Frequently asked
The questions that come up first
Eight answers, in the order they usually get asked. Open any one.
What if it phones a venue at a bad time?
It can’t phone outside 9am to 5pm, and it can’t phone a venue that is closed. There is a hard cap of two calls a day and five a week per venue. If a contact asks not to be called again, that is recorded and honoured. If a venue has a rule of its own — never before 11am, always ask for the duty manager — that gets stored against the site and the agent follows it.
What happens to our data if we leave?
API access is revoked the day you say so. You get a full export of your screens, incidents, contact logs and transcripts. Once you confirm the export is good, everything is deleted within 30 days. No lock-in clause holds your incident history hostage.
What if you’re unavailable?
The automated layer does not depend on anyone being at a desk. Detection, texts, calls and callout flagging all run unattended. Escalation and dispatch are handled by your own team in your own systems, so a person being on holiday never blocks an engineer going out. Support is business hours, Monday to Friday.
Why not just use our CMS’s built-in alerts?
Your CMS already knows a screen is dark. That is not the problem. The problem is that the alert lands in an inbox and nothing happens next. Nodark does the next part: it works out whether the venue is even open, texts the site, phones the contact, talks them through a router power-cycle, and only flags an engineer once the simple fixes are ruled out. If your CMS covers two platforms in one view, applies each venue’s real opening hours, and holds a phone conversation, you do not need this.
Will it text or call our clients?
No. Outreach goes to venue site contacts — the people standing next to the screen. Your client relationships stay yours. Outreach is also enabled client by client, so you choose exactly who is in scope and when.
Does it dispatch engineers on its own?
No. It raises a callout for approval, with the diagnosis attached. A person on your side approves it. Nothing is booked and no cost is incurred before that. In testing, one callout opened at 17:49 and cancelled itself at 17:51 when the screen recovered — nothing was dispatched, nothing was billed.
How fast does it actually detect a screen going down?
Screens are checked on a roughly 15-minute cycle, so detection follows within minutes of a missed check-in — not seconds. It is not real-time, and we will not claim it is. What matters is that a response is already under way before anyone has noticed the screen is dark.
What does it save us?
Around six in ten single-screen outages clear with a router power-cycle. At roughly €85 an engineer callout, every one of those handled over the phone is a visit you did not pay for. One AI call on a 9,000-screen estate ran 21 seconds and avoided an engineer visit outright. The larger saving is the inventory: a screen that comes back the same day is a screen you can still bill.