adamyntium

Category: google

  • Recommendations I gave to the orkut team

    [sorry for the messed up formatting; I copied this over from MS Word!]

    This is a letter I wrote to the Orkut team in August 2004, at the request of a Googler.  I was not a Googler until two years later.  May still be some weird formatting issues due to transferring it over from Word to here.
    *  *  *

    Introduction
    Hi [name],
    Well, it’s taken me longer to get around to this than I expected, and I hope that you’re still able to find my orkut comments useful. I don’t have the time to go into all topics about orkut that interest me, nor can I afford to go into the depth on each topic that I’d like at this point. If you’d like more specific input on particular issues or features, we can talk about that after you read my initial commentary below.
    And one more thing before I get started: I figure I ought to give you an idea of my familiarity with and interest in the social networking space:
    • I’ve been involved with online communities since I was 15.
    • I helped out the Prodigy Online Service a great deal, spearheading the creation of a new forum, beta testing their offline board reader, and providing lots of UI feedback.
    • I was a forum leader for two different areas of AOL.
    • I’ve created, moderated, or significantly contributed to many Web forums.
    • I’ve beta tested quite a few social networking services: Friendster, Tribe.net, Linked In, Ryze, Spoke, Flickr, Multiply, Dodgeball, and others.
    Here’s what I’ll be tackling in this note:
    • Grand overview (‘vision’) stuff
    • General UI suggestions / concerns
    …and not comprehensively tackling:
    • Stuff that’s already been widely harped on (server speed, not enough search options, etc.)
    • Things I’ve already discussed in the Orkut community I created.
    Okay, and with no further ado, I’ll get on with the show here.

    The foundation of a trusted network.

    Orkut advertises itself as a trusted network, and I believe that a trusted network must include all of the following:
    1. Clearly articulated rules and goals
    2. A collaboratively based structure
    3. Useful delineations and appropriate disclosure of ties

    Clearly articulated rules and goals

    In order for people to be able to assess and behave according to the trustworthiness of others on a service, they must first understand the actual service’s rules and goals. Orkut has already taken steps in this direction by articulating that the service is to be, ideally, one that encourages communications amongst actual people, not ‘characters.’ But beyond this, orkut hasn’t really expressed what the service is about, who it’s designed for (‘everyone’ isn’t, in itself, a useful answer), and what its goals are. This ambiguity (accentuated by a not-very-informative front login page) is discouraging to prospective users and also makes current users less apt to understand and trust the network.

    A collaboratively based structure

    Smart Filtering to better interpret and adapt to members’ interests

    Right now, people are thrown in ‘orkut jail’ when they ‘misbehave,’ either judged and sentenced by automated processes, by orkut administrators, or by anunsegmented group of peers. This suffers from lack of appropriateness (computers are only so smart), scalability issues, and untargeted group-think problems respectively. While I understand that a Slashdot karma system has its own flaws, I think that a system based upon a mixture of this and the user-moderation in Craig’s List is an appropriate solution for orkut.
    Over time, such moderation could affect (limit / throttle) privileges in the area of messaging, posting, scrapping, and so on. Eventually, it could also contribute to a scheme whereby communications and search results are prioritized based upon similar-group preferences… that is to say, what I see would be sorted and/or filtered based upon the preferences and behaviors of those I’ve linked to, what and who I’ve modded, and what those I’m strongly linked to have viewed, rated, and modded.
    For instance, let’s say I get yet another FoaF message in Portuguese (ack!). While I consider it ‘not interesting,’ Portuguese speakers might find it to be really funny or useful. Therefore, based upon user/group past behaviors (deleting, replying, etc.), the absolute messaging privileges of the sender should not be revoked, but rather, I shouldn’t see messages from that person (or his/her associated group, perhaps).

    About the current karma system

    I think the trusty, sexy, cool evaluation system should be scrapped. Here’s why:
    • It’s cheesy & a put-off for those who want to use orkut for business networking, or for those who are over 25, married, etc.
    • The concept of ‘sexiness’ is simply too variable amongst different groups of people. Even within my close group of friends, there’s a wide range of disagreement regarding who (or what) is ‘sexy.’ Sexiness is interpreted in completely different ways by age, interest, status, and so on.
    • One’s perception of another’s ‘trustiness’ really depends upon their interactions. I’m going to have a much different take on someone’s trustiness if I’m their business partner than if I’m their gym buddy.
    So here’s my admittedly somewhat-radical idea: Scrap the trusty/sexy/cool thing, and instead implement a system of connection strength metrics, tie descriptors, and enhanced testimonials.

    Connection Strength and Type ‘ Delineating and disclosing ties appropriately

    On ‘friends’

    First of all, change the label of ‘friend’ to contact. I know it may seem impersonal, but overall, it’s a better description of current and future connections. One’s boss is not one’s friend, typically, but many might want to link to their boss or subordinates or other colleagues on orkut. I like my family members, but it just seems weird to call my Grandpa or Mom a ‘friend.’
    And frankly, the word ‘friend’ is widely overused and abused. [Name] once invited me to link to him, and I did so gladly. I greatly respect him and was honored at what I felt he was implying. But ‘ though we had initially planned a lunch meeting ‘ [Name] and I have not yet met. We haven’t even exchanged more than a few quick e-mails. So I’d hardly consider him a friend… and I mean that in the most honest yet non-mean-spirited way. I’m sure you understand.
    Orkut (the service) should not be about (just) friendships. It’s about ties. Strong ties, weak ties… amongst contacts. Maybe you can come up with a better word. But let’s ditch the online ‘friends’ pretense. Like Friendster, it’s so 1999.

    All connections are not equal

    Imagine someone using orkut to find a date for an upcoming concert. And another person using orkut to hire a nanny for their daughter. Two drastically different interests and needs… but both greatly potentially assisted by a smart delineation and disclosure of member ties.
    In the first case, the member wants to see who his contacts know socially… and be able to assess a fit by looks, personality, musical tastes, and so on. Unfortunately, as orkut works right now, he can’t tell how his contacts are really connected to that hottie he spied while browsing. Sure, he shows that HisName –> Jane –> Fred –>ThatHottie… but how well does Fred really know ThatHottie? Does he know her from work, from clubbing, or is Fred just someone who ‘friends’ every ‘hot chick’ on orkut? And for that matter, how does Jane know Fred… and, more importantly, how well?
    The person seeking the nanny obviously has different needs. She looked up ‘Mary Smith’ on Orkut after seeing Mary post an ad on Craig’s List looking for nanny work, and she sees that she is connected to Mary by three degrees of separation. But, again, she has no idea how she is connected; do these folks know Mary from her previous nannying work, from the gym, or from dating, or…?
    Therefore, it’s critical that orkut offer the ability to specify AND view (either directly or indirectly) the strength and type of ties between people.
    Members should be invited to score the strength of their connection with each of their contacts numerically on two axes from 1-100: Professional and Social, with (again) boundaries placed or statistical adjustments made in this context so as to take into account those who give the max points to all their contacts and so on.
    Then, when viewing a network path, one might see this:

    Me –> Jane –> Fred –> Marta [briefcase icon] 87, [martini icon] 13
    You have strong connections to Marta socially, but weak ties to her professionally.

