for the most part, we've all gotten use to texting to people we know who are subscribers on networks other than our own; in fact, in many cases we never know what network the receiver of our message is on, nor should we. just like a voice call, right?
unfortunately, this is not the case for MMS. right now, in order for me to send an MMS to another person's phone i have to a) know they are on the same network b) know they have an MMS capable phone and c) know they've signed up for the service. "c" is especially annoying. as i do carry a camera phone on most occaisions, i have tried more than once to snap a picture and send it to a friend who i know is carrying a sony ericsson t68i, capable or receiving MMS, but i get an SMS back saying "the subscriber you are sending this message to is not capable of receiving the message, it will appear as text". 2 minutes later i get an SMS back with "what was that"? and sending MMS from, say t-mobile to an att wireless GSM MMS capable phone, fogettaboutit.
this is all revenue being left on the table and operators know it. even in the in-network case i reference above...the target subscriber should be given a taste of MMS (say they get 5 for free) and then be texted a link from which they can provision themselves over a wap connection or by replying to the SMS or MMS. at very least, the MMS should be routed to their webmail inbox automatically for viewing, with an SMS notification sent.
in Hong Kong, when cross carrier SMS messaging was allowed between the six carriers at one time, Hutchison alone claimed its SMS traffic increased by 5X off the bat and is still increasing. with MMS currently drawing higher margin revenues than SMS, this is definitely a case where operators working together will benefit the whole, and the quicker intercarrier MMS is implemented, the sooner we all can enjoy the benefits.
inter-carrier MMS has been demonstrated before, in both GSM-GSM and GSM-CDMA scenarios. there have been standard high-level messaging protocols (MM1, MM3, MM4, MM7 ) definined by the OMA and referenced by 3GPP that have made this possible, although i know from working with the myriad MMSC vendors that the production availability of these protocols has been on a "Just In Time" basis, often with development teams scrambling in the background to meet an implementaion deadline or lab test.
now, the GSM Association has published their MMS Interworking Guidelines, which as the title suggests are the preferred mechanisms for operators to get this working in a secure and efficient manner. the thing that stands out to me after reading this document is, this is not trivial!
that is, MM4 in itself is pretty trivial, and one would probably have little trouble getting 2 MMSCs with ESMTP provisioned to pass MMS messages between themselves, but when you impose the carrier environments on top of this, some of the complications start to become apparent.
first all the carriers have to be connected somehow, and in a secure manner. Security of messages is not a primary concern. Spam is a primary concern. the easiest ways in which MMSCs can be setup to communicate also make it easier for spam MMS to get propogated on the network, and in this case, the end user would to pay for receiving the unwanted mail, which in addition to being a nusiance causes all sorts of billing problems. GSM recommends that instead of setting up new VPNs between carriers, they instead use the existing GPRS Roaming Exchanges, which would require an upgrade, but make the task much easier. The fallback is to create secure VPNs, which would seem to be necessary for GSM-CDMA messaging.
the second thing that complicates things is Mobile Number Portability. They highly discourage keeping static tables of which numbers belong where, as it simply doesn't scale - so they suggest some other ways to deal with how to route the message either by phone number or email address as subs start to move around from network to network.
finally, as different operators have different billing models for MMS ( some charge per message, some charge for the bandwidth consumed, some both! ) there's no clear way to settle intercarrier message costs. my guess is that some 3rd parties will pop up to connect the carriers and settle the bills, all for a well deserved chunk of the pie.
one thing is clear - with only 2 percent of the deployed handsets being MMS capable, operators need to come up with some creative ways to tease users into upgrading, with usability of MMS->SMS being key. there are definitely transcoding techniques such as using wap push i've seen that can help convert some of these users, as well as using combinations of SMS and web based email to get people interested. This uptake has to happen in parallel with internetworking of MMS for operators to get the ROI they've put into their MMS infrastructures.
as a shameless plug, our consulting company offers services in helping operators sort out MMS issues such as this.
Posted by Steve at September 1, 2003 10:33 PM | TrackBack