Hawaii Klout Portal Launched

Hawaii Klout

Hawaii entrepreneur Rob Bertholf this weekend launched HawaiiKlout.com, a directory of “Hawaii’s Most Influencial.” (Presumably, “Influencial” is a portmanteau of “influence” and “influential,” and not a typo.) The site connects to an API provided by Klout, which purports to provide “the standard for influence” across social networking platforms.

To find Twitter users in Hawaii, Bertholf explains, the site indexes Twitter posts as fast as it can (a rate limited by Twitter to 150 scans per hour) and accepts user submissions as well. After non-Hawaii users turned up in the rankings, a “flagging” system was quickly implemented to remove bad results (like Justin Bieber).

Klout is one of many services attempting to provide metrics for the fast evolving social media industry. Others include PeerIndex and Twitalyzer. Bertholf says that he is working on integrating PeerIndex data into HawaiiKlout.com, but notes that he thinks Klout may be the dominant service.

Even so, Klout numbers have their quirks.

“Klout is unfortunately biased against frequency of tweets over quality and reach,” Bertholf notes. “The system isn’t perfect, it puts too much weight on outgoing mentions which can be spammy.”

Still, it’s a start. Noted Russ Sumida: “I think for lack of a metric, any metric is a good start. Eventually you get to know who you can trust and who you can’t. Just like in real life.”

“When [Melissa Chang] and I do any training for newbies to Twitter, we don’t point them to Klout. We tell them to look at the tweetstream,” Sumida went on to say. “That wasn’t a dig against Klout. We just want them to get to know what a bot looks like, to learn and not waste their time. For measurements, they can see their follower count. For our clients, we engage as the client, so they gain the followers/value.”

Of course, using a follower count as a metric also has its weaknesses. (Klout agrees, but of course they would.) For best results, you need to look at the whole picture, which is just as true in social media as it is in the real world.

See Also:

Ryan Ozawa

Ryan Kawailani Ozawa has immersed himself in new technologies and online communities since the days before the web. From running a dial-up BBS in the early '90s to exploring today’s dynamic world of "Web 2.0" and social media, he has long embraced and evangelized the ways in which technology can bring people together.

  • Mahalo for the coverage Ryan! (and Kudos for hanging onto #1 position even after a day of scanning!)

  • This reminds me of something I heard about 6 months ago. The Marketing Over Coffee podcast was asking its audience for their favorite reputation metrics and services. I think even then Klout was the only common thread, but it wasn’t used as it was intended. They used the Klout score to filter out robots. I think Chris Penn wrote a script that would auto-block his followers if their Klout score was less than 15 or 20. And, this system worked great for that purpose. But, they don’t give much weight to someone’s Klout once it is above a minimum threshold.

    • Based on what I am seeing with the data Klout seems fairly good at identifying who I would consider “influencers” in the online space. There were several people who use Twitter to chat with friends who we manually weeded out but feel it was about 80% accurate. Klout seems to put too much weight on frequency of conversation (although I get the reason why, engaging and keeping the conversation moving is valuable) but I am confident they are heading in the right direction.

      • “There were several people who use Twitter to chat with friends who we manually weeded out.”

        Rob, could you expand on this further? I recall when you launched, there were a couple of Hawaii users I didn’t recognize, and definitely saw that they were very active with their friends but not particularly plugged into the broader community. But does that warrant ‘weeding out’? Does that weeding out nullify using Klout’s independence — however flawed — as the hook for this directory?

        More tricky, I think, is determining what Twitter users qualify as a Hawaii Twitter user. If they put Hawaii in their bio because they were born here, or love visiting here, or sell vacation packages here, does that make them a ‘Hawaii Influencer’? There are some very active users who position themselves as Hawaii community members — perhaps deservingly so — while living elsewhere. How are you going to make those calls?

        • Ryan brings up good questions. Although there were some users I didn’t recognize (that are gone now), I’d like to think that Twitter and social media are not about cliques. Plus, there are a few accounts in the Top 50 (friends of ours whom we’re all happy to talk to) who use their accounts for some pretty heavy back and forth with friends.

          I’d like to add that I love this initiative, and great job in putting it together. Definitely helps users understand their usage.

          • Hi Ryan/Gene, sorry for the delayed response.

            First and foremost this is a tool for Hawaii and the Hawaii business community so the results are up to us all to view and manage. There have been quite a few accounts flagged, once flagged any of the moderators can remove or re-add (Currently: John, Toby, Neenz and anyone else who wants to help).