    In this example, Marta would be a more appropriate contact for the person seeking a Nanny; the potential employer would probably have a better chance learning about Marta’s professionalism than how well she holds her liquor.
    Beyond quantitative-tie listing, I think it’d be very useful for people to be able to simply note (in a small free-form field) how they are linked to someone, so that when I link to Jane, I could specify that she’s my piano teacher… and this information would be visible when someone moused-over the connecting arrow.
    How is this better than the current system of haven’t met, acquaintance, friend, good friend, and best friend? Let me count the ways!
    • The current system doesn’t fit nicely in a business context.
    • Haven’t met is interesting but orthogonal to the issue of strength-of-tie, and should not be part of this scale! For instance, I once corresponded regularly with a penpal for many years before meeting him, and I certainly considered him a friend (more than an acquaintance) well before I met up with him. This situation is especially common nowadays, with people having friendships (even quasi-romances) with people they’ve ‘met’ only on the Internet. On the flip side, I think it’s quite possible to have very weak ties with someone you’ve met.
    In contrast, my proposed system (with tie-strength and connection descriptor):
    • Works for ties with family, significant other, work colleagues, etc.
    • Doesn’t make a false distinction between those people we’ve met in person and those we haven’t.
    • Lets people accept link requests from others with the understanding that they can indicate their connection appropriately. For instance, right now I’ve noticed that many people link to others to suggest ‘I like you, even though we’re not yet friends.’ With my proposed system, the recipient could decide that, hey, this (requester) person looks cool even if don’t know them at all; I’ll accept their connection request and simply give them zeros for both professional and social tie strengths (to be later adjusted as appropriate).
    • Provides value by letting people understand and take into account connections (‘Hmm… this guy knows her from his college marching band, so maybe he could tell me if she’s got good chops for the new quintet I’m putting together.’)
    • Could be a super-valuable companion to messaging and profile-item-disclosure / privacy (I’ll talk more about these issues later).
    Understandably, there’s also a sticky point associated with my proposed quantitative system: the awkwardness in potentially learning (and knowing that others could possibly learn) how one is rated by friends. Therefore — while perhaps not foolproof  — one protective measure could be to avoid publicly listing connection strengths with single ties. That is to say, with Me <-> Betty, there’d be no indication of tie strengths.
    From a UI and functionality perspective, on the Your Connections (formerly ‘Friends’) screen I’d like to see each contact listed on a separate row with (and sortableby) the following info:
    • Photo
    • Name (underneath)
    • Link description: ‘gym partner’ or ‘college friend’ or ‘former colleague’ etc.
    • They updated: The date of the last time they updated their profile or photo gallery and what they most recently updated, e.g., ‘7/3/04 ‘ New job description’
    • Tie #s added/changed: The date when you added them to your list, or when you most recently changed your tie strength ratings for them. This would help (especially with sorting/filtering) members update such ratings over time by noticing, for instance, hmm… I haven’t updated his ratings recently, and we’ve become much closer friends in the last month.
    • Social tie strength: In the form of a simple 3-character text entry box, possibly augmented by a Flash-based slider.
    • Professional tie strength: Same as above.
    You’ll note that I have omitted the age, gender, and location in this listing. While I suppose it could be included in a different view, I think that I’m already aware of such information for most of my contacts. If I can’t recall something as basic as someone’s gender, why am I even linking to them? wink

    Who’s yer daddy?

    In an effort to help users ferret out fakesters and also add some transparent accountability to the system, I’d like for orkut to show ‘invite trees’ of at least two degrees. In other words, we should be able to click on a link and see ‘Tara Jones was invited by _Suzanne Johnson_, who was invited by _Matt Smith_.’ With this, people could have additional info-ammunition when there is a group of, well, jerks in a particular community; ‘Hmm… all of these twits were invited directly/indirectly by one guy!’ It’d also be useful when looking at testimonials: ‘Hmm… he only has testimonials from people he’s directly invited!’

    Defining how we describe and see each other

    Fandom

    Let’s tackle the most basic issue first, even though it’s something I think you’ve heard time and time again. The ‘fan’ concept as currently implemented is absolutely broken.
    First of all, I question the value of fan indications in general. What use (or even fun) is it for me to see that some ‘hot babe’ has lots of fans? Or that Orkut ‘ the creator of the service being used ‘ has lots of fans. No duh!
    But if, for some reason, this feature must be kept in, then at least make it work more like fandom in the real world. I’m a fan of Larry and Sergey. They don’t know me from Adam. This is not a connection in the way that Fans on orkut is currently implemented (e.g., that Larry and Sergey must first acknowledge my existence). This is merely a one-way indicator.
    In other words, members should be able to ‘fan’ anyone they want. The current ‘fan only if already friends’ is both useless and also harmful, since it often incents people to request connections from others with whom they really hold no (reciprocated) connection.
    On a somewhat related note, perhaps orkut should move to a system whereby anyone can link to anyone and reciprocation is noted but not required. This is something that definitely should be considered, but I don’t have the time to delve into it at the moment… sorry.

    Testimonials

    I’d like to be able to offer testimonials on any orkut member, not just my friends… though of course I think such testimonials should be displayed separately from ‘friend’ (or ‘connection’) testimonials.
    Additionally, I’d like to see testimonials contain two fields: the current main text box for a testimonial-type message, and a separate ‘adjectives’ box. After someone has accumulated more than [x] total testimonials (including descriptors), it’d be neat to display something like the following on their profile:
    Andrea has been frequently described in the following ways by others:
    • By those with whom she has strong professional ties [suitcase icon]: Decisive, funny, smart
    • By those with whom she has strong social or family ties [martini icon]: Loving, smart, athletic
    • By others: Active, sexy, driven
    An aggregate-variety of this descriptor field would also be neat to show in orkut search results.

    Profiles and Privacy

    Do folks want their business colleagues knowing their turn-ons?
    Do I want friends-of-friends to know about my career (‘Wow, he must be rich!’ or ‘He does that for a living?!’)
    I’d like to see more fine-grained user-controls in the profile sections… perhaps based upon strength-of-connection levels and/or connection degrees. For instance, it might be useful to say:
    • Show this field only to strong social connections, limited by two degrees
    • Show these fields only to weak social and professional connections, not limited by degree
    For instance, I might want my sexual orientation viewable only to people with weak professional connections AND strong social connections, but not limited by degree. Or I might want my entire profile viewable only to those within three degrees.
    I think this sort of control would enable folks to worry less about bleed-through of social and professional information, since many of us do like to keep our work and social ‘lives’ separate.

    Navigation

    The current navigation is neither intuitive nor consistent.
    Here’s what I’d change:
    • Move News. This should be under Help, and any new announcement should be listed on the front page. News archives could be under Help. After all, why shouldNews take up part of the coveted top navigation UI when it’s so seldom useful navigationally (news is updated infrequently, and users aren’t notified when there are updates anyway).
    • Don’t remove options willy-nilly! When I’m viewing my profile, I should still have access to the regular options.
    • Don’t switch navigation/edit options from side to side. When I’m editing my profile, the profile-segment links shouldn’t be moved suddenly to the right-hand side.
    • Overall, be consistent! You already have section links at the top and action links along the lefthand side for most of the time ‘ don’t veer from this.

    Private messaging

    Filtering incoming messages

    Messaging issues are the bane of orkut’s existence, to the point where many of my friends are now putting ‘orkut’ in the same mental category as ‘spam.’
    Personally, I’d say ‘ without exaggeration ‘ that about 95+% of the messages I receive via orkut are completely uninteresting to me. I’d turn off FoaF and community messaging entirely in my preferences (as I’m sure countless others have done), but it’s that 5% that grudgingly keeps me from doing this.
    The solution, I think, lies in a mixture or selection amongst Bayesian filtering, collaborative filtering, strength-of-tie factors, and other user-specified restrictions.

    Bayesian filtering

    From an implementation perspective, Bayesian filtering could be potentially handled via thumbs-up / thumbs-down buttons in the orkut mail interface… but this admittedly wouldn’t work for people viewing orkut messages in their regular e-mail.

    Strength-of-tie factoring

    Ideally, users should be able to specify a ‘strength-of-tie’ threshold for receiving FoaF messages. For instance:
    I’d like to receive Friend of a Friend messages only from connections with weighted/aggregate strength-of-tie scores in the following categories:
    [checkbox] Social connection stronger than ___ on 100 point scale
    [checkbox] Professional connection stronger than ___ on a 100 point scale.
    This way, people could opt to receive messages, say, only from people who are strongly connected to them socially, or from people with weak social ties but strong business ties, and so on.

    Other user-specified restrictions

    This could be in the form of ‘Send me 2nd degree but not 3rd degree FoaF messages’ or ‘Send me messages from females aged 20-30’ or ‘Send me messages from people who are in my communities and are 2nd degree contacts’ and so on. This has the potential to get quite complicated, so perhaps these ‘filters’ could be part of an ‘advanced’ user interface option.

    Limiting message sending

    Initially I thought that perhaps there should be some limitations or throttling on orkut messaging so, say, Jeff Bezos doesn’t get 50,000 messages from excited fans or disgruntled shoppers. But then I realized that this issue is indeed better handled from the receiving end. After all, why should I be limited from writing 7th degree contacts when some of those folks may want to hear from everybody… or may want to filter on criteria other than degrees-of-separation and so on. In other words, aside from extreme cases requiring throttling (someone shouldn’t be allowed to send more than 200 messages in an hour!), I think orkut should let messaging restrictions be handled solely by recipients (with admittedly perhaps some default filters in place).
    With that said, I think there should be some awareness involved when people are sending group messages, such as an alert like: ‘WARNING! This message will be sent to as many as 3,719 people. Please consider whether it would be more appropriate to contact a smaller group of contacts individually, or post a note on one of the community forums.’

    Showing messages

    I think that you should show orkut messages only in orkut, and not allow for the receipt of actual messages within one’s personal e-mail for the following reasons:
    • This thwarts the privacy concern that stems from people accidentally writing people back without realizing that their reply will disclose their identity.
    • By placing all communications-reading within orkut, you can amass data on clickthrus, immediate-deletes, and so on. This, in turn, will contribute to user-based and group-based filtering initiatives. For instance, if 98% of recipients delete John Smith’s mail without reading it, then it’s likely he’s sending unwanted communications. Additionally, showing all mail headers in orkut (rather than an external reader) allows orkut to show clickable thumbs-up / thumbs-down icons, further facilitating collaborative filtering.

