I’m still not completely convinced by hashtags on twitter. On the plus side, they can make following what people are saying about a show (#bbcrevolution) or an event (#iod2010) easier. On the other hand, these are a few random thoughts about the downside to hashtags…
Hashtags are common property, which is not a problem when people are cooperating to join threads of conversation, but it’s easy to see how that could #fail:
- I might think #iod is a great tag for Information on Demand, but there are plenty of others who think it means something else.
- An ‘official’ hashtag can avoid some of the confusion but, even if you manage to stake your claim to something unique enough, you can’t control it. While #bbcrevolution was talking about denial of service attacks, I was thinking about how easy it would be for anyone to launch a denial of hashtag attack. Not to mention when marketing tags get hijacked.
- If there isn’t an official or obvious hashtag for something, it’s easy to end up with multiple hashtags. A bit of discussion can usually get things on track, but it always seems a little odd talking about the tag, rather than the subject of the tag.
And finally, the situation that got me thinking about hashtags in the first place. When different people use the same tag for almost the same thing, especially if one of those uses is much noisier than the other(s). In this case, tweets related to the Current Cost meter, and automated Current Cost meter readings have both used the #currentcost tag. Not a huge problem, except it would be easy to miss interesting information if the tag was swamped by even a few tweetjects posing meter readings:
“seems a shame that @mmnHouse is inserting the #currentcost hashtag to their house temp and elec reading. creates major noise” @yellowpark
“thinking we need a new hashtag for #CurrentCost stuff: one for bots and noisy automated stuff, another for discussion. what do people think?” @dalelane
“@dalelane How about #CurrentCostData ? I agree my searches are becoming muddled with people’s bots, and not information on #currentcost” @cumbers
“@dalelane Agreed – It frustrates me no end having countless #currentcost tweets popping up all day!” @markphelan
“thinking we should use a specific hashtag for tweeting #currentcost data to avoid creating noise. any suggestions? #ccdata ?” @yellowpark
“Moving from #currentcost hashtag to #mymeter with a data format for auto graphing.Join in discussion at http://is.gd/7b0si (via @ScaredyCat)” @stuartpoulton
So one solution is to agree on uses for #currentcost, #currentcostbot, #ccdata, #mymeter, etc. which is likely to work reasonably well for the Current Cost audience, but it may not be as practical for every situation.
Alternatively, as well as being able to mute retweets, it would be handy to be able to mute selected people using a hashtag in a way you’re not interested in. Even better if lists could be muted: if I could mute any tweets containing the #currentcost hashtag from anyone in my @jtonline/tweetject list, this problem goes away.
So, a fairly random collection of #thoughts on #hashtags. What are yours?
Thanks for posting this. It’s like the post I should have written.
Hashtags have been less useful for me for the last few months. Over used by many and abused by some. I struggle personally to gain value from #hashtags.
From a commercial point of view, a pure live stream is a no no. A spammer will surely come along and post something fruity.
So that aside. The #currentcost hashtag is it’s own animal. It lives, therefore it tweets. Some good things. Some bad things. Either way, it is a major indicator of the ecosystem surrounding Current Cost monitors. It is a solid indicator for any brand.
Controlling the hashtag is impossible. But shepherding is actually fine. It is very understandable that when someone starts graphing and tweeting their energy usage that they use the #currentcost hashtag. I remember how excited I was when I could graph and tweet my electricity reading. So occasionally we get noise, but a diplomatic ask will normally port to a more relevant tag like #meterme #currentcostbot or the like.
For me, the major thing is BE NICE. And I hope others will BE NICE to me.
Shepherding certainly seems to be working with the #currentcost hashtag, although I wonder if that’s the exception rather than the rule. The Current Cost tweeps are mostly pretty friendly!
I’d not really made the connection when I posted, but I don’t even use the #currentcost tag for viewing- I just look out for currentcost without the hash. So maybe I get even less value from hashtags than I imagined.
Still, the hash is a clue to a shared term which can be helpful, and short-lived tags wouldn’t suffer so many problems, so maybe #flashhashtags #ftw but #statichashtags #downwiththatsortofthing?
Your snap of tweets inadvertently implies that that Chris Dalby’s (aka @yellowpark) complaints resulted in the moving of us ‘tweetjects’ from #currentcost to #mymeter.
A small group of people who have been using current cost electricity monitors and various other arduino and xbee based monitoring have joined together to build a simple tweet -> graph interface. The #currentcost hashtag doesn’t fit into that because it’s too specific. We wanted to be vendor agnostic.
The fact is Chris didn’t even ask if we would drop the #currentcost hashtag until after we’d moved to #mymeter.
For the record there was no ‘shepherding’ and Chris certainly didn’t have input into the resultant #mymeter hashtag – something we’d planned on moving to since May of last year (time is the enemy there).
I’m sure everyone is happier now and at least your stream of “nearly 1 million units sold” and “only 80k more units till we get to 1 million” wont get clogged up by user data.
SC
They were a pretty random selection of tweets I spotted while this post was festering in my list of drafts, so I’m not surprised they don’t tell the whole story. They were just to illustrate that you cannot control a hashtag, which could also be a problem with #mymeter. Perhaps less so though since #mymeter tweets don’t appear to be for direct human consumption.
Given my general skepticism about hashtags in the first place, I’m interested in the idea of #mymeter. Having a vendor neutral tag makes perfect sense, but I’m not as convinced of the value of just tweeting the tag and a bunch of data. One comparison is the #uksnow hashtag, but that’s very human oriented with subjective ratings, and comments, for a short term (hopefully!) event, so a #hashtag works quite well. None of the recent #mymeter tweets seem to have any human element to them, so aren’t there better alternatives for sharing the #mymeter graph data? Pachube or makeHistori spring to mind.
I appreciate you can’t control a hastag, but if your #mymeter data doesn’t parse it’d be rejected not only that but since you know who the user is it’s trivial to filter bad data from processed results or even ignore them completely.
I disagree that #mymeter tweets have no human element to them, they are very easy to interpret and we’ve even done some basic documentation on it http://wiki.homehack.co.uk/index.php/Homehack:Dataformat – just looking at a line of output you can see exactly what’s being said.
In building the idea we wanted something that was simple for both humans and machines to read, personally I think we have managed that.
Pachube is fine, I use it myself, but it doesn’t provide me with the ability to compare with people I’m interested in. People who, like me, might have more than one pc running 24×7, people who might be experimenting with how to save energy. #mymeter and the eventual supporting site(s) will allow you to compare between friends and like-minded people. Pachube doesn’t allow me to run off a quick, one off comparison if I want to.
If you’re a developer you’re also free to pull the data from twitter and do your own thing. Twitter ends up being a short term store and a very simple interface to publishing data which pretty much any device can push to, even via SMS for automated equipment.
There is absolutely no reason you couldn’t push data from #mymeter to Pachube if you wanted to but then you would be reliant on Pachube. Twitter can easily be swapped out for other services (identi.ca for example, nice open source version of twitter) if that ever disappeared.
The data format can be interpreted by a human, but I wouldn’t say it makes for very readable tweets. If they are primarily for automated processing, then noise on the hashtag isn’t much of an issue, but it could still disrupt human consumption.
I think for me, services like pachube, rsmb, etc. are more suited to plumbing the data (e.g. from several people you’re interested in to a tool to do comparisons on the web, on a phone, in a virtual world, etc.), while twitter is a great conduit for processed information (e.g. “Using more gas than the average 2 bed home #mymeter”).
Will be very interested to see how things go with #mymeter though. (Btw. tagdef looks like an attempt to add more meaning to hashtags – might be worth adding something for #mymeter?)