By Simon Woodhead
There’s a story about our termination which we’ve been pretty poor at telling! I’d like to try and address that.
The technical story is pretty evident I hope. Our own high performance modern network gives diversity and low latency access from 5 proprietary global Availability Zones, any one of which is specified to carry our entire billion minutes of traffic and much more. Our real-time call routing not only applies 200 plus fraud controls, but also uniquely routes every call according to telemetry specific to the account. Together they give us control and ownership – if we ever use the word ‘our partner’ to defer blame for some outage or other incident, please shoot me – the buck stops with us and if there’s an issue, it is ours.
But while critical, the network is only the glue in the middle. A spam caller over rural developing world DSL at one end into a SIM bank operated by a spiv somewhere else is never going to be a high quality call, however good the network. That’s why you’ll find neither traversing our network.
Firstly, reputation matters and we simply will not tolerate scam or spam traffic. It isn’t as simple as saying ‘no dialler’ because we carry bank traffic which looks remarkably like that, but we aim for high repute, high quality customers wherever we can and our KYC looks at the intent rather than just the metrics. We’re handing this traffic off to downstream networks and want to be the source of their best traffic, not their worst. In the USA, with our digital certificate being attached to every outgoing call, I’d suggest this is going to become more and more important – not least because that certificate is carried to the final destination, not merely to the next hop. Whilst at the moment the attestation level the certificate is used to sign is used for filtering, I can’t imagine it’ll be long before operators catch on to the fact that a given signer has a reputation, good or bad, and filters traffic accordingly. We want our customers to be unaffected and therefore aim to only carry clean traffic.
But what do we do with that traffic? For the UK, that is easy. As, by our reckoning, the third largest wholesale operator in the UK, an increasing amount of it completes on-net, presuming it is between one of the dozens of operators for whom we host number ranges and the hundreds of operators who have us in route. Beyond that we have direct interconnects with anyone who is anyone, so the vast majority of traffic is a single-hop from its destination. We even expose that to you in our rate sheets by differentiating between Simwood, Simwood-Peer and Transit; so you know before you route a call how we’re going to complete it. This works similarly for the US, although we have to be honest that it is a much bigger market and we’re a much smaller part of it.
Outside the markets where we have network footprint (UK and US) is where I think we have opportunity to improve. Don’t get me wrong: UK operators generally provide good termination by international standards and we rank very highly amongst them, but neither are perfect. The problem is this is a $30bn industry that is increasingly populated by outfits treating voice calls as a commodity. They measure quality in averages and don’t care about the experience of individual calls. It is a bit of a cesspit of “I do VoIP. You buy from me, I buy from you; how much you buy from me?” type relationships. We’ve never gone in for these, favouring relationships with reputable larger entities even if they’re buy-side only (although increasingly we find the sell-side follows). The problem is though the more people regard voice as a commodity, the greater the temptation to save a few points by introducing these lower quality routes somewhere in the supply chain. Price goes down while there’s simultaneously more mouths to feed so by the laws of physics alone, the quality cannot remain constant. It doesn’t end anywhere good.
Instead, we’re aiming for the best quality rate deck on the planet. That means what we offer in the UK – i.e. a single hop to destination – extended, which is a tall order admittedly. That leads to secure, high quality calls supporting wideband codecs end-to-end, with CLI guaranteed everywhere. It also leads to fewer support tickets, for our customers and us – burning hundreds of Pounds in cost, to investigate a few hundredths of a penny of revenue. This is what I call ‘zero defect routing’ and it makes so much sense!
To be clear we’re not there yet and realistically it is a journey not a destination but it is one we’re well on the way to. In the last month we’ve made substantial changes to our routing, and have many more in progress. We’ve strengthened Europe and Latin America, with Africa and Australasia improving too. And have a guess what the result is? Well it turns out that having fewer mouths to feed in the chain leads to significant cost reductions. Furthermore, avoiding those mouths trying to nickel and dime each other improves the quality which leads to less support; dramatically less as it happens. All this feeds through to you not only in terms of improved service but also improved price. Hopefully, we can create a self-perpetuating cycle here.
If you’re not already routing your international traffic to Simwood, maybe now is the time? Please speak to us!