    Communities

    Grouping

    One’s communities quickly become unwieldy, at least for those of us with more than 30-40 communities. Therefore, user-grouping of communities is something that’d be quite useful, especially given the prevalence of communities-as-badge (‘I want people to know I am affiliated with or love [x], even though I don’t plan on writing a lot of stuff about it’).
    For instance, it’d be great if we could group according to anything we want… topics (‘Geeky stuff’) apart from orkut’s default categorizations, level of interest (‘Favorite groups,’ ‘Not favs‘), frequency of desired access (‘Check weekly,’ ‘Check daily.’) and so on. From there, we could also set permission and notification preferences (‘Tell me anytime someone posts something in these communities,’ ‘Make my posts in these communities not accessible by the main [future] orkut Community Search tool’) and so on. Also ‘Show info on these communities on my orkut home page’ and ‘Include / exclude these communities from the list on my profile’ (useful for memberships in personal/sensitive-topic communities, such as spousal abuse discussions, etc.).

    Functionality beyond forums

    I understand that there’s likely some tension between orkut and GG2 in terms of overlap, future functionality, and so on. However, GG2 aside, I’d love to see polls, file archives, photo galleries, and bookmarks (orkut-internal and also Web-external) in the communities.

    Moderation and privacy

    Moderators should be able to mark forums as only readable by members, mark forums as inaccessible by orkut community title and/or message content searches, and should also be able to permanently ban (not just kick) users even from public communities.
    Additionally, moderators should be able to appoint co-moderators and/or temporary (vacation) moderators.

    The forums

    I’d like to see the forums support basic HTML (my fingers are trained to type not ) and/or actual BBcode (I’m so used to type either href…> or [url=] and not [link]!) >

    I think you know the rest in terms of what needs to be done forum-functionality wise. Look at vBulletin & phpBB, then copy what they do. Simple but true, excepting the issue of karma / collaborative filtering discussed earlier here.
    Or if you really want to get fancy, introduce Gmail-style conversational expand/collapse functionality. Yum!
    At minimum, you need to include read/unread awareness within the forums and in the community listing overall. And yes, I realize this (like many of my other suggestions) is resource intensive.

    Notifications

    I don’t care whether it’s e-mail, RSS, a separate page, etc. People need to know when folks have replied to their threads, period.

    Presence detection and live communications

    One of the most common complaints I’ve heard about orkut is that there’s ‘nothing to do.’ While I’m sure that more folks would use the boards more if certain forum-issues were addressed, I’m also confident that the establishment of presence-awareness and live interaction features would greatly increase the ‘stickiness’ of the service. By ‘online’ below, I’m referring to a state of ‘active within past [x] minutes’ (with the threshold increased for those who most recently clicked on a ‘compose’ link).

    Detection

    Specifically, I would like to see the online status of:



    • My contacts and 2nd degree contacts
    • People in each community (from the community page)
    • From a search results page, both proactively and otherwise (‘Show me all people within 5 miles of my zip who are interested in hiking and are online now’)
    • Anyone whose profile I’m looking at.

    This would help make communities feel ‘active’ rather than simply Web pages to browse occasionally.

    One-to-one chat

    In addition to seeing who is online, I’d also like to be able to initiate communication with individuals. orkut could be a great test platform for GIM (Google Instant Messenger). In the meantime, orkut could facilitate one-to-one IM communications by hyperlinking one’s name throughout the service toYahooIM or AIM (click on the name, a Yahoo or AIM window pops up, assuming the user has the relevant client installed).

    Group chat

    Of equal importance, I think that it’d be amazingly useful and fun to be able to have a live chat box on the front page of every community. This would augment, rather than by a substitute for posting and reading messages in each forum.

    Chat implementations

    You could even kludge a chat implementation to begin with, since there are no worries about permanence (e.g., unlike with message forums in which message links should be permanent and accessible over time, chat is ‘ by its nature ‘ preferably transient). For instance, you could simply create andiframe a window into a logically named and described IRC room (‘orkut-[community_name]’ / ‘[description’). Of course, such communications would then not be limited to orkut members, though I suppose savvy moderators could simply password protect rooms to enhance privacy.


    Highlighting interesting data appropriately

    I was hugely dismayed at the original statistics and very pleased when they were taken down. They just encouraged an obnoxious feedback loop… making those who were orkut’ly popular more so, and pandering to why-do-we-need-orkut-for-this ‘hot chick’ instincts and behaviors.

    When showing statistics, the following should be taken into consideration:
    1)      Are these statistics useful or interesting? (obviously!)
    2)      Do they incent desirable behaviors rather than encouraging harmful or annoying behaviors?

    See http://www.orkut.com/CommMsgs.aspx?cmm=781&tid=93 for detailed ideas and commentary on interesting statistics to show.

    Providing greater control without sowing confusion


    I love the way the Google AdWords interface uses javascript/DHTML to show major options (specifically in the ‘custom report’ section) and yet allows for the tweaking of more ‘minor’ options via expand/collapse functionality. Orkut should do the same throughout the service… in the context of moderator settings and ‘ more importantly ‘ user settings.
    For instance, this UI would be nice for allowing users to either turn off all messages from communities or opt to select messages from specific communities or (future) groups of communities. In fact, this sort of expand/collapse UI would be wonderful for permissioning / searching / filtering on orkut in general!


    Miscellaneous UI and other stuff

    And now the moment you’ve been waiting for… where I actually talk about (mostly) small changes you could implement that would make the existing orkut service a lot more useful and pleasant. As noted in the intro, I’ll try not to harp on the obvious stuff that everyone has already (often justifiably) ragged on (e.g., optionally persistent logins, etc.)

    • Show more useful data in lists! For instance, on the my community list, there are little three-pronged icons to the left of every group. Why? This fancy but useless bullet point would be great to replace with an icon or text indicator of sorts… whether a group is moderated or not, whether I’ve accessed it in the last [x] days, the name of the last thread author, anything at all!

    • The current classify-a-friend widget is really horrible from a usability standpoint! Anything in this area would be an improvement… a dropdown menu, radio buttons, etc.

    • Allow users to indicate languages they speak. Correspondingly, have people note the language in any FoaF messages they send. And then… don’t send messages written in French to people who don’t speak French, and so on.

    • Use confirmations more consistently! When someone clicks on the delete link on a post, this request should be confirmed!

    • Include useful icons on more buttons… e.g., include an arrow for next page, a pen+paper for new topic, etc.

    • Don’t put critical links next to each other without padding… e.g., edit and delete! (icons here would also be useful)

    • Don’t allow identically-named communities. I guess it’s a bit late for this, but… ack! It’s just wrong for so many reasons.

    • Throughout the service, it’s important that you indicate change. The ‘news’ link is largely useless because there’s no reason to click on it. It should be bright red when there’s news a user hasn’t yet seen

    • The upcoming birthday thing should be reworked into ‘Upcoming events’ and should be minimizeable.

    • ‘Next steps’ should be smarter (I’ve already filled out my profile, darnit).

    • Ditch the dumb fortune. This could be replaced with much more relevant text (‘Bee Ling has recently added new pictures’ or ‘Your Chicago barsgroup needs a moderator. Interested?’ or ‘Did you know that on orkut you can…’ and so on)

    • Allow users to specify pagination (max records per page) that they like per section.

    • Show network connections like Kartoo does! J Or at least in a graphical format of some sort.

    • For the love of God, please move the location of the ‘Send message’ button. For an action that is appropriate only on rare occasions, it should notbe in the (highly visible) upper-left location! And while you’re at it, kindly add the warning above the text entry box, so people will see it before writing a message that’s bound to annoy several thousand people.

    • I’m especially fond of this thread, and (honestly!) not just because of the head-swelling compliment at the end.

    Discovery

    I definitely enjoy orkut’s new ‘related communities’ feature, and I also noticed that there were some personal recommendations in a recent orkut newsletter. I’d like to see more of this! Perhaps orkut could highlight a different person on one’s homepage at each logon, including their pic, info from a random field, and an explanation of the connection (‘Joanna is in many of the same communities you are.’ Or ‘Many of your friends are also friends with Joanna.’ etc.)


    The bottom line

    Here’s my harsh but heartfelt recommendation:
    Scrap the existing orkut, or at least the entire UI.

