A New Metric for Chat
It’s time to pay attention to attention
There are few things people do online more than chat — as Kleiner Perkins’ Mary Meeker recently pointed out, messaging apps now count their user numbers in billions — but the industry hasn’t had great ways to measure that activity. Traditionally, the owners of chat apps have focused on message volumes, message frequency, or time spent in app, among other similarly rigid metrics.
As we have seen chat behaviors evolve, at Kik we’ve become less interested in how many messages are exchanged or how many times an app is opened, and more interested in how people engage in chatting. Put more simply: we care more about attention than app taps.
In recent months, we have developed a new metric to gauge a user’s behavior in chat. We call it the attention metric. It focuses on people’s chat sessions. By looking at the time taken between messages sent, we have found that chat sessions (which may include multiple conversations) can be grouped into three key categories: active; passive; and sporadic.
An active chat session is one in which the chatter is fully focused on a conversation and will send a message to her correspondent within 20 seconds of the previous one. For instance, an intense back-and-forward between girlfriend and boyfriend would qualify as an active chat.
Someone in a passive chat session, meanwhile, is not so engaged and might take several minutes to send another message. This is the sort of on-again, off-again conversation you might have with a friend while you’re slightly distracted, perhaps while watching TV or cooking dinner.
A chatter in a sporadic chat session will just check and send messages occasionally throughout the day.
We’ve found that the attention metric accounts for real-world chat behaviors that aren’t well represented by traditional metrics. It captures multi-tasking and acknowledges that a conversation can be ongoing even when the chat app is closed. It also accounts for individual chat styles. Here are some examples of what we’ve found since employing this new metric:
- The average U.S. teen on Kik engages in 6.1 chat sessions a day, at 12.7 minutes per chat session,
- U.S. teen females on Kik chat 35% longer than U.S. teen males
- U.S. teen “super users” (who make up 10% of Kik’s user base) engage in 10 chat sessions a day on average, accounting for 157 minutes of daily chatting
- Kik users are more engaged in active chat during the week, but they have more chat sessions during the weekend
- Overall, Kik users spend about 21% of their time in an active chat state, and 79% in passive
We are finding the attention metric a useful way to capture a lot of chat behaviors that other metrics miss. Engagement on the web has traditionally been measured in session times, which are assessed to expire after a certain period of inactivity. With apps, however, the expiry point has been clear: it happens when the app is closed. That method of measurement has worked for games, and things like news feeds. But with chat, the behaviors become much more fluid, and the opening and closing of an app fails to capture what’s really going on.
As the internet evolves, so too must our metrics. It’s never been more important to pay attention to how people are paying attention.