I can't remember the ACCC ever being quite so blunt and prescriptive. Article here
Decreeing across the board new advertising standards and requiring Telcos to measure and report actual speeds four times a year.
I'd like to think it will shake the industry up a bit but no doubt it will adapt. New speed 'tiers' will now be Standard, Standard Plus and Premium (not the most imaginative of names but workable )
Assuming the results would reveal what the public and Telcos already know - that peak hour speeds are generally well below the advertised tier speed - I'm not sure how what the industry response will be. Already Optus have had some plans that says in the fine print "Your connection may be slowed in preference of other users in times of peak usage." I can only assume that Optus and other Telcos will embrace this approach further. Each tier will have two flavours - cheap but one step down and throttled as needed, or a bit pricier and maintained. If the expensive one is not throttled then those able to afford it will be able to eat into the bandwidth of those on the cheaper tier.
Also interesting will be what happens when the Telcos fail to meet their promised speeds now. Will they bring in refunds on request as the ACCC suggests?
I'd like to see these new service tiers look like this:
Premium: Your top speed with be 100down/40up with a variation of no more than 20Mbps in times of congestion. Refunds will apply if congestion occurs.
Medium: Your top speed will be 50down/25up with a variation of no more than 20Mbps in times of congestion. A refund will only apply if speeds are lower than 15/10 during congestion peaks.
Entry: Your top speed will be 25down/15up with a variation of no more than 20Mbps in times of congestion. A refund will only apply if speeds are lower than 5/5 during congestion peaks.
Pretty easy going on the entry level @TommyB ? Delivering just 20% of what you promised considered acceptable?
I didn't realise but the ACCC has already written in depth on what it now wants. Using your bands its opting for a 10Mbps shortfall in peak, 20Mbps shortfall (for 50/25) and 40Mbps shortfall (100/40) - interestingly enough they've opted to ignore upload speeds in this. This is only a "preffered" model (so as @10000000000GB says be prepared for a dozen different variations on what is reported.)
The ACCC has also laid down a method for measuring actual speed (tested 4 times a year on a minimum of random 75 homes, oce each hour for 54 hours). On the surface it sounds pretty straight forward but I'm skeptical. Take the 'random' requirement. Seems to me the a RSP can (and probably does already) monitor speeds 24/7. So it could take 75 random houses and do a speed check. If it doesn't like the answer it takes another 75 random houses and does a speed check and so on until you get 75 random houses that indicate everything is working as it should be. Alternatively they could stick with the same random 75 houses and run 60 speed tests across the hour. Pick the minute that crosses the required threshold.
Ultimately, it seems more like a shot across the bow (The ACCC being unable to demand any of this). Hopefully it will be a lot easier for those customers gettting really bad speeds to ring up and get a refund or exit their contract without issue.
They can't hold car manufacturers to diesel emission standards.
I wrote an enquiry to the ACCC, to ask them how they will prevent the RSPs from detecting the testing and giving those houses more bandwidth. The testing methodology is a hardware device in the home, that does random things. They presumably download a repetitive file set which the RSP can easily figure out by becoming a test house. The RSP then detects those other houses using the files and gives them priority.
I saw something like this with Skymush satellite. They had a bandwidth test that would always give you 5mbs, but when you tried to download anything real, it was congested after 3pm. I mentioned my suspicions on Whirpool to a Skymush rep, and he was most indignant. But it was plain as day.