    Take what you’ve learned and apply it to a new system built from the ground up, structured and coded by a substantial team of UI folks, social scientists, experienced online community people, collaborative filtering geniuses, and so on. Make goals up front and communicate them internally to engineers and externally to members and potential members. Establish a team presence as a group of facilitators / moderators, not policemen. Actively solicit input and assistance from members constantly. Hire more customer service people; temps are cheap (I’d think).

    Google would then easily have the ability to envision and implement a truly outstanding trust-based social & professional networking service with the UI simple-elegance of Gmail and the far-reaching change-the-world-power of Google search.

  • Why I’ve given up on orkut.com

    Sometimes I’m wrong. I initially hoped for and expected great things from orkut.com, and sadly the service has not delivered. Worse yet, I feel that orkut.com is moving steadily in the wrong direction.

    It does pain me to write this, partly because I’ve had the pleasure of meeting the creator of the service — Orkut B. — in person. He’s clearly passionate about doing Good Things with his networking service, and he’s a very fun and engaging guy besides. Indeed, largely because of this, when I was asked by a Google friend of mine to submit a detailed list of my recommendations to the orkut team, I didn’t hesitate to do so.

    But after a relatively enthusiastic blog entry about orkut.com about a year ago and a history of strong participation on the service, it’s time for me to concentrate my interests and energies elsewhere.

    Seeing the implementation of Orkut Media was the last straw for me. While I had envisioned (and was excited about!) the idea of a Content Manager drawing from and encouraging ongoing participation within the orkut.com network, the orkut.com team instead unwisely decided to create something wholly different… something akin to a high school paper-based literary journal.

    Why was this such a bad decision?

    • It creates a relatively static Publishing (newspaper / book) environment, rather than facilitating a real-time community feel. Want to respond to a photo or an essay on the Orkut Media site? Send an e-mail to the editor. Isn’t that a bit 20th century? Even everyday blogs offer live commenting.

    • It results in an odd separation of community and content and a disincentive to contribute one’s best work to orkut’s communities. Pithy, funny, insightful essays written within communities won’t get highlighted unless the author affirmatively submits them to the Orkut Media editor. Wouldn’t it have been better for the Content Manager to serve as someone identifying interesting posts and quirky but not-yet-popular communities instead of creating a pre-approved island of public content?

    • More than half of the members of the service speak (Brazillian) Portuguese, but not a single initial Orkut Media entry is by a Brazillian or even tangentially related to Brazil. That’s simply inconsiderate, IMHO.

    With that said, perhaps the greatest disappointment isn’t what orkut.com has done with the new Orkut Media feature, but rather what the service has not done in general.

    In particular, the service has done little to put Trust in a supposedly trusted network.

    • Connection type and strength is not factored into any substantive part of the service. When looking at someone’s profile, you can’t tell if any of their ‘friends’ is a close connection, a distant acquaintance, a work colleague, a family member, or a spouse. And given the complete lack of any barriers or disincentives to adding anyone and everyone willy-nilly as a “friend”… the connections have become completely meaningless.

    • Members have no thoughtful or innovative tools to help them moderate their own communities, choose whose posts they’d like to see (or not see), rate others’ contributions (karma points), and so on. Instead, members are given the crude and often backfiring “bogus” button.

    • The system of friends and ‘fans’ is ridiculously obtuse and backwards overall. In real life, I am a fan of several prominent ‘famous people’ but hardly one of their friends. In most cases, they don’t even know I exist. So why can I not, then, mark them as a fan on orkut without bugging them with a ‘friend’ request? Other social networking services have devised such systems in much more effective, intuitive ways. At the most basic but still effective level, for instance, on several other services anyone can mark someone else as a friend; if it’s one way, then you’re a fan, and if it’s reciprocal, then you’re friends.

    Of course, on a more obvious level, the orkut.com UI has scarcely been improved since its inception. Navigation is inconsistent, one can re-assign friends to groups only one at a time (and each after a full screen redraw at that!), and so on.

    But the greatest shame, IMHO, is that the orkut.com team never made communication a priority. They didn’t set up a feedback forum, and very rarely posted notes acknowledging, much less thanking members for their contributions. Mass ‘jailings’, lost memberships, security exploits… generally brushed under the rug… exactly what beta programs are supposed to NOT do to win trust and engage members in improving a service.

    Indeed, instead of a Content Manager, orkut should have hired a Communications Manager, or — better yet — a Community Manager. Someone to guide new members, encourage and recognize feedback, offer insight into the decisions of orkut.com, and so on.

    I recognize that orkut.com was the “20%” project of a well-meaning Google engineer. But, as I brashly suggested in the document I sent to the orkut team, it’s clear that this engineer and/or Google should simply scrap the entire system and start over. Online Networking is too important to simply ignore or give a half-ass ‘fun’ treatment to. A Google-worthy effort deserves a substantive team of engineers, social scientists / sociologists, UI folks, community managers, customer service people, and so on. Whereas it’s fine to give ‘20%’ to a news index or most any other beta project… when you involve a few million people (and their sensitive data), it’s time to up the commitment. Here’s hoping Google eventually sees the light.

  • Too shy or lazy to sneak peeks in the bookstore? Here’s Google to the rescue.

    I know that Google Print — Google’s initiative to scan and place online full-text and searchable book pages — is not really new news.

    But tonight is the first time I recall actually seeing a book result within a regular Google search:

    One thing that’s both interesting and refreshing to note is that Google does not get a kickback from any book sales. As usual, it makes money solely off the clearly-marked paid ads on the right-hand side of its pages.

    (But sorry, folks, a search for Kama Sutra doesn’t yield any book hits).

  • Google Desktop Search — A review via an EXCLUSIVE interview!

    I was able to score an exclusive interview with Adam Lasnik, supreme geek, Google connoisseur, and Google Desktop Search expert, and I’m very pleased to offer the full transcript below.

    Adam, thanks for coming today. To start, why don’t you give us a quick overview of what Google Desktop Search (“GDS”) does?

    It’s delightful to be here!

    Well, GDS enables any personal and business user to search their computer’s hard drive much as they would search Google… typing in a search term using Google’s general search syntax and getting a results page in under one second.

    Specifically, GDS searches both the filenames and contents of the following: Outlook and Outlook Express e-mail, AOL instant messages, Internet Explorer (Web page history), text files, and also files from Microsoft Word, Excel, and Powerpoint.

    Do note that GDS only works on Windows XP and 2000 at this time.

    So is this all pretty easy to use, or is it a tool just for geeks?

    Perhaps unsurprisingly, GDS is quite user-friendly from end-to-end. It installs very quickly (it’s 400K) and it politely uninstalls just as easily. And actually using the tool is a snap: When you click on the GDS icon in your system tray, a browser window opens; you simply type in a search query, and BOOM, Google lists results are shown, formatted very similarly to regular Google searches.

    Whoa… regular Google searches… does this mean other people can search my hard drive? Or can Google see what I’m searching for or what’s on my hard drive? I better start removing those porn…er, confidential business plan documents!

    No, no… unlike with Google Web Search, the index of your files remains on your computer. While you have the OPTION to let Google learn about your search stats (not terms!), Google doesn’t peek into your hard drive contents or examine what you’re personal searching for. So if you’re looking for a big ass barbeque to grill a fluffy bunny, you don’t have to worry about Google notifying the ASPCA (I
    hear rabbit meat is tasty, though… but I digress). In a nutshell, Google’s GDS privacy policy should make you feel better.

    Okay, so GDS is easy to use and it’s not going to get me in trouble. So far so good. How about a few more techie details?

    You got it! First, let me say that — while GDS isn’t a power tool yet — it performs well on machines of power users like myself.

    I have a decent rig (2.4 ghz Pentium with 512 megs of RAM), but it’s loaded down with other file indexing programs, numerous Outlook plugins, more than a dozen apps in the systray, two routers (one wireless, one phone adaptor), a firewall (XP SP2-based), a virus checker (AVG), and many general programs running concurrently (Trillian, Dreamweaver, Outlook, and so on… and GDS has worked flawlessly. No install problems, and no noticeable slowdown of the system during indexing. Furthermore, searching with GDS is LIGHTNING FAST… both on my rig, and on the machine of a less-geeky friend of mine who’s sadly cursed an older machine. The downside of that, understandably, is that GDS is pretty conservative. It initially indexes quite slowly in the background (you’ll need to leave this puppy on overnight to get a full index!), and overall the product doesn’t try to do too many things for too many people.

    Hmm… well, tell us a bit more about the limitations of GDS, then

    GDS provides ease-of-use at the expense of power-tool complexity. Whether in the interest of not overwhelming non-geeks or simply due to the fact that this is still a version 1 (0.9?) beta release, Google has chosen not to offer much functionality customization or ANY UI customization.

