There seems to be a great deal of divide over where, when, and how hashtags should be used on social media services like Twitter.
So – what’s the best route to follow? When is it ok to create your own? Why would you want to use them at all?
Why use hashtags?
Dating back at least to the heady days of IRC, hashtags give a convenient (if implemented) way to finding related content to what you are looking at now. (In IRC they are used to call attention to a user and/or name a channel.)
For example, on Twitter, if you search for, say, “#ObamaCare“, you’ll see a variety of recent tweets that talk about or reference the “Affordable Care Act”. If you want to join an ongoing discussion, it’s a good way for people to be able to find you and what you think.
When make new ones?
Some folks seem to get a thrill out of hashtagging everything they say. Like this: “#crazy #TSA #waittimes #patdowns #cavitysearches #whowouldeverwanttofly?”
So when should you make new ones? When you’re tweeting something that hasn’t been before, or you want to repurpose an old tag, or some other Really Good Reason™.
For example, when I was at Moab Con in 2011, I live-tweeted many of the sessions, appending or prepending (depending on where I thought it should go) #moabcon2011 to my tweets or folks could find and follow easily. Or when a family we knew was doing an adoption fundraiser 5k, they used “#r2b1h” for “Run to bring 1 home” (which, very excitedly and unexpectedly has now turned into bringing 3 home!).
Where do they go, then?
The general “best practice” (though I despise that term) seems to be to tag inline, and append additional tags at the end (and post any links that may be in the tweet). For example, you might tweet a news story thusly, “#Amtrak considering fleet replacement and new stations http://t.co/somelink #train #transit #rail” (story and link made up).
Overall, hashtags should not be repetitive (eg “#Ubuntu #Linux 13.10 betas available http://t.co/somelink #linux), and every tag should add something to the post – the first tag tags the entire post, and adding more makes the post harder to read, and adds no semantic value to it.
This same concept applied when tagging blog posts, question on sites like the Stack Exchange family, etc. In many ways, it’s no different than the old card catalog at your local library. Tags are like a selective concordance – there may (or may not) be a reason to have a truly exhaustive concordance for something (eg Strong’s Exhaustive Concordance), but in all likelihood, you don’t really care every time someone uses a simple word like “the”. You care to find the meatier terms in the text.
Maybe we can all enjoy SMS to the masses a little more now.