Every caller says everything is down. One of them is right.
A single user who cannot reach one application and a site whose line is actually dead describe the outage in the same four words. The difference decides whether this waits for the morning queue or wakes an engineer, and it takes two questions to find out. A voicemail asks neither of them.
Hear it before you read another word. Dial (314) 784-8835. It is an A.I. voice, it answers for several kinds of business, and it does not assume yours — say “I run an IT company” and it will take the call the way your front desk would, including the one at two in the morning where nobody can log in. Or tell it what a typical call sounds like for you and let it answer that. The demo has no calendar or text line behind it, so it takes the job and hands it on; yours books and confirms.
The call you are not there for.
It is 02:40 and the call is from a warehouse supervisor at a client site. Nobody can log in. It could be one switch, one expired certificate, one user who changed a password, or the thing you least want it to be. What the engineer needs before they open a laptop is how many people, at which site, since when, and what changed — because that is the difference between a remote fix and a drive.
Three calls an IT company gets, and what happens to them.
What it gets out of the caller.
- 01
How many users and which site, before anything else. One person and forty people are different incidents with the same opening sentence
- 02
Whether the internet works from a phone off the wifi — one question that separates an ISP or circuit fault from something on your side
- 03
Whether it is one application or everything, which is usually the difference between a SaaS status page and a call-out
- 04
What changed in the last day: an update, a password, a new device, a cancelled card on a renewal. Most out-of-hours calls have one of these behind them
- 05
On anything that looks like a compromise: the indicators in the caller's own words, how many machines, and whether it is still spreading — captured and put at the front of the summary, not buried in it
Why a general answering service loses this call.
- 01
It writes down “everything is down”.
That sentence is what a single failed login and a dead circuit both sound like from the other end of a phone. An engineer reading it at three in the morning learns nothing and has to ring the caller back to start, which is the call they were paying not to make. Two questions asked while the caller is still on the line are worth more than any message written after they hang up.
- 02
It treats a ransom note as a ticket.
A service that grades urgency by how upset the caller sounds will rank a calm description of renamed files below an angry one about email. The indicators that matter are specific and a caller will volunteer them if asked, and the only useful thing a phone can do with them is get them in front of a human immediately. What happens next is your incident plan, and no answering service should be improvising it.
- 03
It wakes you for a printer.
The cost of an out-of-hours service is not the subscription, it is the engineer who answered three calls that could all have waited. Severity is not a tone of voice; it is whether anyone is blocked and how many of them there are, and it is answerable in one question.
- 04
It does not know which client the caller belongs to.
An end user at a managed site does not think of themselves as your client's staff, they think the number they were given is the IT department. If the site is not established on the call, the response clock cannot be attached to the right agreement and the work cannot be billed to it either.
What it costs.
There is no published price yet. The first five businesses run it free for thirty days, with no card; before day thirty we tell you the monthly price, and you decide whether to keep it.
Questions IT companies ask.
Can it tell a real outage from one user with a problem?
It asks the two questions that separate them — how many people and whether the internet works from a phone off the wifi — and puts both answers in the summary. It is not deciding your severity levels for you; it is collecting the facts your severity levels are defined against, while the caller is still there to answer.
What happens on a call that looks like ransomware?
It captures the indicators in the caller's words, how many machines are affected and whether it is spreading, flags that at the top of the summary, and transfers to your on-call number straight away rather than logging it. It does not give remediation instructions. Your incident response plan is yours and a phone line is the wrong place to improvise one.
Does it start our SLA clock?
It timestamps the call and captures the scope the clock is measured against, which is the part that usually goes missing when a call lands in voicemail at 02:00 and is read at 08:30. It does not write into your PSA. Where the summary goes is a setup decision — email, a webhook, or your CRM.
Will it wake our on-call engineer for everything?
No, and not waking them is most of the value. A single user with a work-around is logged for the morning. A blocked site, or anything that looks like a compromise, is transferred. The line between those is a rule you set, not a judgement it makes freshly each time.
Can it handle calls from our clients' staff, not just our clients?
That is most of the volume. It establishes which site the caller is at before anything else, because an end user does not announce whose agreement they fall under — they assume the number they were given is the IT department, and from their side that is exactly what it is.
Find out what your phone is costing you.
The audit is free, it runs on your own numbers, and it takes one short call.