Six Twitter rules

Consciously or not, everyone has a set of guidelines for Twitter that govern both how they tweet and what they’re willing to tolerate in the tweets of those they follow. Most of these rules have to do with content. Political tweets, religious tweets, anti-religious tweets, what-I-had-for-breakfast tweets, hashtag-of-the-day tweets—any of these can make you beloved or instantly unfollowed. I have content rules, of course, but I also have stylistic rules.

Use Twitter’s own automatic link shortening

I’ve mentioned this before, but it’s worth repeating. Twitter shortens all links, regardless of length, with its own t.co service. There’s no advantage anymore of using bit.ly or j.mp or any of the dozens of other services that popped up like dandelions during Twitter’s pre-t.co days. In fact, there are two distinct disadvantages:

  1. Your links will include two levels of indirection. First to t.co, then to bit.ly (or whereever), and then, finally, to the site you’re actually linking to. Giving Twitter the original URL and letting it shorten it reduces the indirection to just one level. And because t.co links include the original URL in the tweet entities metadata, Twitter clients can, if they choose, go directly to the original URL with no indirection whatsoever.
  2. Your link won’t look as nice to your followers and won’t be as informative. With a third-party shortener, your links will be obscured.

    Rails 3.1.0 is out w/ asset pipeline, HTTP streaming, jQuery default, and a trillion other cool new things: bit.ly/qhFLMk
      — DHH (@dhh) Tue Aug 30 2011

    But if you let Twitter do the shortening, your followers will get a sense of where they’re going if they follow the link.

    Adapting to Tweet Entities and t.co-shortened links: leancrew.com/all-this/2011/…
      — Dr. Drang (@drdrang) Mon Aug 8 2011

It’s not surprising that spammers always use third-party link shorteners. They don’t want you to know where the link leads.

I suppose social media professionals will always want to use third-party shorteners because they need the click statistics those services provide. This puts social media professionals at the same level as spammers, which seems somewhat unfair, but only somewhat.

Use native retweeting

I know the objection to native retweets: they don’t allow you to add your own comment. I had the same objection when Twitter introduced native retweets, but I’ve come to appreciate Twitter’s wisdom (!) in disallowing additions to a retweet.

When you RT and add your own bon mot, you’re not showcasing someone else’s tweet, you’re just using it to show how terribly clever you are. Be generous and let the other person’s words stand on their own. You can always follow up with your own tweet if it’s absolutely necessary. It seldom is.

You may note that two of the people I follow, Rob Corddry and Albert Brooks, violate this rule repeatedly, and it hasn’t caused me to unfollow them. True enough, but be honest with yourself: Are you as funny as Rob Corddry? Are you sure about that? (I won’t bother asking if you’re as funny as Albert Brooks because that would be impossible.)

Use Twitter’s own photo service

This one I’m a little less doctrinaire about. While I use Twitter’s service for linking to photos to tweets, I can understand the reluctance of some others to do the same. If you regularly archive your photos to a service like Flickr or SmugMug, reuploading a photo to another service is redundant and a pain in the ass.

But if you’re using TwitPic or yFrog, you’re not really archiving, are you? You’re just using them out of habit—a habit established when those were the most convenient ways to tweet an image. These third-party image hosts are like third-party link shorteners, once useful but now unnecessary.

The advantage of uploading your images to Twitter is that Twitter turns them into tweet entities, which makes them easy for clients to display inline. For a Twitter client to display images hosted at a third-party site, it has to include code to interface with that site. Many clients do that, but all clients can’t incorporate all third-party image APIs. Displaying a Twitter-hosted image, on the other hand, requires just one call to an API the client is already using. You’re more likely to get your tweeted images seen if you upload them to Twitter.

Don’t dot-reply

Dot-replying is sticking a period before the username when replying. Doing this puts your reply in the stream of everyone who follows you, not just those who follow both you and the person you’re replying to.

I hate this. It’s more self-aggrandizing than the RTing with an added comment. Just because I’m following you doesn’t mean I’m desperate to read every word you write. If I were interested in your conversations with other people, I’d be following them. And then I’d get the full conversation, not just your half.

