nytimesWe’re still very early into The New York Times’ paywall experiment, but research company Experian Hitwise is already collecting data on how the move has affected traffic to NYTimes.com. Its finding? Traffic is down.

Specifically, the company looked at the site for 12 days before and after the paywall went up, and found that traffic was down almost every day afterward. Visits were down an average of 5 to 15 percent, while pageviews declined 11 to 30 percent. The one exception was April 9, when the number of visitors was actually up 7 percent, presumably because readers wanted to find out about the narrowly-averted government shutdown.

Now, it’s way too early to say anything definitive about The Times’ traffic, especially since Hitwise was comparing relatively short periods of time, but the fact that the numbers were down consistently suggests a significant change in reader behavior.

If there is a change, it’s not exactly surprising. A website doesn’t put up a paywall because it wants to increase its readership but because it wants to make money from the readers it has. A drop in traffic is only natural, and it doesn’t say much about the numbers The Times really cares about — how many readers are signing up for a subscription.

The real danger is if the decline continues. The Times has said that the vast majority of readers should never encounter a paywall because they can view 20 articles per month for free, and they can also follow links from search engines, social networks, and other websites without being blocked.

The challenge is to remain open to visitors while also convincing a dedicated minority to pay. If, instead, The Times is frightening casual visitors away because they’re worried about the paywall, or if the more frequent readers choose to read less rather than pay, then the company may want to rethink its plans.