    In some cases, this is a mere minor annoyance. We can’t opt to have the system go into a turbo mode (using more processor cycles) to get everything indexed quickly.

    There’s no way that I can see to have the system wholly refresh the index without doing an uninstall and reinstall.

    Some holes are a bit more frustrating. GDS doesn’t monitor e-mails and files after it indexes them, which can result in an inability to pull up items, or a duplication of listings in search results. Additionally, while you can tell GDS not to index certain sites or folders, you can’t block it from indexing specific Outlook folders. Luckily, it does ignore (perceived) spam folders by default.

    From a UI perspective, GDS is generally streamlined and will please folks who are looking for an experience that matches what they’re used to with general Google search results. Power users, though, may be a bit disappointed.

    The key issue is that GDS doesn’t yet offer search results in a contextual way. For instance, when you’re searching for an image on your hard drive, you probably want to see image thumbnails. When you’re searching for a particular e-mail, you likely want an easy-to-scan list of mail-related headers. However, GDS treats all search results alike, except for a little icon next to each search listing. In a way, this is understandable. In order to offer more strongly contextual search results, GDS would have to do one of the following:

    1) Show only one type of results per screen, which would require extra navigational clicks.
    2) Divide up results by column (e.g., e-mail results on the left, image results on the right), which could be problematic when there are more than 2 file types returned or when someone’s browser window space is limited.
    3) Limit search queries to one file type at a time.

    As you can see, none of these options are particularly desireable. Furthermore, many folks might find it disconcerting to see a shifting UI with search results.

    Therefore, it’s not surprising that Google has decided to go with a more generic but consistent results UI at this point.

    So GDS is sounding like a useful and effective, if not revolutionary product. How does it compare to existing desktop search solutions?

    It’s easier to install, it behaves more nicely, and it returns results faster than any other desktop search program I’ve seen.

    And speaking of other solutions, here’s a list of other desktop search programs I know about, with an asterisk by those I’ve tried:

    – X1*

    Seemingly one of the most popular desktop search tools, X1 distinguishes itself in several ways: it displays search results as you type each letter, it shows previews of documents, it offers special fields for different searches, and it supports a ton of different file formats. Downside? It’s pricey, it can be a resource hog. [See license giveaway; still some left as of 10/15/04 2:46pm PST]

    – LookOut*

    Recently bought out by Microsoft, this program integrates into Outlook, but searches many different types of files. It’s more flexible and powerful than GDS, but it’s not as fast… and it’s not useful for those who don’t have Outlook.

    – Enfish*

    Big, pricey, and highly customizeable, with an especially useful saved-searches feature and decent integration with ones Outlook data overall.

    – Copernic
    – dtSearch
    – 80/20
    – Superior Search
    – Wilbur

    Wow, that’s a lot of competition for Google. So what makes GDS so special and important, then?

    Several things:

    1) Speed, stability, and ease of use.
    2) The trusted and liked Google brand.
    3) Integration with Google Web searches (you can configure the program to show desktop search results when you do a general Web search)
    4) Potential later integration with existing popular Google tools (toolbar, Gmail, etc.)
    5) Perhaps somewhat controversial… but there’s also the possibility that Google could — at the user’s option — use desktop search fingerprinting to steer or filter that individual’s Web results.

    Okay, I can see why GDS may be particularly noteworthy for consumers, but what does it mean for Google?

    It will enable Google to take over the world, though not all at once.

    But seriously, folks… I think GDS suggests several significant ramifications for Google (and yes, its shareholders):

    1) This may increase the frequency and quantity of Google Web (or integrated Web + GDS) searches, thus augmenting ad revenues.

    2) As many others have noted, it’s a clear strike-ahead at Microsoft, who is building desktop search capabilities into their future operating system. Why does this matter from a revenue standpoint? GDS will help insure that more people remain loyal to Google (and its advertisers), instead of defecting towards the possibly-OS-default MSN search feature.

    3) Each foray into Windows tools allows Google to build up an aggregate competency in this area, strengthening not only each individual tool, but the broad set of desktop-based tools overall (Picasa, Deskbar, etc.).

    Well, I’m pretty sold on Google Desktop Search at this point, I must admit. But how do I know if it’s right for me?

    I’d suggest that you simply give the tool a try. But if you’re really gung-ho about this space and have some extra time, you may want to give the other desktop search program a look-see as well. It’s notable that all of them are either free or offer free trials, so you have little to lose. Just remember to fully un-install any desktop search programs you decide not to use to insure that their likely-sizable indexes aren’t continuing to clutter up your hard drive.

    And in the meantime, here’s my humble list of what to look for in a desktop search tool:

    1) Ease of install (and, indeed, uninstall!)
    2) Comprehensiveness of indexing (how many different file types does it support?)
    3) Speed of indexing (initial and ongoing)
    4) Load on computer (during initial indexing, ongoing indexing, searching)
    5) Speed of searching (how long does it take for it to deliver search results?)
    6) Power of searching (Boolean expression support? Ability to search particular fields?…)
    7) Usefulness of search results (relevance, completeness, formatting)
    8) Cost of program
    9) Support from company (FAQs, e-mail support. etc.)
    10) And I almost forgot the most important thing — does it work on your system (platform, necessary specs, etc.)

    Adam, thanks so much for all of this information and for taking the time to stop by here today

    Absolutely a pleasure. It’s always been my dream to appear on BLADAM, and I thank you for the opportunity to address your wonderful, smart, and attractive readers who will no doubt eagerly link to this page and share it with all their friends. But tell me one thing: Why has it been more than four months since you posted anything in this blog?

    Ahem… um… well, I was in the process of changing blog software, but never got around to actually finishing the transition, plus I started two new jobs recently and…

    Ah, no worries! Just glad to see BLADAM back up, even temporarily.

    :blushing: It’s nice to be back.

    —

    Relevant and cool links:
    – A Net Takeway examination of the desktop search space
    – Excellent overview of GDS from SearchEngineWatch
    – SearchEngineWatch – Privacy and Google Desktop Search
    – John Battelle’s take on GDS

    —

    Edits: (Times are Pacific Standard Time)
    – 10/14/04, 21:45: Actually linked to the product. What a revolutionary idea, eh? [smacks head]
    – 10/15/04, 14:46: Added x1 license giveaway link (not likely to be applicable for long!)

    —

    [P.S. — Please feel free to leave a comment with any corrections, or suggest any additional aspects of GDS to cover. Thanks!]

  • Gmail: Do we really want a TERABYTE of space? (a.k.a. "Be careful what you wish for…")

    Gmail indicates a terabyte of space available.

    [See other BLADAM entries on Gmail… and also, as I’ve noted earlier, please do not ask me for an invite; I don’t have any to spare… sorry! – Adam]

    A number of Gmail users — including yours truly — have noticed that Gmail has seemingly upped the service storage limit to one terabyte. That’s right… 1,000,000 — one million — megabytes (compared to Hotmail’s two megabyte free storage limit, for instance). To put this into perspective… even if you got 250 e-mails each DAY (averaging four kilobytes each), it would take you about 2,739 YEARS to fill up this much space.

    Of course, there’s certainly the possibility that the new indication of a terabyte of space is a Gmail bug. Gmail is, after all, still very much in beta. But what if it’s not a bug? What if the terabyte storage limit is really, well, real? Would we truly want a terabyte of space? Or would this lead to more problems than benefits?

    Okay, those may sound like silly questions. After all, isn’t storage space like money… the more the better?

    Maybe not.

    I’m beginning to ask myself about the issue not only of diminishing returns (at what point does the space-race become a bore?), but — more importantly — whether such a huge capacity could lead to some unexpected challenges.

    True, Google search is excellent. But here’s what worries me as I think about the consequences of really dumping in and keeping many tens of thousands of even the most banal / temporarily-significant e-mails into my Gmail account:

    Let’s say I know I’m looking for an e-mail I got a few years ago from a woman named Jen. Or maybe Jenny. Or Jennifer. Something like that. And we were talking about going on a skiing trip together. When I have 200,000 e-mails in Gmail…

    – I have to remember to specify (“Jen” OR “Jenny” OR “Jennifer”) AND (“ski” or “skiing”) since Gmail doesn’t support partial word searches (though admittedly this isn’t different from what one must do now for this sort of search).

    – I may be deluged by false positives… e.g., notes from travel agents named Jen/Jenny/Jennifer, friends talking about going skiing with their fiance Jen or colleague Jenny, an article in the New York Times written by journalist Jen Smith talking about a bank robbery by someone with a ski mask, and so on.

