
consumers' concerns so deep.
Fundamentally, Big Data will not be the final
frontier. Organisations still have a way to go
in building resilience to the shifts in consumer
mindset that accompany technological progress.
Undoubtedly, the information age ecosystem can
now make or break a business faster than ever.
T
he limits to how much value organisations
can create - using data, predictive technology
and ambient intelligence - are not defined
by a company’s hardware or data processing
capabilities. Rather, the value is limited only by
its creativity, vision and the design quality of its
relevant systems and processes.
The new value chain of predictive technologies, where
incentives align for business and the consumer, is as
follows: envision the potential value, identify what
data is required to make it happen, ensure data is
collected and analysed appropriately, and finally
deploy the solution in a personalised manner
6.1 Enabling & Realising Value
6.2 A Virtuous Circle
ETHICS
BETTER
OFFERING
BETTER
DATA
BETTER
BUSINESS
TRUST

Trust & Predictive Technologies 2016

Trust & Predictive Technologies 2016

More Related Content

What's hot (20)

Similar to Trust & Predictive Technologies 2016 (20)

More from Edelman (20)

Recently uploaded (20)

Trust & Predictive Technologies 2016