The Web Of Trust
David Amerland ????
Working on a new book. Can't tell you on what just yet, but I currently share snippets of the insights I gleam as I write.
As a social construct trust is a communal cognitive and affective response that, traditionally, has helped us survive. The communal threats these days may not be quite as life threatening but the need for trust is as important as ever. As the digital space is heading towards a semantic web can we hope to reach a level of transparency and connectivity that the online, connected, digital space becomes the village square of our more bucolic past? In this chapter we shall be looking at concepts such as algorithmic trust and the creation of trust networks in the digital space and how they affect us. What practical lessons can we draw from them? ????
We live not just in the digital age but in the digital domain. Whether we like it or not our physical self has a digital doppelganger whose digital footprint can be traced from our Tweets about politics and the lunch we had today to our Facebook shares and LinkedIn articles to the Google+ interactions. As a matter of fact, we could argue that over time our digital self, stripped from many of the requirements of our flesh and blood existence that force to co-exist in social structures which we have not consciously chosen to be part of, is pure intentionality and therefore a truer picture of us.?
If trust is the one constant, ubiquitous requirement for any type of relational exchange to take place how is it manifested in the digital space? What are the elements that make it strong or weak? Can it be manipulated? Is there such a thing as algorithmic trust? And if so, how is that even possible when trust is so precious and hard to obtain because it is, precisely, a very human emotion?
It is only natural that we start with questions. The digital realm is new to us still. We are still rapidly evolving in it and just when we think that we have got a handle on it, just when we feel that we understand what we are doing, we blink just a little longer and things change again, and we are, once more on a learning curve.?
While the digital realm is characterized by incessant change and non-stop challenges, it is also rife with opportunities we have never had before. Made up of data it also transforms us into data. On the web we are nodes, our interactions and engagements become edges and the tensile strength between nodes and their relational connections to each other becomes an intricate pattern of nodes and edges whose numerical values, ever changing, tell a story of friendships, acquaintances, jobs, careers, hobbies, likes, dislikes and new discoveries.
In many ways the digital realm is pure magic. It is almost anything we could have ever wished for except for the fact that we are still struggling to understand it, its impact and out place in it. In the digital realm the story of trust is born anew and within it, we have the chance to get some things we messed up in the real world, right.
The Semantic Web and Calculation of Trust
We looked a little at algorithmic trust in the second section of this book in Chapter 4 so here we are free to build on that foundation and actually explore things a little bit further. The ultimate dream of the semantic web is that it becomes a vast, transparent arena of connected data and people where one could digitally venture into any space with the same confidence that they leave their own front door in the morning.
“Lack of trust is one of the most frequently cited reasons for consumers not purchasing from Internet vendors” begins a detailed research paper on the subject that appears in the International Journal of Human-Computer Studies. Written by researchers from the University of Klagenfurt in Austria, the paper cites one instance after another of online situations where trust is not present and online transactions do not take place, despite the fact that the tools to carry them out are all there.
“Trust, in general, is an important factor in many social interactions, involving uncertainty and dependency. Online transactions and exchange relationships are not only characterized by uncertainty, but also by anonymity, lack of control and potential opportunism, making risk and trust crucial elements of electronic commerce.”
Since the articulation of the Semantic Web vision, it has become the focus of research on building the next web. The philosophy behind the Semantic Web is the same as that behind the World-Wide Web: anyone can be an information producer or consume anyone else’s information.
There are specific problems associated with that:
?"One major difficulty is that, by its very nature, the Semantic Web is a large, uncensored system to which anyone may contribute.”
The problems presented by new technology usually have a technological solution to them, we just need to find it. Just one suggestion coming out of the IBM Almaden Research Center employs the measurement of personalized trust scores:
“One major difficulty is that, by its very nature, the Semantic Web is a large, uncensored system to which anyone may contribute. This raises the question of how much credence to give each source. We cannot expect each user to know the trustworthiness of each source, nor would we want to assign top-down or global credibility values due to the subjective nature of trust. We tackle this problem by employing a web of trust, in which each user maintains trusts in a small number of other users. We then compose these trusts into trust values for all other users. The result of our computation is not an agglomerate “trustworthiness” of each user. Instead, each user receives a personalized set of trusts, which may vary widely from person to person.”
