#innovationneversleeps

Whether it’s the number of stars on your eBay profile, connections you have on LinkedIn, your Klout score, or your response rate on AirBnB: there are a number of different ways to quantify reputation value and create trust online.

Why is quantifying reputation important? It helps signal trust in reputation, and serves as data that can:

1. Be compared against other users
2. Provide peace of mind by establishing a user’s history of positive experiences / feedback 
3. Establish a new reputation by leveraging existing relationships and networks 
4. Prevent fraud and harm to other users through transparency, aids in identifying bad-reputation individuals
5. Be easily interpreted  

Here’s a quick look at some of the general ways that online marketplaces and social networks quantify reputation data:

• Social Network: # of friends / followers / connections

• Feedback: # of positive reviews

• Activity: # of contributions / participations / transactions

• Credentials: quality/skill, novice or expert

(badges representing learning modules completed in Codecademy)

Increasingly, networks are relying on multiple measures of trust:

(AirBnB profiles display information on number of places traveled, number of guests hosted, number of friends, a badge ‘credential’, and number of reviews)

Networks where trust is highly valued are also combining their platform-specific data with information on reputation from other networks as well:

(AirBnB prompts users to connect Facebook, Twitter, and LinkedIn profiles for added trust verification)

A particularly exciting trend in online reputation, represented above, is the rise in importing information from outside networks to bolster reputation on another network (ex. attaching your Twitter account to your AirBnB profile). This adds new potential layers for trust, as the combination provides a more valuable authentification of identity and reputation. It also helps accelerate adoption of new marketplaces, where in the early days there are not enough users / transactions to provide a history, and thus the ability to important related social data brings in a level of trust.

There are caveats to quantifying online reputation, of course. Boiling an important and often complex and contextually-driven piece of information such as reputation into a single number has its drawbacks. The Klout Score, for example, has drawn criticism for creating a single score from 1-100 to represent ‘social clout’, what some consider a gross oversimplification of an individual’s social media influence.

Despite the challenges in how to best measure and present reputation data, simple numbers and icons such as badges will continue to play a necessary role in signaling trust online.

*Note: this post is published as part of a series, ‘Online Reputation’, inspired by Harvard’s CS105: Privacy & Technology & taught by Prof. Jim Waldo (you can find his writings here).

Online reputation plays a crucial role in some of the biggest technology companies founded over the two decades, many of which are marketplaces. In 1995, eBay! launched it’s online auction marketplace, facilitating transactions online for new and used physical goods: today it’s the largest online marketplace with over 100M users and $68B of goods sold annually. Companies like eBay!, AirBnB, Etsy, Skillshare, Rent The Runway, Taskrabbit all rely on a critical piece of data: trust, most often in the form of user reputation.

In order for these marketplaces to facilitate transactions between users efficiently and effectively, user profiles representing reputation data is necessary to help establish trust amongst users and within the marketplace. To be able to rent out your home to a stranger, send money to Russia for a laptop, or take a class on web development from someone with a college degree, some missing information has to be fulfilled: trust in a user/marketplace’s quality and ability to satisfactorily provide what it says it will. The result of these various online reputation profiles almost creates what might be thought of as an ‘online community’, one in which are online behaviors and actions are used as the basis of our individual profiles.

My friend David Haber (@dhaber) thinks about this stuff all the time. He recently published an amazing post on Web Marketplaces in which he identifies some of the most important characteristics in evaluating the success of online marketplaces, namely:

1. Size of the market
2. Excess Capacity
3. Friction/Opacity
4. Fragmentation
5. Customer Experience

David provides a great snapshot of the dynamics facing marketplace technologies. For an even more in-depth analysis of the way these type of platforms evolve and scale over time, check out Prof. Tom Eisenmann’s working paper with HBS colleagues: Platform Envelopment.

But shying away from a more detailed analysis of marketplaces themselves here, we can still learn more about online marketplace characteristics by adding another category to consider:

      6.  Trust: the ability to measure, maintain and access online reputation profiles

The creation of online reputations for the consumers (and producers) in online marketplaces is essential for providing the fundamental trust necessary for individuals to transact and form relationships with other users and businesses online.