            You can see a list of the accounts flagged here: http://hawaiiklout.com/inactive/ and I added the ability to “request review” and have them added back to the main directory. All future flags will contain the reason they were flagged, I am thinking about creating a category for “conversationalists” that would be filtered out of the top ranked influencers but could be visible in their own tab… what are your opinions on this?

            Got some big news about a new relationship with Klout I am excited to announce soon!

          • Thanks for the reply, Rob. I loved that flagged accounts, i.e. the inactive list, is public, although not (yet?) linked publicly. And that there’s a way to ‘request review’ to be returned to the directory. Also glad to see you have some help with moderation!

            But I still don’t think ‘conversationalists’ should be segregated from ‘influencers.’ However Klout measures influence, flawed or not, it identifies both types of users (and other types of users) as having a high score. If you’re going to pull users who don’t talk to the rest of us, who don’t play nice with some clique, you’re undermining the directory’s value.

            After all, I think I’ve only seen Mufi Hannemann’s account reply once to another user, and Neil Abercrombie’s account is mostly announcements these days. Another user on the list interacts almost entirely with people outside Hawaii. News organizations definitely don’t operate the same way as individuals.

            They may not fit the same usage pattern as the people you want to highlight, either, but if Klout says they provide value, they should stay on the list.

          • Thanks Ryan, all totally valid points! What is important to put into perspective is Klout is not perfect yet and is looking for ways to be more accurate (I would love to have a discussion offline which may interest you based on a recent conversation with their CEO, its just not public knowledge yet) so when we identify “Influencers” this is not determined by a metric, this is only visualized by a metric. It is up to the community to decide who is an influencer and who is not. As such, the new version I am about to launch (hopefully tonight or tomorrow morning) will allow nominations or peer recommendations if a member should appear in the homepage’s influencer list.

            I am really interested in hearing more opinions and am happy to factor in their responses into our local methods.

            (Missed you at GB today, we need to connect soon!)

          • I think if there’s a list that’s based on ‘nominations’ or explicit moderation, that should be a tab, rather than the main display. Again, the development of a ‘clique’ mentality is something I think many are concerned about.

            If anything, I’d rather see a distinction made between ‘individuals’ and ‘organizations’ before we talk about ‘influencers’ versus ‘conversationalists.’ For example, news organizations (which are all well represented on the site). Separating them out seems a much less tricky call than, say, taking out a massively eager Justin Bieber fan simply because she doesn’t speak our language.

            In any case, thanks for taking the feedback, and remaining responsive!

          • As an aside, I can confirm that Mufi is currently running his own Twitter account, at least sometimes. He DM’ed me recently about something!

          • Rob, I love the whole flagging aspect, and the fact that flagged accounts are visible. It adds some transparency in the process and that’s always good.

            I don’t know if you’ve heard of the hashtag #wjchat before, but it’s used every Wednesday by Twitter journalists to talk about issues and trends. I was touting HawaiiKlout.com as a good tool to make some sense of the metric, particularly the fact that you’ve laid out infrastructure to flag accounts like Bieber nonsense.

  • Anonymous

    I just posted on how Klout is like PageRank and how it just needs some time to develop: http://rob.bertholf.com/software-developer/klout-the-standard-of-influence/

  • Borisjk99

    This sounds like a good idea, but if  Mr. Kolf’s previous business ventures are any indicator of things to come(which I believe they are), then this thing is dead in the water.

    ZeppOS and his Empowered Internet Solutions was one of the biggest scams in a long time. They promised a kick-ass software that would be fully customizable, but easy enough for any newbie to use…not to mention full support. With a prices for their service starting in the lower-mid 4 figures, you would expect quality, but that’s not what you got.

    Instead, you got a software that:

    1. Was full of bugs
    2. Had terrible coding
    3. Not fully customizable as promised
    4. Not very newbie-friendly at all.

    I have seen open source WSYWIG(what you see is what you get) website software that you can download for free that was much better than ZeppOS.

    The hosting was pretty standard, but the support team was not very helpful. They were slow to respond, and rarely available. They finally “topped the cake” by not notifying any of their clients that the company tanked.

    By the time I got word of the company failing, their phone #s were disconnected, thier office was cleared out, and there was no contact information available for customers to find out what was going on.

    Finally, as far as SEO goes(search engine optimization), Rob is about a novice level(definitely not worth the extra couple grand he tries to get from you for his SEO services).