    * * *

    Even in searching my own 2-gig-or-so Outlook e-mail collection nowadays with a fabulous search tool (“LookOut“), I find myself increasingly regretful that I didn’t throw away a lot of the newsletter e-mails, temporary “verify your subscription” e-mails, one word reply e-mails with a ton of quoting, and so on. Specifically, I’m starting to suspect that being a packrat — and, in particular, having the ability AND encouragement to support this habit — may not be all it’s cracked up to be.

    Don’t get me wrong. I still love Gmail and think that, overall, it’s a humungous step up from the miserly accounts one gets with Hotmail and Yahoo and so on. But at the same time, I wonder if people are prepared for potentially frustrating consequences of REALLY saving absolutely all their e-mails.

    What do you think?

    Added later the same day:
    Google, please don’t vengefully remove my terabyte of storage. I may be concerned, but I’m still greedy. And at my current rate of overall mail sending/receiving, I’d fill up a gig in 440 days! :O

    Added May 19, 2004 (the next day)
    Oops. Be careful what you diss; looking a gift horse in the mouth may cause the horse to kick you. Er, in this case, the extra storage has gone *poof*. Bummer.

    And wait… adding to the confusion… some of my friends still show a 1TB limit in their Gmail account. Hmm!

  • Google stole my idea for discussion lists! (an early review of Google Groups2 / Google e-mail lists)

    Okay, not really :-). But maybe I should go into fortune telling, because I had written about the newly released Google Groups2 Discussion List Service two weeks ago on the “What Should Google Do” group on orkut.com:


    Email Discussion Lists

    The competition is simply awful. Yahoo! inserts annoying banner ads everywhere and sometimes even full-page interstitials. MSN lists are awkward to set up and use. And Topica (free version) has suffered from poor reliability and over-ad’edness.
    Google could create a stellar e-mail discussion list service, especially given that:
    – it already is rapidly acquiring mail-related know-how from Gmail (anti-spam issues, smart threading, etc.)
    – AdSense can be even MORE targeted (and profitable!) on many if not most discussion lists, and Google has in fact already experimented with putting AdSense ads in e-mail newsletters.
    Overall, I’m excited about how Google could create some fabulous synergy with Google Groups (definitely in need of a UI overhaul, IMHO), Google E-mail Lists, and Gmail… really leveraging threading, anti-spam protections, and of course, search!
    What do you think?

    Well, given my earlier-written note above, you can guess what *I* think about it. It’s pretty damn exciting!

    However, a few things concern me:

    1) Will it stay banner-ad free?
    Google writes this on the Google Groups Help page: “Google Groups never displays pop-ups or banner ads.” Sounds pretty absolute. Note, though, that it doesn’t say that Google will NEVER display banner ads (or even the nasty interstitials that YahooGroups annoys the hell out of folks with). And given the fact that — to my disbelief — Google started including up to 50Kb in size banner ads on AdSense content sites, I’m not quite what to think about Google’s commitment to a streamlined user interface.

    Of course, I’m minorly hypocritical. I include some banner ads on my sites, including this one. But that’s my own personal choice. I’m not sure I’d want to see a banner ad on every Google Groups page :(.

    2) There still seems to be no protection against evil spam-bots harvesting Web addresses
    Unlike most standalone Usenet (newsgroup) readers, Google Groups2 (like the regular Google Groups) offers no way for posters to hide their e-mail addresses. Why does this matter? Well, as soon as I post to rec.arts.whatever with my coveted Gmail account address, it’s probably just a matter of days before one of those Godawful spammers uses a tool to scrape my address (and millions of others) off of Usenet postings… using them not only to send spam, perhaps, but also to sell as part of ‘spam address’ CDs, guaranteeing me bucketloads more spam in the future. Indeed, I believe it’s largely due to my earlier unobfuscated Usenet postings in the distant past that have resulted in a couple of my e-mail addresses getting several HUNDRED spams per day.

    3) Right now, the user interface doesn’t come close to matching the elegant simplicity of Gmail… or even various software Usenet readers.
    While tree-view is a bit more palatable, the default Groups2 view is functional but, frankly, a bit unwieldly and not something I’d want to stare at for any length of time.

    * * *

    With that out of the way, let me offer my assessment of some of the GOOD stuff I’ve found in my early testing:

    1) It’s fast
    Going from screen to screen is quick, and time-to-receive-sent-messages is darn fast, too (Google claims messages are sent within 10 seconds! Take THAT sluggish YahooGroups! :D)

    2) So far, sent messages are naked!
    I can’t imagine that this’ll persist, but right now, group mail is delightfully devoid of obnoxious taglines, and even ad-free so far!

    3) Rather rich functionality
    If imitation is the sincerest form of flattery, Yahoo! and Topica should be positively blushing right now. GG2 offers pretty much the core of those services’ functionality, at least in terms of e-mail options, though it lacks the extended features that YahooGroups offers, such as calendaring, file storage, and so on.

    * * *

    Will people make the switch? It’s hard to say. I’m guessing that many folks — understandably fed up with the flakiness and intrusiveness of the existing discussion list services — may strongly consider GG2. However, there are rather high switching costs in many cases. While Google makes it easy to import new subscribers (add OR invite… hmm… I wonder how they’re guarding against abuse with that!), many members will have to update their spam or general rules filters, and — more commonly — get used to a new interface. And those of us with e-newsletters may seriously have cold feet; I’ve already moved my subscribers from list service to list service (one folded, one became too expensive, and so on), and I am quite hesitant to once again drag everyone to a new home.

    Where I think GG2 has the most potential is in the creation of *NEW* groups, since there’s no user or moderator baggage to deal with. With similar feature sets and Google’s trademark reliability (except for AdWords :cough cough:), I see few reasons for anyone to start a list on anything other than Gmail.

    There are, of course, a few exceptions ;-). Here are some cases in which GG2 may not be the best choice for aspiring moderators:

    1) Branding is critical
    GG2 is NOT the right choice for most companies, who (IMHO) should really be sending out mail listed as from theircompany.com, or at minimum, theircompany.trusted-mailer.com. “Free” — no matter how pre-IPO fashionable — is generally less respected by customers who may expect an ad-free and custom communications channel from the companies they do business with.

    2) Your group isn’t ‘critical’ or ‘required’ or ‘differentiated’
    There are switching costs for members, as described above, and some may not wish to establish Google Acccounts, no matter how easy this is to do. As a result, some may bail out, preferring to stick with tried-and-true YahooGroups or Topica Groups.

    3) You make use of YahooGroup’s extended features
    If you maintain group files (photos, databases, FAQ’s, etc.) or use the calendar function or rely upon online-presence-detection (who’s online Yahoo), then GG2 is not (yet) the right choice for you.

    * * *

    With that said, I encourage everyone to give GG2 a try. You can create groups for free and with surprising ease… and so you now have new and spiffy tools to start the 387th Britney Spears fan club. On the other hand, maybe it’s best to follow the Google creed; Don’t Be Evil 😀

    * * *

    Want to talk about GG2? Feel free to post your comments here, or join the official (not run by me!) Google Groups2 Discussion List. Be sure to read through the GG2 info materials first 🙂

  • Gmail may not change the world… but will it substantively change the way we e-mail?

    [As I’ve noticed in my earlier Gmail commentary, I unfortunately do not have the ability to offer invites… sorry 🙁 – Adam]

    With all the coverage of Gmail (to which I’ve somewhat guiltily contributed), one might snicker that Gmail’s being positioned as something that’s going to change the world.

    Well, almost. 😉

    I seriously think that Gmail may substantially change the way people deal with e-mail… sending, receiving, storing… leading to some interesting and not-insubstantial behavioral changes.

    First, there’s the admittedly obvious possibility: that Gmail will acclimate users to archiving rather than deleting mail.

    But I’m thinking there may also be quite a few other ways that Gmail will change users’ behavior, assuming the service catches on widely and wildly, as I think it will.

    – Reduced quoting
    Since all previous correspondence is right there in plain (conversation) view, I think people may be more apt to eliminate the redundant quoting… especially if Gmail ceases adding this into replies (particularly Gmail-to-Gmail replies) by default.

    Why is this significant? For starters, it will help folks pinpoint what they’re looking for when they search through e-mails. Right now, with many folks quoting from the last bazillion e-mails in a thread because they’re too lazy to trim the quotes, their e-mails show up in searches when they may not be relevant… specifically, when the searcher was really attempting to pull up the original e-mail that they merely quoted!

    In contrast, when people are encouraged to trim their quotes or at least begin to see that quoting entire previous correspondences is unnecessary, e-mail searches will become more relevant, with fewer but more targeted hits. The haystack, in effect, becomes smaller, but the needle remains the same size 🙂

    – Greater sensitivity to the subject line
    On one hand, I think people may be reluctant to change the subject line in an ongoing conversation for fear of ‘breaking’ the conversation (I actually got hollered at by a friend for doing just that).

