I recall recently on one of my flights to a recent event looking at a packet of smoked almonds handed to me by the cabin crew which contained the warning “May contain nuts.” I couldn’t help but chuckle to myself, as I bloody well hoped it did indeed contain nuts!
Unfortunately, the world is increasingly oriented towards the lowest common denominator. The apocryphal case of a woman successfully suing McDonalds for her coffee being too hot (which has been somewhat warped by the internet misinformation machine – it’s actually a fascinating case and if you read the full details, you’ll see why) is just another example of warning labels seemingly gone mad because of the left hand side of a common sense bell curve.
At the moment, we default anyone with channels with us to unlimited burst. There’s a reason behind this. Imagine whoever is responsible in your business for interconnect planning is off on holiday, or off sick. Your sales team then brings on a new customer, who underestimated their peak demand… it’s not your end users’ fault a channel limit was hit – why should they receive number unavailable, or busy, or some other error message when trying to place an outbound call? Why should they miss an incoming call? Why should we both miss out on charging for a minute?
Unfortunately, our good nature has been taken advantage of. While we have many excellent customers that manage their capacity professionally and co-operatively, there have been a few outright mickey takes. We still maintain our underlying approach is the correct one; that we connect the odd call when a channel limit would otherwise be hit. Only very rarely have we turned bursting off, and it’s been in response to egregious behaviour, which was not remedied after asking nicely.
But, like packets of almonds, or cups of coffee, we are going to have to change things so they work the way they should across the entire bell curve of capacity management common sense.
In the coming weeks, our approach will change with the deployment of new logic governing channel usage. We will allow some bursting of your allocated channels, and then you will hit the buffers (save for 999 calls). This balances our intent to not disadvantage you, or your customers, for the odd seasonal blip, while removing the issues we have had with the veritable jokers. For the avoidance of doubt, this is not buy some, get one free – this is breathing room for you and your customers while you upgrade your capacity to a better baseline. Think of it like a temporary channel overdraft – it is not intended to be lived in 24/7/365, and we expect all customers to be right-sized at all times. We just want to give a margin for error for the reasons above.
If it appears you’ll be affected by this, your account manager will reach out to you in advance and increase your channels as needed. Feel free to get ahead of the action in advance though, if you know you’re in the red!
Oh, and channel limits are not a means to game our calls-per-second restrictions. We are now setting those independently of the channels allocated to an account, on an as justified basis. All accounts will have a default of 30 call attempts per 10 seconds. No, this is not the same as 3 calls per second – the reason being that our, and our customers’, metering is never going to be perfectly in sync, and we do not want hundreds of pointless tickets complaining about why a limit was hit on this nanosecond versus that nanosecond.
If more are needed, then that’s a discussion with our team – but if your traffic looks like robodialling scrotitude, then you’re going to have a tough time getting it increased.
And one more thing… We have blogged extensively about our nuisance call filtering, so I won’t repeat that here, but the CPS limit on an account is going to become more dynamic. The worse a traffic stream is, the more the whole of it will be throttled.
CPS appears to be the main attribute that scrotes want from a carrier to be able to engage in scrotitude. If you want to avoid trouble with your blue-chip enterprise conversational traffic, make sure you don’t let a bad actor onto your network in the first place, otherwise you’ll find yourself route advancing and it costing you money as we bring the shutters down.
If you don’t hear from us again on this, you’re not likely affected, otherwise we’ll be in touch.
