-
Website
http://coldacid.net/ -
Original page
http://coldacid.net/blog/2009/04/09/hashtags-are-dead-long-live-real-time-search-and-filtering -
Subscribe
All Comments -
Community
-
-
Top Commenters
-
Popular Threads
#bushfire gives me more relevant results than just bushfire.
#police will be about police forces and police behaviour or news, but 'police' without the tag will pull up many other kinds of items such as 80s rock bands.
It all depends on the context and what you are hoping to find in your search.
I am new to this, and possibly missing something, but my current and short experience of just over a month is as follows:
Time being at a premium, I'd prefer a perhaps incomplete result of 10 relevant tweets, rather that 100 that I wont be able to wade my way through anyway.
Thus the value in a hashtag resides in a specific nuance or clarification of intent in the users post: choosing to hashtag it is a declaration of topic, where topic interest uis likely to be shared with a specific community of practice or interest.
That is where hashdictionary comes in.
And with in-line tags (e.g. #tag) it can be used anywhere you type, effectively threading things together.
Relevance and threading are why they will only become more powerful.
Like, for HealthCamp Boston we're using #hcbos. Fine with me if ppl use just hcbos. Otoh, the # does clue newcomers to what we crowdsourced as the (hashless) tag. And Tweetdeck will detect and auto-include the hashed tag in a reply.
So yeah, we don't NEED it need it, and my searches don't use it, but I sure do still use it. Am I missing something?
I'll start off by saying that Robert Scoble has yet to actually blog about this; he might do so still, but the FriendFeed discussion continues (in fits and starts) and he did state he might not bother posting after I got mine out. I hope that Robert does actually write up a post about this (or at least comes here and replies to some of your comments), because he does a much better job of this blogging thing than me.
I agree there's still a lot of use left in hashtags, and that sounding the death knell for them is a bit premature at this time. Still, the death of hashtags is certainly on its way, as methods of searching and filtering improve. For Chris Peoples' Lakers example, hashtags will become obsolete in that case due to the development and growth of semantic search technology. It'll help with Yvonne Thompson's examples as well.
With the growth of new services and new technologies, we'll see hashtags dwindle into insignificance in over the next two years or so, not instantly. But their days are numbered.
One thing I forgot to mention in my Tweets (and it's easier this way anyway): hashtags are kind of like team jerseys. When you're playing soccer and all of your players are wearing blue, and the opposing players are also wearing blue, you sort of get disoriented when trying to pass the ball. But if you're wearing orange instead, without thinking about it, you can make a killer pass and score a goal for the team.
I probably dragged on that metaphor too long and the first sentence would have sufficed, but my point is, visually, it's easier to sort Tweets when they have a hashtag appended. Even if you're using something like Tweetdeck that sorts things for you... if you're catching up on a favourite person and you're reading her Twitter page, you can glance over the text and ignore (or zone in on) the hashtag a lot easier.
Because this is folksonomy we're talking about, it's not an exact practice... that's the problem I have with hashtags. I was following the Mesh conference in Toronto, and some people were using #mesh09 while others were just using #mesh. So I used TweetGrid to sort both and probably missed very little of what was going on. But of the people who weren't using *either*, I have no idea...
Also, when hashtags are unnecessarily long and take up precious character space, it's also annoying. Maybe what Twitter should do is disregard anything that comes after a # in the character count and keep it invisible... and when you do a twitter search for a key word, it comes up with all that invisible text after the #s? That would be a good future for hashtags, I think... the same way you tag blog posts and it all goes in meta...
Even despite Twitter's flaws, such as not tracking conversations in any really meaningful way and making people have to include tags right in a post, they're helping with this too by providing comprehensive, real time search technology.
Let me posit something here. For Mesh 2010, you won't need to worry about people using a hashtag or not using one. Location services (like BrightKite or Google Latitude) will give a fix on a person's location, so when they're posting about a session they are attending, their post will be automatically cross-referenced with that session. Then, when you go and search for posts about Mesh, you'll get those posts, even though they don't explicitly mention Mesh at all.
It's because we can draw on this information when it is created, and check it against other information, and build the relationships on the fly, that hashtags are becoming obsolete. They're not quite dead yet, but it's only a matter of time (and not a long amount) before they won't be needed any longer.
And isn't the semantic web built on tags, essentially? You add information in the code about location, attribute, topic, etc., and that's what powers the intelligent search.
Re: Your reply to Phil Baumann: Blackberrys and other devices will evolve their SMS functionality to include metadata eventually. If the future of the web is increasingly going wireless, it's the next logical step. Hell, smartphones today already have more powerful hardware/software than the laptop I'm currently using.
This accomplishes the task of hashtags (to specifically mark a tweet, versus just as a searchable term).
Until Twitter provides a tagging feature, hashtags will still have their value. It's in Twitter's interest to have metadata on tweets. And ours.
This isn't just about Twitter, by the way, although everyone seems to think that it is. Twitter's old, metadata-poor model makes hashtags useful there, but for many of the newer services, it's not too hard to draw correlations between data from several services and figure things out from there.
Therefore, I proudly declare: blog posts that pronounced the death of internet behaviors are dead!
Yesterday, for example, I made lamb meatballs, took pics of the process then posted them on Flickr with the recipe. I then Tweeted the link and used #food. Before posting my Tweet I compared a Twitter search for "food" with one for "#food". The latter was far more narrow in scope so I included it in the Tweet.
This seems trivial, after all how many people really care what I cooked for dinner? But I felt that foodies might and if there were such people out there who regularly used and followed the #food hashtag then I might as well use it to reach those who already have an interest in food related Tweets. The same principle can be applied to more important Tweets.
So to me, #hashtags can be used as a way of crowd sourcing Tweets on a specific topic. Whether one is looking for Tweets regarding a certain subject, or whether one wants to reach that particular audience.
Natural searches are great as well, but #hashtags are like a filter that help us to narrow the focus. Each has it's time and place depending on the topic, but I like to compare each then decide which is most appropriate to my situation.