Telephony: the number is the asset, not the system

The phone system is the easiest equipment to sell in this family, and it is also the one whose purchase is least often justified by what appears in the quotation.

The reason is simple: your customers do not know your equipment. They know a number, often for ten years, often written on a vehicle, on a sign, in a directory, and in the phones of people you have never met.

That number is a company asset on the same footing as a domain name, and it is protected with far less care. The system, by contrast, is a consumable: it gets replaced every eight years and nobody outside notices.

This article starts from there. It covers what happens to the number, what a phone becomes when it runs over the network, what the old analogue handset did during a power cut that the new one does not, and the share of your first contacts that no longer arrives over a line at all.

A conversation needs very little throughput — far less than an ordinary web page — and yet it breaks where everything else gets through. The reason is that voice is the only thing on your network that is not allowed to wait.

A file arriving a tenth of a second late still arrives. A word arriving a tenth of a second late arrives after the next one, or does not arrive at all. Regularity is what counts, not quantity.

The three causes we find, in order: the upstream path saturated when somebody sends a large file, equipment that does not distinguish voice from the rest, and a wireless link on the path — wireless being excellent for convenience and mediocre for regularity, as the family article says.

The first two are corrected by configuration rather than by purchase: give voice priority on the equipment going out, and reserve a share of the upstream path. That is half a day’s work and it settles most cases.

The third is corrected by cabling the fixed handsets, which brings back the previous section. A phone sitting on a desk that never moves has no reason to be wireless, and it is precisely the one whose conversations matter most.

A modern phone system offers dozens of functions. Four genuinely serve a twenty-person business, and the rest weigh down the configuration that the switch page’s section 8 asks to fit on a page.

The four: distributing calls to several handsets, transferring a call without losing it, forwarding by time of day as the previous section describes, and keeping a log of incoming calls with their numbers.

The fourth is the most underrated and the cheapest: knowing which numbers called, at what hour, and which got no answer. It is the only function on this page that produces revenue, because it makes calling back possible.

What is almost never justified in a small organisation: queues with music, agent statistics, systematic recording, and multi-level menus. Each of those is excellent at a certain size and costs a customer at yours.

The rule we apply is the same as for network equipment: if the configuration does not fit on a page, it is more complicated than the problem. And on this subject specifically, a complicated configuration is audible to the caller, which is not the case on the pillar’s other pages.

The physical handset is the heaviest line on a telephony quotation, and it is the one where the count is almost always copied from headcount rather than from use.

Three categories of people exist in every business. Those receiving calls all day — reception, sales, after-sales — who need a real handset, with a receiver, a screen, and keys.

Those who call occasionally and receive rarely, for whom software on the computer with a headset does exactly the same job for a fraction of the price. And those who do not use the phone for their work, who need nothing, even if they have a desk.

Honest counting gives, in the businesses we equip, about a third real handsets. The standard quotation proposes one per desk, and the difference is a substantial share of the budget — with no effect at all on what matters, which is whether somebody picks up.

A useful nuance: the physical handset keeps a real advantage in noisy environments and for people who talk while doing something else. That is not a status question, it is a question of free hands and microphone quality.

What we will refuse: migrating telephony onto cabling we have not checked. Voice is the only thing on your network that makes a defect audible to a customer, and mediocre cabling becomes a commercial problem on migration day.

We will refuse to install a voice menu in a business under about twenty people, and to install one without a key leading immediately to a human. It is a function that sells well and costs calls.

We will refuse to enable call recording without the questions of notice, retention and access having been settled in writing. Our default position in a small organisation is not to enable it at all.

What we do: the number verified as being in the company’s name and its portability obtained in writing before any migration; the telephony chain on a power protector; forwarding to a mobile configured and tested rather than promised, including for the midday slot; real handsets counted on use rather than on headcount; and the configuration exported with its dated page.

And what you can do today without us: ring your own number from a mobile, three times — in the morning, around one, and after closing — and listen to what happens each time. It is this page’s only measurement, it takes three minutes, and it decides half of what there would be to do.