Experience Planner

On the final post of this series on reputation systems, I’ll talk about how online reputation can work as an incentive for participation. I’ll provide examples of online communities that use reputation elements to promote positive behavior and engagement. Let me start with a brief story.

Last year, I traveled to Tennessee for the first time to attend the IA Summit. As a full-time grad student, I couldn’t afford staying at the hotel that held the conference. So, I went to Yelp to look for reviews about cheaper places to stay in Memphis. I ended up finding a hostel in Midtown, which had positive ratings and a reasonable price. It turned out that the Pilgrim House was exactly what I was looking for: cheap, neat and well-located. After the trip, I left a review for it on Yelp. I also sent a “Thank you” compliment to the reviewer who convinced me to stay there.

Yelp is a catalog of local businesses, designed to help people find the best services. There are over seven million reviews on the site. Someone who is looking for a hotel or a restaurant, for example, can access Yelp to look for tips provided by the community. Users can assign ratings to the reviews and they can also offer compliments to the reviewers. This is a way to stimulate participation from both sides: the reviewer, who is encouraged to keep the good quality of the reviews; and also the reader, who wants to show gratitude for finding good information. When you’re navigating profile pages on Yelp, you can see the compliments each member has received.

Communities, such as Yelp, use reputation to promote positive engagement. Reputation can be used for multiples purposes, not only to narrow choices ─ as on Amazon and BookMooch ─ but also to increase participation and to incentivize certain types of behavior. As mentioned on Reputation Systems are Everywhere, we should think of online reputation as a cycle: “quality contributions attract more attention, which begets more reward, which inspires more quality contributions”.

Here’s another interesting example: Yahoo Answers. Yahoo Answers is a Q&A site. Any user who signs up for an account can ask and/or answer questions. Users can build their reputation by accumulating points. In order to get points, they have to perform different tasks on the site. The most rewarding way to participate is by providing a good answer, which could be selected as the best one (user gets more points). Yahoo Answers benefit from reputation because as users compete trying to provide the best answer, the site naturally filters the good from the ordinary content. Similarly, users benefit from the status and authority they get by building a positive reputation with points.

This post ends the series on reputation systems. The key takeways this time are: (a) reputation can promote positive behavior and engagement within communities and (b) there are multiple types of reputation systems; some of them are more collaborative (Yelp, compliments), while others are more competitive (Yahoo Answers, points). The challenge for designers is in finding the type of system that best fits each community.

One final tip: the book Building Web Reputation Systems is now available.

(This post is part of my research for the Information Architecture and Knowledge Management academic program at Kent State University).

If there’s an old book in your house and you don’t know what to do with it, here’s a tip: try using BookMooch. BookMooch is a non-profit site for book lovers who enjoy exchanging books. The site was built with a very simple idea. Users have profile pages where they can list the books that they are willing to give away. Anyone can sign up for an account and registration is free. Once you have an account, you are able to see the inventory of other members and mooch their books. Every time you mooch a book from somebody else, you “pay” with one point. And every time someone mooches a book from you, you get one point. Books travel from sender to moocher via regular mail.

Unlike the e-commerce sites mentioned on my previous post, BookMooch doesn’t have a central authority to control the transactions. The site is based exclusively on mutual trust. Therefore, it’s important to display clues about past mooches, so everyone can see who is playing fair and who is not. Nobody wants to use a point and mooch a book that will never arrive. Each user has a profile page that lists basic information on past activities, such as overall feedback score, books mooched and books given. Members can use that information to decide which members they should mooch from.

Reputation is a fundamental part of the experience on the site. Needless to say, BookMoochers with low feedback score are less likely to exchange books than those with higher scores. Other than score, the profile pages display additional clues about the community members. Their list of books given and mooched as well as their inventory and wishlist can indicate reading habits and preferences. Those records are open to any other user on the site, and they offer more information to assist BookMoochers during the transactions.

When a member puts Huckleberry Finn on her wishlist, she is not only increasing her chances of getting books. She is also letting everyone else know a little bit more about herself, as a Mark Twain reader. In the long run, the list of books given and mooched could serve as a way to cluster users with similar preferences. As noted by Bryce Glass on his post Reputation is Identity, on the Web “visible histories reveal shared affinities and allow users with common interests to find each other”. BookMooch doesn’t rely exclusively on user feedback to build reputation. The site also tracks its members’ regular activities and makes them visible to those who are interested.