The approach suggested uses path algebra to calculate the strength and quality of the connection between a small set of people on the web. The idea behind it is that when a statement is made and another statement is derived from it, the trustworthiness of the original statement can be calculated based upon the trust values of the originating source and the group that source belongs to and this then allows a further calculation of trustworthiness to be made to the derivative statement. Do this across the web and you end up with cases where you have sufficient information to make a direct judgement call on trustworthiness and cases where you have just enough information to make a probabilistic judgement call on trustworthiness.
Here’s what’s at stake, which makes this approach as exciting as it is important:
?
?
Already there are algorithms that are used to establish a personal trust score and determine the suitability of an individual for a loan. Already, an algorithm called Vital sits on the board of directors of deep Knowledge Ventures, a firm that focuses on drugs for age-related diseases. The algorithm looks at a range of data when making decisions - including financial information, clinical trials for particular drugs, intellectual property owned by the firm and previous funding.
This highlights an interesting problem we have not really had to address before: while we know that a machine can potentially do a specified job, tirelessly, without making mistakes and much faster than a human (which is why Vital sits on the board tasked with making very specific decisions), which are we predisposed to trusting: the algorithm or intuition?
The question brings into focus the way we use cognitive and affective trust to reach at decisions and the weight we are prepared to put on each part of trust in each instance. Of particular importance in this area is a study conducted of Uber drivers by researchers of the Human-Computer Interaction Institute, Carnegie Mellon University. Uber drivers basically interact with an app that is backed by a recommendation and reward algorithm. The algorithm determines specific driver ratings based on a specified set of statistics, determines rates and who will or will not get the chance to pick up a ride. In addition it provides informational support and evaluates the driver’s job performance.
?What is striking in that study is that the app interface becomes the contact point between a human and an algorithm and trust from the human develops in the same way that it would had it been a human-to-human contact. Namely:
?
?
What came across most clearly was that those drivers who were savvy enough to go online and acquire a little bit of knowledge of the strengths and limitations of the algorithm they worked with, through online forums, also reported the best work experience.
In the semantic web online businesses that want to create a trust-inspiring environment need to work harder at employing a specific set of steps:
?
?
The online and offline environments are converging through the use of mobile devices and the presence of reputation systems like Yelp and Trip Advisor that bridge the online and offline worlds and drive reputation-based trust in both.
Think of the relationship you want to build with your customers first and the outcome you want to achieve as a result and the technology second. The technology will evolve and change, again and again. The kind of relationships you establish with your customers and the way you work in your business shouldn’t.
Creating Trust Networks
We talk about social networks and digital platforms spring to mind like Facebook and Twitter, Google+ (now defunct) and LinkedIn, Instagram and Pinterest. They differ in functionality as well as demographics and even their intent and purpose are different so it is good to actually define what it is, exactly, that we mean with the words “social network”.
First coined by the social anthropologist, John Barnes in 1954, in the article "Class and Committees in a Norwegian Island Parish" a social networks is represents relationships and flows between people, groups, organizations, computers or other information/knowledge processing entities. The activity that occurs within a social network is usually expressed through the topological mathematics of graph theory where the words connectedness' and connectivity' may refer to properties of the distance between persons, the number of paths between them, whether there is a path at all, or the proportion of possible paths actually in existence.
Simplified a little a social network is a grouping of people drawn to it by a broadly common purpose defined by the network’s intent and connected through several degrees of separation that are determined by how close their particular experiences and interests are.
Barnes, in what must have been a moment of epiphany, thought that the edges in a network represent interdependencies between nodes so that they could be:? interactions, invitations, trades, values, and so on. While social networks like Google+, Facebook and even MySpace are global in nature and massive in size but still work according to small-group dynamics where people make personal connections in them based upon their sense of affinity for other members of the network. But that is not the whole picture.
In the Figure above: The network graph formed by Wikipedia editors (edges) contributing to different Wikipedia language versions (vertices) during one month in summer 2013 (some rights reserved Wikipedia)?
Network connections exist because the network environment itself always has a trust mechanism in place. The social-network-based method of assessing trust uses social relationships to rank nodes (i.e. people) in the network. While this may be done in any number of ways within the network itself in order to help the network administrators determine the value of content and the focus of activity of network members, it also happens at the human level in an ad hoc way, through the observation of specific social proof elements like popularity of a person posting, recognition, familiarity with other members, perceived authority and so on.