In the context of the categories above, online reputation is a key underpin for both 3) Friction/Opacity & 5) Customer experience. For friction/opacity, this deals with the information that is readily available (or not) in the marketplace, and a valuable piece of this information often deals directly with reputation, e.g. for reliability and accessibility. On AirBnB for example, the response rate (i.e. frequency of responding to messages from other users) is prominently displayed for users looking to rent out their homes. The response rate is one of AirBnB’s most valued piece of data, as it represents how active, attentive and engaged their hosts are in interfacing with potential guests: a potential friction point if hosts are slow/disengaged from future guests. High response rates means fast and fluid communication between users, a strong value for AirBnB as a marketplace for facilitating connections, and a well-monitored metric for their business.

Online reputation is also an essential component of 5) Customer experience. Looking again to AirBnB as an example, we see reputation manifest itself in customer experience via user testimonials (i.e. the quantity and content of reviews from individuals that have stayed with you). Having a reputation for being an amazing host on AirBnB means a large quantity of positive testimonials from guests that have stayed with you. This data acts as a signal (an important sociological concept to explore in a future post) of trust, inviting in potential guests with information around the safety and quality of the relationship. This applies to other marketplaces as well: the number of students and quantity of positive reviews for a teacher on Skillshare indicates what type of experience a prospective student can expect, and the number of feedback stars of a Taskrabbit provides security around the history of the errands they’ve run successfully. Ensuring great customer experience is a tenant of marketplace companies, particularly at the earliest stage when just starting: it’s hard to operate without a liquidity of data points that speak to the quality and effectiveness of what your platform promises to offer.

Online reputation acts as a digital analog for a concept we’re very familiar with in our real-world day-to-day, as trust is a powerful piece of information. It may be that it’s easier for us to take for granted basic trust around identity and security, but online where there is a computer and internet separating you from other users and companies: the need for trust and reputation online becomes glaringly necessary in the absence of in-person information and assurance. 

Next post: How online reputation also acts as the basis of online social networks such as Facebook, Quora, and LinkedIn. The role trust plays in these communities is also critically important, and worth it’s own discussion given the greater focus on social relationships rather than transactions between users.

*Note: this post is published as part of a series, ‘Online Reputation’, inspired by Harvard’s CS105: Privacy & Technology & taught by Prof. Jim Waldo (you can find his writings here).

I want to take a moment to step-back & appreciate how lucky I am to have my professors this semester. Running counter to much of my college academic career (sadly), this fall I strove to achieve nearly-perfect attendance to get more out of this opportunity to learn from great thinkers and doers.

College isn’t for everyone, a realization that is thankfully coming further into the light. But that said, there are certainly merits to education, even traditional education, and it’s worth appreciating this perspective alongside the rise of new learning models and formats.

The assignments and grades for these classes don’t really matter: it’s the teachings and personalities of my professors that will leave their imprint on my thinking and perspective from here on out. And for that I am grateful.

Roland Fryer | Robert M. Beren Professor of Economics

Faculty Director, The Education Innovation Laboratory at Harvard 

Class | Economics 1820: Education Reform in America

Jim Waldo | Gordon McKay Professor of Computer Science

Chief Technology Officer, Harvard University

Class | CS105: Privacy and Technology

H.T. Kung | William H. Gates Professor of Computer Science and Electrical Engineering

76th-Generation Direct Descendant of Confucius 

Class | CS143: Computer Networks

David Jones, MD, PhD | A. Bernard Ackerman Professor of the Culture of Medicine

Ackerman Program fostering collaborations in the medical humanities and social sciences across Faculty of Arts and Sciences and the Faculty of Medicine

Class: ER33: Medical Ethics and History

Suzanne Blier | Allen Whitehill Clowes Professor of Fine Arts and Professor of African and African American Studies

Co-Chair of Electronic Geo-Spatial Database AfricaMap, and Chair of the Steering Committee of Worldmap

Class | HAA1: Landmarks of World Art and Architecture

Follow-up post: learnings from each professor once finals are finished!