Greg Day made a good point the previous week which could explain part of the utility of this morse-code type texting. People who use texting a lot can touch-text. That is, they turn predictive texting off, and without even looking at the cell-phone, they can send messages. Therefore, for them, this type of input is very natural. Similar to how I find the Qwerty keyboard very natural just because I have learnt to touch-type.

Every new thing has a cost and a benefit associated with it, and that includes the cost of change. In the case of texting, for many people the utility of being able to send a short message “good luck” before an exam, or “10 min late” before a meetup (or however you do it in text speak) instantly and directly to a person outweighs the cost of messing around with the keyboard. And let’s not forget that the technology is very simple to learn. Even my mother got her new cellphone and sent my sister a message without opening the instruction manual. With something like Xero, the cost of trying the system out and potentially not liking it is sufficiently high that many people won’t try it, at least until they have been told enough times by friends how great it is and have been seen it in action.

Perhaps my point isn’t very clear. What I’m saying is that you need people to be able to make a small step that has an obvious benefit to them. Then they will take it. One there, they might then take another. And so on. Essentially, people act just like local gradient based optimisation algorithms much of the time.