This is an interesting approach for two reasons: (a) it takes advantage of the natural actions taken by users, instead of requiring extra efforts (e.g. filling out forms to describe preferences) and (b) it offers a granular perspective of each member, which is something particularly important for new users with no transaction history yet. Transparency is an important tool used on BookMooch and it helps BookMoochers make better decisions.

On my next post, I’ll go further and talk about reputation as an incentive for participation within online communities.

(This post is part of my research for the Information Architecture and Knowledge Management academic program at Kent State University).

A few years ago, I listened to a very interesting podcast called The Architecture of Participation. It was a panel discussion that took place at the Web 2.0 Conference in San Francisco, 2004. During the talk, the Chief Technology Officer from Amazon, Allan Vermeulen, described his team’s approach for using ratings and reviews on the site:

“So our choices [on Amazon] are: either we can hire a bunch of people who sort of know how to get things done and that can write a bunch of guides and so on, on the net; or what we can do is build a technology so that everybody else out there who actually know this stuff better than we do, can go ahead and write about it and build these guides and push them out to people.”

This podcast brings up a very important fact: most online reputation systems are built with user-generated content. Amazon does not rely on experts to assign ratings to its products. Instead, the site uses a platform that is capable of capturing users’ inputs in many ways. Ideally, Amazon’s reputation system becomes more effective as more users contribute to the site. By combining multiple and diverse opinions, Amazon is able to filter the entire catalog, narrowing the list of options for each customer. As I mentioned on my previous post, ratings-and-reviews can provide guidance and therefore, improve decision making.

It is important to note that just because reputation systems allow users to send feedback it does not mean that users will actually do it. In an interesting study about ecommerce, Jared Spool compared the number of reviews for the book Harry Potter 7 on different sites, including Amazon, Target, Walmart and Barnes & Noble. One month after the book’s release, Amazon had the largest number of reviews – 1805 -  in contrast with Target, which had the smallest number -  just 3. Both sites use the same platform to offer product recommendations, despite the huge difference in the amount of user feedback captured by each.

Even though ratings-and-reviews are becoming increasingly popular and expected, designers should remember that this functionality does not rely on technology alone. Users have to be motivated in order to participate, and each community fosters participation in a different way. As noted in the example above, some communities are more active than others. And because user feedback is the foundation of most reputation systems, an early challenge for sites like Amazon or Target is in creating engagement.

Still, there are alternative methods for capturing user feedback. On my next post, I’ll talk about a smart approach for assigning reputation values to entities.

(This post is part of my research for the Information Architecture and Knowledge Management academic program at Kent State University).

On my previous post, I provided an overview of online reputation systems with some examples. Here, I’ll explain why they are important.

As we know, online shopping can be very convenient. Sites such as Amazon offer an incredibly big variety of items. A search for “laptop” on Amazon retrieves more than 450,000 results. For any item, a customer can choose the best price, compare it with similar products and create bundles to get more for less. From one hand, this is a good thing and that’s what makes Amazon so attractive to customers. But it also poses a new problem: with so many options out there, how do I choose the best one?

Unlike a brick-and-mortar store, in which customers have the ability to see and sometimes even test the products before buying them, the Internet lacks some of the crucial elements a real shopping experience has. Whenever customers go to a store inside a shopping mall for example, they are able to make immediate judgments from their interaction with other people and the environment. Customers can hold the products, ask questions, evaluate if the staff is friendly, look around and see if the store is well-organized and observe the other people that buy things there.

All those physical cues are not present on the digital space. For that reason, e-commerce sites try to minimize the problem with good design, online assistance, discounts on shipping and so on. Reputation systems are also part of that strategy and they are built to facilitate the online shopping process. Amazon provides ratings and reviews for products. This is one way of helping customers overcome 2 problems: (a) the overwhelming number of available options for products and (b) the lack of personal interactions.

In a nutshell, reputation systems can improve decision making by aggregating user feedback and presenting it in a way that allows people to narrow their options. On my next post, I’ll talk about a little bit more about user feedback.

(This post is part of my research for the Information Architecture and Knowledge Management academic program at Kent State University).

Online reputation systems have become part of the experience of many websites, helping users make smarter decisions and stimulating them to participate on a regular basis. A popular example of online reputation is the star-rating system that tells you if a book is worth reading based on users’ ratings. Amazon, for instance, uses this feature to assist customers during their shopping process. Those who are unsure about whether to buy book A or book B can look at the number of stars for each book, and then make an educated guess on which one is the best.

