Last Friday MX ran a story about Metro plans to shut down its SMS text alert service, in favour of pushing people towards the mobile web site and Twitter. This was sparked by an ad earlier in the week in MX, station posters, and an email sent to SMS subscribers, all encouraging people to “say goodbye to SMS messages that fill up your inbox and hello to the Metro mobile web site.”
On Monday, in a followup story, Metro denied they were shutting down SMS.
Interestingly, they said people “preferred real-time information, not possible via SMS but available on Metro’s mobile site“. A surprising argument, given it appears (from the outside) that both get updated generally simultaneously.
Now, I like Metro’s mobile web site… it’s great for seeing the state of the trains, and (unlike the Twitter feed) it includes individual service cancellations and delays. I certainly recommend people have it in their phone. It’s particularly useful when travelling on different lines and at different times than you have set up for SMS alerts.
But you have to go and look at it, and it’s only “real-time” if you happen to be looking at it when something happens.
In contrast, the SMS alerts tell you that something’s happening, and although some people have complained of the late delivery of alerts (which could be Metro sending them late, or issues with phone providers), I think most of them to get to my phone in a timely manner.
How timely are Metro SMS alerts?
As it happens, I have a lot of SMS alerts from the past few months in my phone. I’m not sure why, but I haven’t deleted them since July. So I thought I’d do a quick check of the time of the train being cancelled vs the time the message arrived. (The “arrival” time appears to be the time the message was received by my phone provider for me, or perhaps the time it was sent, as some times appear early in the morning when I know my phone wasn’t on.)
Leaving out alerts that are less time specific (eg general delays to services), my sample is 50 messages over 2 months (early-July to early-September) covering the alerts I’m subscribed to, which is morning trains from Bentleigh into the city, and evening trains going the other way.
Of the 50 messages, 35 were cancellations, 3 were “will now run” (eg negating an earlier cancellation) and 12 were about Loop trains altered to run direct to/from Flinders Street. There were none about delays to individual services (these are normally only sent if the delay is 15 minutes or more.)
Overall 37 (74%) were received before the relevant event — with a median of 52 minutes. 13 (26%) were after the event, with a median of 8 minutes.
Counting only notices of cancellations, there were: 30 (86%) of which arrived early, 5 (14%) late.
Of the 13 received late, 9 were in the afternoon peak, reflecting that some cancellations and alterations are made at the last minute (eg due to a fault found in the loop-facing cab), and my alert is for Flinders Street departures, so there’s often less notice. Headed to the city, Bentleigh is about halfway along the Frankston line, so there’s probably often more time for passengers there to receive alerts.
Of the 13 received late, 8 were loop diversions (a Loop train altered to run direct), and most in the afternoon peak. Doesn’t make it any less annoying of course, if you were waiting in a Loop station to catch a train — some forward notice means you can try and get out of the Loop to Richmond/North Melbourne/Jolimont and pick up your diverted train there. In some cases they’ll make announcements in the stations as well, of course.
Morning vs evening
There were 13 morning cancellations. None arrived after the departure time at Bentleigh. The median was 104 minutes ahead of the event. The lowest was 31 minutes ahead. (This concerned a 7:31 departure from Bentleigh, which originated at Carrum at 7:02. Those waiting at the first few stops got almost no warning, but they had the option of catching an express scheduled 4 minutes later.)
So while being halfway along the line on the way to the City helps for me, it appears every morning cancellation alert was received before the service was expected to run, sometimes hours before, so people could plan around it.
So my conclusion would have to be that while some SMSs are received too late (and on-time delivery is not guaranteed), most are on received time and provide genuinely useful information about disrupted services.
It’s costing Metro a lot of money of course. SMS alerts have 25,000 subscribers according to Monday’s article. I had heard a figure last year that it was costing them about $50,000 per month to send the alerts, which is a lot.
But it’s a valuable service, and as noted in Friday’s article, the best way for them to cut costs is to reduce cancellations and delays.