Hi again,

I just had a very interesting email question from Craig at Neustar regarding my presentation on mobile social networking. In short Craig wonder why I omitted talking about moderation and how to make it work with scale and still not ruin the business case.

It is a very valid comment. First of all it was omission by accident rather than purposely avoiding a tricky subject. If you have a time limit to stick to you have to cut some thing out. Thankfully there are ways to correct this in the digital world.

To some extent this particular question fits under the overall question that I did stress in my presentation: Can you serve the users with the service they and you want them to have in a financially viable way?

Let me divide the answer in two parts:

1. Public vs. Private

Is it a public text/picture exchange or is it between two persons in private? If it is public it should most likely be moderated.

If it is a one-to-one conversation it should not. It would be equivalent of wire tapping a phone. You need a warrant from a prosecutor to do that. Having said that, I have had several discussions with mobile network operators who have wanted to impose moderation in one-to-one exchanges. In all these cases I relate to the wire tapping and prosecutor approval to get them on board and forget about the moderation of one-to-one exchanges.

2. How to moderate in a scalable way

Public chat rooms are at one extreme of the scale. It is costly and requires some complexity in your system and interface to perform. My recommendation to a developer is to build the interface for real time moderation but put it on a third party or the mobile operator to perform the moderation. They often have a call centre for customer care and they can spread that work among their support staff in some cases.

If not there are dedicated firms that work with chat moderation and they work on a performance pay often with some kind of guarantee. The good ones will also stimualte use in the community and help diving the revenue side to some extent. In some cases you can make that a viable business case to work with a firm like these.

Also it is key to give your community members tools to wash out bad users themselves. We have built in a “block user” function in our services where appropriate. With them a user can block another user immediately if they decide they want nothing more to do with another user. In this case interaction between the two users in blocked. Both users are still live in the community.

Then we have the usual “report abuse” button where users can report to us about other users they feel should be blocked. Here we then make an investigation and make a judgement if we should block a person or not. The quality of the moderation you can do here depends on what you keep in your logs and how easy to use interface you have to the back end for these investigations.

For photos I can only see manual human batch processing working well at the moment. This means you have to give up real-time moderation. We pre-approve all photos that goes up in user profiles and photo galleries. We promise 24 hour turn-around and send an SMS when approval (or dissaproval) has taken place. Our experience is that the SMS communication makes it a very acceptable process. We keep the user informed about the status and they do not have to keep monitoring their account.

I trust this answer makes some sense to a very good question. Thanks again Craig! Keep asking more questions if you have them!