There are different types of reputation systems, and some of them have people as entities that can build reputation. BookMooch – an online community that promotes book swapping – uses a feedback score to help users evaluate the reliability of their peers. On the other hand, Yelp – a website that displays reviews for services, such as hotels, bars and restaurants – lets its users give badges to those who provide useful information.

Even though online reputation systems can be very useful, they may also bring new challenges to the sites that use them. An early obstacle for any reputation system is in capturing user feedback. Just because a website allows users to provide ratings and reviews, it does not mean that users will actually do it. Another obstacle is in figuring out which reputation model is the most suitable for the product or service that is being offered. Because there is no universal reputation system that will work for all sites, the choice of which one to use depends on the type of community, the site’s purpose and the technology constraints of each project.

On my future posts, I’ll try to highlight the importance of online reputation in an information-overloaded space. I’ll also pinpoint the effects that may arise from interaction under such community-controlled rules. The mail goal here is to provide a quick report of issues to be considered when designing your reputation systems.

(This post is part of my research for the Information Architecture and Knowledge Management academic program at Kent State University).

As anyone who has tried already knows, creating community is hard. But Christina Wodtke recently explained why it’s so hard in way I thought was pretty enlightening. She refers to Lewin’s Equation, which is…

B=f(P,E)
Behavior is a function of a Person and his Environment.

Wodtke goes on to explain that, as websites turn over more and more control and content to users, we have less and less control over the environment. Therefore, we have less control over how people behave.

So I thought I would highlight a couple of unique, innovative ways to use our limited control of Environment to positively affect user behavior.

Example 1: Mail Goggles
How often have you written an email late at night, mindlessly clicked “Send” and then regretted the email the next morning? Maybe it was filled with typos. Maybe it said something horribly offensive. Either way, you wish somebody had pointed out that you weren’t thinking and should have just gone to bed.

Thanks to the new Gmail Labs program, that kind of mistake can be a thing of the past. Google engineer Jon Perlow developed an add-on for Gmail called Mail Goggles, which cleverly stops you from sending email while your brain is off.

When you send an email late at night, it asks you a series of math problems (difficulty can be adjusted) that must be correctly answered before the email can be sent. If your mind isn’t totally sharp, it prevents you from hitting that all-important Send button.

Example 2: Audio Playback of YouTube Comments
If you’re not already familiar with the web comic XKCD, then you need to get up to speed. It not only jokes about internet memes, it creates them. In fact, it sometimes has a profound impact on the web’s most influential sites.

In this case, an XKCD comic illustrated what life might be like if YouTube required everybody who writes a comment to hear their comment said back to them before it gets posted.

The people at YouTube liked the idea well enough that it’s now a live feature you can try out. It’s not required like the comic joked, but it is the first button you see after writing a comment. Even before “Post Comment”.

The Future
Do you think this is what the future of influencing user behavior looks like? Even before these examples came out, people commonly pointed to sites that use an informal tone to influence user behavior. Maybe the secret to influencing user behavior is presenting it in a way that’s fun, lighthearted, and which people choose to engage with.

Note: For more thoughts on building websites so that users choose to engage with them, check out the Marketing with Meaning blog, written by Bob Gilbreath, the Chief Marketing Strategist at Bridge Worldwide. He regularly discusses examples of how marketing can be so good that people choose to engage with it. Good stuff!

Q: What is a “widget”?

A: To understand widgets, we first need to consider the history of the word.

Originally, the word “widget” was used in business classes to describe a non-specific product. For example, business students might play a simulation game in which they were put into teams that sold “widgets”. It was their job to figure out how to differentiate their widget and market it to consumers.

More recently, the word “widget” has been adopted to describe any small application or tool, ranging from a blog badge to a Facebook application. This usage was originated by executives who were shown various small applications and tools and didn’t know how to describe them. So they used a nonsense word.*

The bottom line is that the term widget has been used to describe so many diverse things that it’s not really a useful word any more. If someone starts talking about widgets, or asks you to design one, the first thing you need to do is work with them to figure out what they really want.

*This is not actually true. To the best of my knowledge, the word “widget” was coined by the people who developed Konfabulator (now called Yahoo! Widgets), which was was one of original platforms for hosting small applications on a PC desktop. But I do think the term “widget” has been co-opted by executives and stretched to the point of uselessness.

Q: How do I help someone figure out what kind of widget they want?

A: There are three important attributes to every widget. If you define those three things, then you’ll be well on you’re way.

1. Purpose
2. Audience
3. Platform/Technology

Let’s look at each of those attributes in detail.