Within a social network setting trust is a measure of confidence that an entity will behave in an expected manner, despite the lack of ability to monitor or control the environment in which it operates. Social networks usually provide some form of deterrent that includes reporting bad behavior to network administrators and/or blocking a person from having any other interaction with members who are find his behavior causing unnecessary friction.
There are two key concepts that anyone with an online business needs to keep in mind. The first is that clustering within a social network (grouping which can include communities, posts or groups) occurs along two distinct types:
?
A.????? Homophily – There are two distinct types of that status homophily, which is based on ascribed status such as race, ethnicity, age, religion, education, occupation, and so on, and value homophily which is based based on values, attitudes, and beliefs, that is, a tendency to associate with others who think in similar ways, regardless of differences in status. Depending on the type of social network or even the type of clustering that occurs within it, there can be an overlap in these two types of homophily so that status-based categorization may also bring people together who share similar ideas.
B.????? Small-world phenomenon - there are two different views on the “small world” phenomenon in social networks. Stanley Milgram, the original developer of the “small-world problem”, famously popularized the notion of there being “six degrees of separation” between any two people in the world. Later experiments involving computer networks later investigated the “small-world problem” in the modern computerized world, where small world networks were assumed to be everywhere (e.g., online social networks; email networks; networks of movie stars, boards of directors, and scientists; neural networks; genetic regulatory networks, protein interaction networks, metabolic reaction networks; World Wide Web; food Webs; and so on). The view is that instead of being one highly connected small world, our world consists of many loosely connected and some disconnected small worlds.
In the Figure above: "Six degrees of separation" by Daniel' (User:Dannie-walker) - Own work. Licensed under CC BY-SA 3.0 via Commons - https://commons.wikimedia.org/wiki/File:Six_degrees_of_separation.svg#/media/File:Six_degrees_of_separation.svg ?
Homophily and the small-world phenomenon where the Friend-of-a-Friend (FOAF) concept occurs make it possible to generate trust quickly by association. The underlying assumption behind this concept is that the relation “friendship” is transitive. The foundation of every friendship is trust and in a social network setting trust is propagative rather than transitive, spreading along the connections between nodes (people), its quality and endurance dependent upon the strength of the edges (connections) that form between them.
Social networks render the problem of how a business can scale being personal, solvable. Suppose you are looking for a car to buy and you already have a general specification and price range in mind, you’re faced with the problem of now having to do a considerable amount of work to narrow down all the available choices to you and then convince yourself of the final decision you’ve made.
The natural shortcut here is to ask your friend network for a recommendation that will point you in the right direction. A car dealership that is successful at forming trustworthy relationships could benefit from a recommendation in this context, finding prospects and customers without having to resort to expensive advertising or hard-sell techniques.
One organization who has spent a lot of time looking at trust and online sales for automakers is DrivingSales. As Fig 1.3 shows, the influence of friends and family feature heavily amongst the factors that lead a prospect to a particular car dealership.
In the Figure above: Original research carried out by DrivingSales showed that car dealership recommendation by friends and family was the strongest of three group factors affecting a car buyer’s decision.?
When it comes to using social media to increase brand awareness traditional marketers balk at the investment of time and effort required because there is no direct return-on-the-investment (ROI) metric they can apply that will show a bottom line advantage.
Yet, social media engagement allows a conversation to take place where brand values and brand ethos can be shared without the added pressure of selling, which lead to brand trust which has both a brand equity and a bottom line in terms of market share impact. We live in an age where we don’t trust brands, companies or corporations just because they tell us to. We need to feel we can connect with them and understand them under our own actions, not their instigation. We need to feel that we like them because want to, not because their advertising slogans tell us how likeable they are.
Those companies that don’t get this, brands that don’t understand it and corporations that do not respect it are on a downward value spiral from which they may never recover.
Summary
Trust is also present in the digital environment. Social networks propagate trust so that brands, business and companies that enjoy a good reputation in one part of the network soon find themselves discovering new friends and new potential customers.
While there is no direct impact of social media brand-awareness and trust-building activity to a company’s sales, it is a fact that without trust nothing will happen. No sales, no contact, not even a phone call. Many companies fail to understand the need to build a trust bank. They still rely on brand equity the old-fashioned way forgetting that relationships take time and the older they are the more valuable and more robust they are.
Extract from: The Tribe That Discovered Trust: How trust is created lost and regained in commercial interactions (You can download a free PDF copy of the book, no-strings whatsoever, here .)