We use CopyBuoy via Hoster Broaster, because it streams really easily into a Plaxo/LinkedIn yak-fest meld. … If you already have a blog, make sure you spray-feed your URL in niblets open-face to the skein. We like Reddit bites (they’re better than Delicious), because they max out the wiki snarls of RSS feeds, which means less jamming at the Google scaffold. Then just Digg your uploads in a viral spiral to your social networks via an FB/MS interlink torrent.

But let’s talk about Twitter influence for a second. There are a number of tools that threaten to measure influence, Twitter or otherwise. Edelman just launched a new fancy one, and from Brian Solis this morning I saw that Klout measures influence over a topic.

Just what does influence over a topic mean? Let’s stay grounded in the reality of the topic I know: VMware. We have a great conversation around the enterprise use of VMware on Twitter. It started around VMworld 2008 and took off from there. The conversation consists of a few thousand IT professionals hanging out, swapping links and advice with some grace notes about our personal lives. Most of the virtualization bloggers hang out there. If you want to join our conversation, a good place to start is this hand-curated list by @ericsiebert, Top 100 VMware & Virtualization People to follow.  If you really want to have a deep technical conversation about VMware, though, I don’t recommend starting at Twitter at all — go over to the VMware Communities and ask away.

From the vantage point of my personal involvement, Eric’s list contains some of the most influential people on Twitter talking about enterprise virtualization. Those people set the current “conventional wisdom” and have “authority” in our community. The part of the enterprise virtualization community that’s active on Twitter is small compared to some – it’s not MomBloggers or the Social Media Ouroboros – but I can assure you it’s a community that is of extreme interest if your company wants to participate in the virtualization revolution.

OK, you’re a marketing professional and you’d like to start talking to these people, because you have something they’d be interested in. (Exactly how to approach a community of technologists is a topic for another time.) You might start by searching for influencers on the topic “VMware” at Klout. Let’s look at the Klout list for VMware and compare it to Eric’s hand-curated list. At the current time, there are a few overlaps, with @CXI, @daniel_eason, and @Mike_Laverick being the most notable. Certainly, if you were a PR professional and used this list of tweeters to start talking to people about enterprise virtualization, or to introduce a product that was destined for the data center, you’d be very off-track, and I’m not saying that just because I’m not included on the list! What’s gone wrong?

1. Keywords are not conversations. With Twitter’s 140 characters, often the general subject of your post is omitted.  Sometimes people will stick a #vmware in there, but we really don’t use hashtags outside of some events or special conversations. We are usually talking about VMware’s products rather than mentioning the name of the company in our tweets. Looking for tweets with VMware in them, which is what I assume that Klout is doing, catches only a fragment of the actual conversation. Compare that with a Topsy search for VMware, which looks at tweeted link destinations, and is dead on.

This observation was driven home to me as we looked at the results of one of our corporate announcements a while back. I watched a lot of conversations that morning in my normal twitterstream — actually quite a bit of back and forth considering how hard it is to keep a dialog going on Twitter. Point, counterpoint, argument, counterargument, predictions, reactions, positioning, counter-positioning — and a few poison pen letters. Our agency also did a search for a few keywords (company names in this case) – which showed a few hundred tweets and some mostly-positive pronouncements about the industry. The keyword search had missed much of the dialog in the community that day, because most of that dialog had moved past the keywords we were searching for.

2. Influence is not communty. The beauty and frustration of Twitter compared to forums or IRC channels is that, instead of having separate places to discuss separate topics, Twitter is a mess of overlapping conversations and overlapping communities. I think by relying on keywords, Klout has netted fish from several fishing holes.

In the current Klout list, there are actually quite a few people participating in the conversation on cloud computing, like @samj, @jamesurquart, and my new colleague @wattersjames. The Klout Score has a Network component, which transfers the influence of your followers and retweeters. Evidently the cloud conversation has a lot of influential mavens who have lifted the influence of people participating in the cloud computing community.

Another group of tweeters in the Klout list are people that work with VMware Fusion on the Mac. Mac users are obviously another community with lots of heavy hitters in the social media space. The Mac conversation is usually very separate from the enterprise data center conversation that I’m involved in.

3. Influence is not authority. From the explanation of the Klout Score, it looks like the score also considers reach and some cultural factors (how often people are retweeting & using lists). Nobody in our little enterprise virtualization community has hundreds of thousands of followers – I’ve got about 2K and @vmware about 6K, so I’d say those are the upper limits right now of our active community. I’ll speculate that our community is more close-knit and less self-promotional than some others I know — I’m aware of other communities that do a lot more logrolling and self-backpatting than ours does. That may affect our influence scores.

So for all these reasons and probably some others, a Twitter-wide influence score doesn’t adequately reflect authority within our community, nor can the participants even be found with a simple keyword search.  (Hmm. Sounds like a job for lists?)

tl;dr Keyword searches don’t identify communities on Twitter. Influence scores don’t correlate with authority.