Purpose
It’s sometimes hard to know whether you should begin planning a widget by thinking about your audience or by thinking about the purpose of the widget (i.e. the business objective of the widget). My feeling in this case is that you need to decide on the type of activity you want to encourage in users before you go too far down the path of figuring out what your users want. But purpose and audience really go hand in hand, so you need to keep both in mind as you work.

Based on my own observation of widgets, I am proposing 4 high level purposes that a widget may have.

1. Perform a task
2. Provide information
3. Gather information
4. Connect people socially

Note: I will add a page of example widgets that demonstrate each of these purposes in the near future.

Actually, a widget might fulfill several of these purposes at the same time. The key is to ask questions that help determine which ones are relevant.

Example Questions

• Can you do something useful for your audience, like perform some calculations or sound an alarm at an appropriate time?
  Then maybe your widget is meant to perform a task.
• Do you have content that you want to push out at relevant times?
  Consider a widget that provides information.
• Do your users want to provide feedback or give you information?
  Maybe you want to design a widget that gathers information.
• Would your users find it valuable to be connected with other users? Do they have something to say to each other, or do they want to compare themselves against each other?
  Consider a widget to connect them socially.

Audience
Some widgets have only one audience and others have multiple audiences. It’s your job to figure out who the audiences are.

Possible Audiences

• The user who installed the widget
  In most cases, the widget is going to be used by the person who installs it. For example, if I add a widget to my iGoogle page, it’s because I want to see it every time I go to my home page. The widget should be designed solely to meet my needs.
• The friends of the user who installed the widget
  Many bloggers add widgets to the sidebar of their blogs. For example, on this blog I might add a widget that searches the email archives at IxDA. As the blog writer, I already know about IxDA, and I already know how to go there and search the email archives. I would only add it to my blog as a service to my readers. They are the real audience.
• Both the user and her friends
  A lot of Facebook application fit in this category. They provide some value to the user, but they also provide value to the user’s friends who visit her profile page. For example, a Facebook application might be useful to me because it helps me keep track my favorite recipes. But it’s also useful to my friends because they can see the kinds of recipes I like, and recommend more they think I would also like.

Once you have identified your audience(s), then you can start designing an appropriate UI. If your audience includes both the user and that user’s friends, then you want to enable both audiences to do the things they want to do, without making the interface confusing for either audience.

Platform/Technology
Platform is the third and final attribute for a reason. It should flow naturally from the first two attributes.

For example, let’s say you chose to design the following kind of widget…
Purpose: Perform a function
Audience: Only the user who installs the widget

In that case, you probably want to choose a platform like iGoogle or the PC desktop.

However, if the chose to design a widget more like this…
Purpose: Connect people socially
Audience: The user and her friends

In that case, your widget probably belongs on a platform like Facebook.

Since there are so many platforms, and their pros and cons aren’t well known, I’ll provide a brief overview of the most popular options here.

• Yahoo Widgets
  This is a framework that users download and install on their PC. The framework is for both Mac and PC, which is good, but it does require a separate installation process. So I tend to think this platform is only for people who really want widgets — not for average users.
• Google Gadgets
  Google Gadgets are a lot like Yahoo Widgets, except they can work in more places. A Google Gadget can be placed on a user’s iGoogle page, it can be installed as a PC desktop widget using Google Desktop (which is just as distracting as installing Yahoo Widgets). And I understand a Google Gadget can even be installed on any regular web page.
• Facebook
  Facebook is probably the most popular social network for widgets (Facebook calls them “applications”). Since so many people are already using Facebook, it’s almost ideal for any widget that needs to be social. The downside is that Facebook has been overrun by applications, so the managers are taking steps to minimize the role the applications play. Also, Facebook applications are often written in a proprietary language, making them more difficult to port to other platforms.
• OpenSocial
  Just about every social network that isn’t Facebook uses a Google technology called OpenSocial to power its applications. That includes MySpace, LinkedIn, Ning, and dozens of others. Although I haven’t seen many OpenSocial applications yet, it’s a good bet that they’ll become more popular in the future.
• Blog badges
  Most blogs give their owners a place to add “badges”, usually along the left or right hand side of the page. These badges are typically just a snippet of HTML and an image, offering minimal interactivity.
• PC Desktop
  Some PC widgets are actually just specialized applications. A common example is WeatherBug for the PC. It gets installed like any other PC application; it just has a small window and it hides in the Start bar when not in use. Alternately, you could develop a cross-platform desktop widget by compiling a Flash application, or transforming your Flash program into an Adobe AIR application, or by writing your application in any other platform-agnostic language.
• Mac OS X Dashboard
  The Mac operating system comes with a tool called Dashboard that overlays a set of widgets whenever the user presses a special key. Dashboard widgets are generally very popular with Mac users, but it does require you to build a widget just for the Mac platform.
• Vista Sidebar
  In Windows Vista, a tool called Sidebar provides a “dock” for widgets on the side of the screen. It’s actually very similar to Yahoo! Widgets, except it’s built into the operating system. However, Vista Sidebar has the same drawback as Dashboard: it only works for a single operating system.
• iPhone
  The Apple iPhone now allows developers to write 3rd party applications for the phone. Because the screen is small, these applications look like widgets, so they’ve been wrapped in under that term too.

