Only 27% of Tweeps Use Clients–What’s that Mean for Your Tweet Strategy

Last week, I was reading @waworld’s stream, and I saw this tweet:

I was surprised at the number. In the late spring, I’d seen another statistic that showed almost a 50-50 split between those using the default Twitter site and those using one or another Twitter client. When I went to the CrowdSource report, the picture was a little more mixed. It says “43% of Twitter users employ a third-party application at least some of the time, and 26% as their main mode of access.” The graph in the report does seem to show that almost three quarters of people access primarily through the website.

This has some implications. I’m wondering, in particular, about repeating tweets. I’m now promoting my product, TweetPackage, using Twitter. I repeat tweets on those accounts. I also use my personal account to pass on relevant links when I find them. And I promote my own blog posts. When I find a good article, I usually tweet the link several times over the course of a week. Likewise with my own posts, particularly if I think they are really, really good ones.

But my understanding is that people who use Twitter clients often don’t like these repeat tweets. (Because of my usability bent, I still access Twitter through the website for the most part, in order to stay closer to the experience of novice users.) They create groups or columns containing smaller numbers of twitterers or tweets with certain keywords, in order to better facilitate viewing. That means they can put me in a category, say Marketing Info, with maybe ten other people, and go to that column when they want to see the latest tweets about marketing. If I repeat my tweets, they are more likely to see those repeats.

For people using the default UI, it’s less likely they will see the repeats. It doesn’t take many follows for your stream to get thick. Tweets flow by quickly and you aren’t likely to see the repeats.  Even if you’re not following a lot of people, if you check Twitter only a couple of times a day, you’re likely to miss a lot of tweets.

This presents a dilemma for people trying to promote themselves, their business, or their content. They have to tweet the same item more than once in order to have any chance of being seen. But people using clients will get annoyed at the repeats. Likewise, there are Twitter addicts who are checking their stream every hour or feel compelled to read the entire stream every time, so as to never miss a tweet. They too will see repeats.

Where does one err? On the side of the addicts and client-users or the side of the newer, less frequent users?

Given the latest data, I would say err on the side of more repeats, unless you are fairly certain that your followers are made up of more addicts and client users.

What about Twitter’s repeat policies, though? For the time being, I think they warrant a reexamination.

Twitter doesn’t like repeat tweets and will block them. If you, like me, use a site that allows you to schedule tweets, you can’t repeat the same tweet too often within a short time period or they will be blocked. Some tools offer features to help avoid this—like rotating tweets—allowing you to enter several slightly different version of essentially the same message, in order to avoid being blocked.

Several factors require that Twitter rethink how it handles this situation.  First, there’s the change in the number of client users, with all the implications discussed above. Second, there’s the expected change in Twitter’s demographic as it becomes more mainstream, to a larger percentage of people who will only view the stream intermittently. (The same demographic change may also have implications for search.) Third, there is the fact that Twitter—rightly—sees business as playing an important role in its future.

The latter means that Twitter needs to carefully balance the need of businesses to effectively promote themselves against the need of consumers to not be inundated.

Luckily, I think there are technical solutions to the problem. If they haven’t already, the clients need to implement some simple filtering based on whether a twitterer is in a group or not. If I’ve put Joe in a column in TweetDeck and Joe tweets the same text several times in a given time period, TweetDeck just doesn’ t show the repeats.

Likewise, if Joe uses different text in several tweets, but they all contain the same link, TweetDeck doesn’t show it.  (The clients will need to be smart enough to resolve short links behind the scenes to check the destination. Otherwise, clever spammers will use several different short links all going to the same URL.) Twitter will surely be adding groups or channels in the future. When it does, it needs to handle things similarly.

Twitter and the clients can go further even, by checking for addict behavior. If the user repeatedly clicks More at the bottom of their stream until they reach the point that was at the top of the stream last session, they’re exhibiting addict behavior. (Yes, this requires monitoring and storing some data about the users’ sessions.) Likewise, hourly access (or some other determined frequency) would suggest an addict. Those people should have repeats automatically removed from their stream, or at least have the option.

I think this is an example where technology can resolve the proble of conflicting requirements, and make all parties happy: average Twitter users, addicts, and even businesses.