    But on the other hand, I believe folks may perhaps become more sensitive to the realities of both personal and public (discussion list) conversations floating to wildly different topics… and feel compelled to change the subject to split the conversation. Of course, Google’s implementation (or not) of conversation joining/splitting tools will certainly have an effect on all of this as well.

    Overall, I think that the proliferation of Gmail may encourage people to pay more attention to subject lines which, IMHO, is a great thing. Personally, I’m sick and tired of people talking about hiking shoes when the subject line is still unchanged from the first e-mail in a 38-note thread that started with “Best pocketable camera for the outdoors.” After all, I may be interested in one topic and not the other, yet loathe to either waste my time reading through everything or miss discussions that I’d want to peruse.

    Nowadays, it’s mostly us geeks who modify subject lines to read something like: “Hiking shoes (was: Best pocketable camera for the outdoors)” and it’d sure be great if others were as considerate… or even if Gmail helped encourage such changes! Ah, but I digress…

    – Lessened reliance upon HTML e-mail
    I’m assuming that Gmail will eventually support the creating / forwarding of HTML e-mail, but in the meantime, I’m wondering if people will be less apt to make use of HTML mails (creating in other clients, forwarding them in Gmail) due to the current lack of HTML e-mail support in Gmail.

    * * *

    What do you think about the scenarios described above? And how else might Gmail change the way people use e-mail?

  • A mercifully very short entry on the Google IPO

    Many people know I have significant ties with Google; I use AdWords personally and for my clients, I use AdSense on my own Web sites (including this one), and I have many friends who are Googlers. I am hardly an impartial observer.

    I also know that this dang Google IPO has been covered ad naseum. So while I feel compelled to say something here, I’ll keep it blissfully short:

    Google is an honorable company, with good people (execs and all) that I believe really do want to make the world a better place. I hope they continue to succeed in every way.

  • Gmail’s not the problem: A look at misguided and dangerous privacy positioning

    HARK! A powerful new technology! Let’s [try it / ignore it / ban it]!

    History repeats itself. When new technology is developed, there are the early adopters who coo excitedly, those living under a rock who grunt “huh?” and varying degrees of those who issue warnings, which — at least as of late — seem to fit into one of several convenient categories:
    1) Fighting against terrorism
    2) Safeguarding values (e.g., “Protecting the children!”)
    3) Protecting privacy

    To simultaneously save me from getting writer’s cramp and enable you to finish reading this essay in under two hours, I’m just going to tackle the issue of privacy for now.

    * * *

    Examining privacy

    First, let’s take into consideration the two fundamental aspects of privacy:

    1) SECLUSION: Right to be left alone; protection from intrusion, interruption, etc.
    2) SECRECY: Right to keep one’s personal life from prying eyes

    Much ado has been made of Gmail’s alleged or potential violations of both types of privacy by other individuals (advertisers, hackers, employees) or entities (Google, other companies, government agencies). And admittedly, with seemingly so many potential loci for abuse, it’s understandable that many folks are concerned about a project of the scope of something like Gmail.

    However, in this essay, I’d like to offer my opinions as to why these risks are outweighed by benefits and — more importantly — why focusing on ‘reforming’ (or even suspending) Gmail harmfully takes our eye off the real problems in our society.

    * * *

    Assessing privacy issues in the context of Gmail

    It’s easiest to dispense immediately with the ‘right to be left alone’ aspect of Gmail. Unlike the flashy, gaudy, intelligence-insulting “You’re winner number 314159!” pop-ups that pollute leading Web mail services (and much of the rest of the Web), Gmail’s ads are deliciously unobtrusive. They don’t blink, they don’t flash, they just sit there quietly on the side — barely noticeable. And since Gmail does not insert so much as a tag line (much less ads) in any outgoing mails, it becomes even more difficult to forward an argument that Gmail creates an interruption or other intrusion into one’s daily life.

    One might counter that the very presence of contextual ads may create a feeling of queasy intrusion. However, Gmail has (in my testing) been pretty cautious about placing ads next to apparently-personal or otherwise sensitive e-mails, and of course, it does not place any ads next to e-mail marked (automatically or manually) as spam. Still, there’s admittedly the possibility of an unpleasant juxtaposition of a heartfelt e-mail with a angst-inducing related ad, but such an occurence is also possible via a song on a radio, an ad on TV, a seemingly random phone call, and so on. Our minds are rich in imagination, and capable — for better or worse — of astounding leaps of correlation. Once again, this is not something one should hold Gmail responsible for, IMHO, and it hardly seems to fit into the issue of privacy. And of particular note, the Gmail text ads are lightyears away from the ludicrous example of a billboard in one’s living room, as one clearly-clueless senator has suggested in tandem with her anti-Gmail legislation.

    So what we’re left with, then, is the more meaty and complex aspect of privacy-as-secrecy. With Gmail’s one gig of storage, it’s no doubt that many folks will practically be entrusting their life’s story — AND secrets — to Google. And in this area, I honestly have no easy answers. I will, however, attempt to address the various pieces:

    – Security breaches / the bored employee
    Google has an excellent history of data integrity, having (to my knowledge) avoided any data theft from its multi-billion dollar AdWords or AdSense programs. But other big players haven’t been quite so successful. Many of Amazon.com’s anonymous viewers found their screen names exposed a few months ago (frankly, much to my evil delight, since many if not most of the anons were shills and twits). On a less innocuous note, many Hotmail users found their accounts suddenly ‘open to the public’ due to a security exploit that Microsoft then understandably scrambled to fix.

    Could Google stumble here? Nothing’s impossible. Unfortunately (and in fairness to Google), however, this is a universal problem. Though I’d frankly trust Google with my data over most other companies, I think it’d be a mistake to assume that anyone’s data is 100% safe, anywhere.

    On a similarly sobering note, I’ll add that it’s trivially possible for an Information Systems worker at your company or your ISP to read your incoming and outgoing mail. Google has vehemently highlighted precautions it’s taken to prevent this happening at their company, and frankly, I believe them. Would you, as a highly-compensated employee, risk your job at such a coveted organization for the opportunity to snoop into likely banal exchanges? Seems unlikely to me. But that aside… once again, this is not a Gmail issue. This is a general e-mail issue.

    – Selling/giving data to other companies
    As a Google (AdWords) Advertiser and Publisher (AdSense), I can tell you point-blank that Google does an incredible job at protecting the data of both its users (visitors to Google.com and AdSense Web sites) and its advertisers. As both an advertiser and publisher, I am absolutely unable to glean any personal information about people who click on my ads. Google actually maintains a very strict separation between its departments, and were it (or another company) to risk this integrity in the future, the brand would be tarnished beyond recovery (and it’s not like there’s not competition!). In other words, unlike with almost every other transaction we partake in In Real Life, I firmly believe we can trust Google not to share our personal info with other companies.

    – Giving our secrets to federal governments… oh yeah, and lawyers
    If Ashcroft applied enough pressure to Google next month and insisted on wiretapping a few dozen “suspected terrorist” accounts, I’m betting that Google would buckle.

    Wow, that’s probably not what you expected me to say, is it?

    I’ll surprise you further: I have no doubt that lawyers are indeed salivating at the likelihood of millions of people getting Gmail accounts and storing voluminous amounts of discoverable data, perfect for future lawsuits.

    Of course, Ashcroft and — to a less chilling but likely more common extent — Dewey Cheatem and Howe have been subpoening mail records from Hotmail, Earthlink, Comcast, Big 10 universities, Fortune 500 companies, and so on. Gmail just expands the scope.

    But Gmail isn’t the problem

    And now we get to the meat of this essay, in which I argue — after acknowledging Gmail’s unwilling-but-likely facilitation of government and lawyerly snooping — that Gmail is absolutely, positively not the problem.

    Electronic Freedom Fighter and longtime smart-geek Brad Templeton gives us a good start here via his entry on Privacy and Gmail, which includes:

    …there are also some deep issues here, worth discussing with not just Google but all the other webmail providers

    …but Brad doesn’t go far enough. E-mail privacy concerns aren’t — or at least shouldn’t — be just limited to Webmail. As noted above, your mail is susceptible to snooping no matter where it resides… unless you’re one of the 0.0000002% of geeks who use encryption, and that’s likely to send up a red flag to Ashcroft anyway ;-). And though Brad’s suggestion that Gmail incorporate a more user-friendly form of encryption has merit, I think it’s a mere bandaid on larger problems. In fact, I’m generally uncomfortable with the feeling that Brad places the onus of responsibility (and solutions) upon technology, rather than those who abuse human rights via such technologies.

    As we move these things [record of peoples’ lives] online and outside, we build some of the apparatus for a surveillance society.