.@deepgreendesign Aso too, George Burns: “Dying is easy. Comedy is hard.”
  — Thomas Levenson (@TomLevenson) Fri Mar 2 2012

I like Tom Levenson. I like his Isaac Newton book. I like his blog. But I stopped following his Twitter account long ago because he’s always doing this.

Don’t repeat tweets

This seems to happen mostly when people sign up for a web service and it starts tweeting their activities without them knowing about it. You may have just started with Service A, which sends out a tweet when you do something and simultaneously sends a notice to Service B. If Service B is also set up to tweet automatically, your followers will get two identical, or nearly identical, tweets in rapid succession.

Autotweeting isn’t to my taste, but autotweeting run amok like this is maddening. I understand how this can happen accidentally, but I can’t understand how you could let it go on for days. Don’t you see the repeats in your stream?

Don’t link excessively

This may seem like I’m complaining about content, but I’m not. Even if I generally enjoy what you’re linking to, I don’t want to see twenty or more link tweets a day. If I’m following you, it’s because I want to hear what you have to say; I don’t want an exhaustive detailing of your reading list. Most of them are going to be links I’ve already seen, anyway. You’re not the only thing I read on the internet, you know.

Update 3/3/12
On the same day I write a post that suggests I’m losing my memory, I write another one that confirms it. Overnight Daniel Jalkut gently reminded me, via Twitter, that he’d written a similar post “Elements of Twitter Style” about two years ago.

Daniel’s post is more concerned with content than mine is, and he offers advice on what would be considered traditional stylistic matters (avoid abbreviations; treat a tweet as a paragraph, a single thought), but there’s a lot of overlap when it comes to retweets and dot-replies. It’s hard to read Daniel’s post and and not come away thinking that mine is a response.

And yet, even though I read Daniel’s post when it was published and have been thinking about this post for a month or more, I never once put the two together. This is either an unfortunate consequence of middle age or an opportunity to increase my blog output with little effort. “Oh, did you write a post from the perspective of an anthropomorphized version of Apple’s brushed metal theme? I guess I’d forgotten all about that.”

Anyway, my apologies to Daniel, and I encourage you to read his post—it’s an interesting dip into recent history. Even though many of us had been tweeting for a couple of years by then, Twitter was still new to a lot of people, and he was trying to turn back the tide of ugly, illiterate tweets that was rising with Twitter’s popularity. I’d like to see him do a followup with two years of perspective.

In the meantime, I’ll be writing a post about how you should archive your email to clean out its inbox. Still trying to come up with a catchy title…


5 Responses to “Six Twitter rules”

  1. Dan Wineman says:

    Fully agree on all of these, especially the dot-reply, because it also suppresses the in-reply-to metadata that makes it possible for clients to follow the conversation thread (in the unlikely event that any of your followers are actually interested in doing so).

    One exception, though: it’s appropriate to dot-mention someone (not reply) when you want to start a sentence with their name, e.g. “.@drdrang’s Twitter rules are awesome.”

  2. Matthew Frederick says:

    Third-party link shorteners aren’t just for spammers and “social media professionals”, they’re for marketing in general. In my case I’m trying to grow my local iOS developer/Cocoaheads group. Knowing which tweets actually get people to follow the link and potentially attend a meeting helps me narrow down the bet wording and timing.

    Far from an evil task. I suppose I could write my own tracker and host our tiny website somewhere I’d get query string traffic reports (and not on WordPress), but my limited volunteer time seems better spent wrangling speakers and organizing events and just using Bit.ly/j.mp tracking. No?

  3. Matt Reed says:

    As to point 1, sometimes you’re forced to use a link shortener if you’re using a third-party phone or tablet client. I share your view on this one, but any time I’m posting from my phone, I have to use another shortener if I want the post to go through.

    As to point 2, I generally agree there too, but there are some times when added commentary is called for. Unusual, maybe even rare, but it does happen.

  4. Alex Reid says:

    I have very mixed feelings about the dot-reply. Happily, I see much less of it (and I’ve only done it once myself, I think). They don’t even work all the time, thanks to a Twitter bug.

  5. Ben K says:

    Dan: The unsightly dot leader can be avoided in that case by re-wording one’s message in such a way that the user’s name falls later in the sentence.

    Matthew: You might just find you get better uptake when readers can actually see the URLs you’re linking to.

    Matt: Clients that force the use of shorteners are broken by design.

    b