This is not a complete list of widget platforms, it’s just the most popular options. All the same, it’s more than enough options. The good news is that many of these platforms use the same basic technologies (HTML, CSS, Javascript, Flash, etc.), so it’s usually not that hard to rebuild a widget so it works on multiple platforms. Once you’ve picked a platform (or a few platforms) that make sense for your purpose and audience, you need to talk with your developers to figure out which ones you can reasonably support within your budget and schedule.

So what do you think of this framework? So far it has made sense for my projects, but I’d love to hear some other opinions. How are you thinking about the widgets you’re designing?

So, I’ve been thinking a lot about leveraging of user data recently.
Actually, to be more specific, I’ve been thinking about how to add value and one possible method is to leverage user data (data you already have).

When I say user data I’m thinking about three things:

1. User Generated Data
2. User Imported Data
3. Captured Data (i.e. user account data, analytics, tracking, etc.)

There are probably better categories out there but this entire idea needs to be fleshed out more anyway…

Anywho,

Let’s start trying to use our data to provide some added value. Finding new ways to use data you already have can breathe life into an older digital ecosystem or it might allow one to build a whole new system (stand alone or complementary).

Maybe you already have what you need to make a good site into a GREAT one. Maybe, you just have to look and start planning for it. I’m thinking that talking to your measurement/research team is the logical first step.

I’m thinking that the backend has to do all of the heavy lifting, if the user needs to do most of the work then the new feature, the new site, etc. will not be sustainable in the long run.

Maybe I shouldn’t say leveraging user data, maybe it’s leveraging user input
Or
Maybe it’s more about add value and the different ways to do that

Still, i’m just thinking “out loud” at the moment

One caveat:
Emergent Properties

Examples:

Take Last.FM for example, for the SXSW conference they created a system that took their user’s music preferences and allowed them to filter through the SXSW music lineup. Last.FM also rolled out an “event calendar” of sorts that looks for music in your area that meets your musical tastes (as well as it can). Those are great value adds for an already strong site.

Group Recipes – food/cooking/recipe community
Different browse options (e.g. beauty, intereresting, stumbler, local; eateries), Roger the recipe robot, friend “recommender” – all based on the community’s input

Ravelry – knit and crochet community
Popular patterns, project finder (based on type of yarn and other attributes), flickr integration, public project tracking (what are people working on), project and resource tracking (e.g. needles, hooks, and yarn), pattern browser – all based on user behaviour and input

Facebook
Facebook introduced the ability to import friends from a myriad of different web applications and address books – is it a good thing? That’s another post but it did leverage User Data (albeit the fact that the data came from another site/application) – user initiated import

Etsy
Search by colour – based on user tagging

There was a review site a few months ago that had a really interesting navigation system. You could browse by a taxonomy and then refine results based on a folksonomy. Seemed like a great idea. I need to find that site again!

During today’s Design and Architecture of Social Web Experiences workshop, I took 5 pages of notes and designed a very simple, yet very cool social website. So, yeah, I’d say it was a good session.

Here are some of the highlights from my perspective…

The Webb/Butterfield/Smith Model

This is an illustration that shows 7 aspects of social networks in a way that makes it easy to describe the functionality of a social web site.

Open design patterns

Trust and monitor

The problem of the Cold Start

Group size

I arrived in Miami, FL late last night for the IA Summit, and right now I’m sitting in the first all-day pre-conference session called “Design and Architecture of Social Web Experiences“.

So far, this session has provided a grounding in what social networks are, and we’ve had our first “workshop” of the day. Everyone has been assigned a different website to consider how social experiences might be integrated. My group is considering how to design a social web experience for a non profit that helps children with cancer. First step: figure out the business goals, user goals, and overall strategy.

Most of all, I’ve really enjoyed meeting the people here. Really smart people. And people who understand what I do! What a rare thing.

I’ll update later with more thoughts from the session.