    I strongly disagree. This is like saying that when Japan set up high-speed rail transit, the country created the apparatus for criminals to flee farther and faster. Or as long distance calling costs have become insanely cheaper over the last 15 years, it’s made it much, much easier for people to plot nefarious acts over the phone more affordably.

    The introduction of new technologies — or the exponential improvement of existing technologies — facilitates opportunities for good and for not-good. But blaming the technologies (particularly communication technologies) is horribly wrong for two reasons:

    1) It stifles innovation, and keeps costs artificially high.

    Can you imagine how much more work — and perhaps more value — Google could be putting into Gmail if they were able to spend less time fighting PR fires?

    2) It takes our attention off of the real culprits: federal powers without sufficient accountability or checks or balances, not to mention a legal system which rewards, even necessitates antagonism over consensus.

    Let’s stop blaming tools, and start fixing the deeper problems

    We should channel our indignations towards privacy-enroaching intrusions such as the Patriot Act. We should vote our representatives — Democrat or Republican — out of office when they value false security over liberty. We should demand accountability, and insist that our journalists actually ask tough questions in Washington D.c. (and around the world). We should, as Brad rightly notes, fix The Electronic Communications Privacy Act (“ECPA).

    Wringing our hands over Google isn’t just misguided, it’s dangerous in that it blinds us to the root of deeper problems in our society:

    1) As a society and individually, we’re generally unable to intelligently weigh risks.

    In our quest to “fight terrorism” we forget that far more people die from spousal abuse, drunk driving, malnutrition, etc. Will Ashcroft start scanning our e-mails for keywords relating to those causes of death? That’s (thankfully) doubtful, but only because he — and most of the FOX-watching public — is unable to thoughtfully weigh risks and benefits.

    In the vast storage that Gmail offers, non-profits can save and find key documents that help them better deliver social services. Grandparents can pull up pictures of their grandkids. Self-employed folks can be more productive, boosting their bottom line and doing their part to also buoy the economy. Sure, these are all butterfly-flaps, so to speak, in the ongoing history of society, but in the aggregate it adds up to a lot of good.

    2) We are tragically misinformed and poorly educated and we don’t seem to care.

    We fail to take our liberties seriously until we see sensationalist articles in the paper or hear about grandstanding bills proposed by a tech-ignorant senator.

    When I recently chatted with some friends here in the generally tech-savvy area of San Francisco about Gmail, nearly half suggested to me that they wouldn’t touch Gmail with a ten-foot pole. “They’re gonna read all the e-mail I get!” was the oft-expressed concern. Clearly, these people had no clue that their e-mails are already “read” many times enroute to their destination, nor did they think to actually visit Google’s Gmail pages to get the real info about what the service would and would not be doing. Funny, but I don’t recall my friends worrying about the privacy of their e-mail before.

    I don’t think the masses are misinformed because they’re dumb (though there’s certainly some of that). Rather, we absorb AND support the lowest common denominator in journalism. Not only does sex sell, but so does any bad news — whether it’s relevant or even true. “Duh, Sherlock” I hear you saying, teasing me for pointing out the obvious.

    But if it’s so obvious, why aren’t people like Brad and other prominent technologists decrying the lack of journalistic integrity and dearth of basic technological education and knowledge in our society? I have no doubt that if more people knew and truly understood the tenuousness of our personal privacy in society (with regards to credit card purchases, political affiliations, and yes, e-mail), they’d be concerned and mobilized to change the fabric of society and government… rather than throwing stones at a Webmail provider that’s providing a richer offering than its competitors.

    In summary…

    We need to focus on the fundmental roots of the erosion of our privacy, and realize that Gmail is a serendipitous wakeup call, not a culprit.

    True… Gmail, in itself, is not likely to lead to absolute world peace, and indeed, it’s admittedly easy and sometimes tempting to overstate what may be more of an evolutionary rather than revolutionary technological offering. But every moment people spend attacking this or any other communication tool amounts to time and efforts distracting us from the real (and real serious) issues facing our nation and our world.

    * * *

    Related entries:
    – Brad Tempton’s essay on The GMail Saga
    – My review of Gmail
    – My writeup of Gmail tips and tricks
    – Blog posts (via Technorati) on Gmail and Privacy

  • Got a Gmail account? Here are a few interesting tricks ‘n’ tidbits

    [In case you didn’t already notice, I wrote a pretty detailed review of Gmail earlier, in which I also noted that I’m unfortunately unable to procure accounts for folks. Sorry! For those who already have accounts, I hope the tips below are useful and/or fun 🙂 – Adam]

    So, you’re one of the Gmail testers and you want more bang for the buck, eh? Try these tips on for size:

    SENDING & SHORTCUTS
    – Send mail to fellow Gmail’er by just entering their username in the TO, CC, or BCC spot. No need to include @gmail.com 🙂

    – Did you reserve .(JavaScript must be enabled to view this email address) and now wish you had left out the darn period? It’s not too late! For whatever reason, Gmail treats that e-mail address the same as one without a period (and visa versa), so .(JavaScript must be enabled to view this email address) works just as well as .(JavaScript must be enabled to view this email address). As you may have guessed, Gmail is flexible with regards to capitalizations, too!

    – Using the handy-dandy spell-checker and want to quickly Resume Checking without using the mouse? Just hit ‘R’, an undocumented shortcut key in this context.

    SEARCHING
    – “or” is not the same as “OR.” Only the capitalized version (sans quotes, by the way) will work with searches. So if you want to find mail from your friend Jen, you can use this in search: FROM: (jen OR jenny OR jennifer). Note, by the way, that the actual search terms are not case-sensitive. “jen” works just as good as “Jen.”

    – But, using the same Jen example, it’s important to realize that the search engine of Gmail (and Google, for the most part) does not search partial words. So “jen” will not find “Jennifer.”

    PERFORMING ACTIONS ON A GROUP OF E-MAILS
    – Let’s say you have 150 e-mails, listed over two pages (100 max per page), and you want to archive all of them. I initially made the mistake of clicking ALL, then hitting ARCHIVE and thinking that this would do the trick. Nope. Commands — whether TRASH or ARCHIVE or LABEL — only affect those items that are both selected and on the page you’re currently viewing.

    – And speaking of groupings… don’t forget that when you archive or label e-mail, you’re affecting the entire Conversation of e-mails by default. If you want to trash just one of the e-mails in a Conversation, you can do this by expanding that particular e-mail, clicking on MORE OPTIONS, and then clicking on TRASH THIS MESSAGE.

    ALL THIS *PLUS* A LITTLE BIT MORE
    Gmail supports “plus” addressing, which means that if your address is .(JavaScript must be enabled to view this email address), you can receive e-mail at maryhadda+littlelamb@gmail.com or maryhadda+longaddress@gmail.com, etc. Why is this useful? Well, Mary (or you!) could use one, er, I’ll call it a “plussing,” for mailing lists (“maryhadda+lists”), and another for shopping online (“maryhadda+shopping”) and so on, and then create filters to put useful labels on the different types of mail.

    Some have suggested that this could also be a useful spam deterrent (e.g., using maryhadda+2004q2 and then discarding e-mail sent to this address the following quarter), but I think this suffers from two key flaws:
    1) Spammers are probably smart enough to start stripping off the plussing :(.
    2) After a while, you’d have to create a LOT of filters, and — at least for the moment — we only get an allotment of 20 filters total. It’d be a shame to use those all up in a (likely futile) attempt to thwart spamjerks.

    *** Edited 4/30/04:
    Some folks had expressed concern that plussing was seemingly not working for them. However, I’ve worked with them to track down the cause: Due to the way Gmail handles discussion list mails — not showing one’s own contributions in the Inbox to avoid duplicate views — people who were testing the plussing feature by mailing themselves via Gmail wrongly assumed the mail was ‘lost’. As it turns out, the mail was indeed received, but — since it appeared to be part of a ‘discussion list’ — was not shown in the Inbox, which is what caused the confusion.

    THE BOTTOM LINE: Plussing works; just don’t try sending tests to yourself FROM your Gmail account TO your same Gmail account :D.

    * * *

    Well, that’s pretty much everything off the top of my head for now! How about you? Got some cool Gmail tips or tricks? Speak up below, or feel free to contact me 🙂

    * * *

    *** Added 4/21/04
    Uh oh! I have competition! 😀 It’s been pointed out to me that there’s already a blog dedicated to Gmail tips and tricks, called Gmail Gems. Definitely worth checking out.

    *** Added 10/11/05
    Want to read more stuff about Gmail and Google? Check out the BLADAM Google category! and don’t forget to subscribe to this blog! (see the options in the top menu!)

    src=”http://pagead2.googlesyndication.com/pagead/show_ads.js”>