Support & SLAs
Something has broken. Here is exactly what happens next.
Support is the part of hosting that only matters when it matters, and by then it is too late to find out how it works. So here it is in advance.
Getting hold of us
Three routes in, all of them answered by engineers
Use whichever suits the urgency. If it is genuinely urgent, use the first one — nothing beats a phone for a P1.
Ring the desk
0118 384 2175Answered by an engineer, on average in about 15 seconds. Best for anything down, anything urgent, and anything easier to explain out loud than in writing.
Email the desk
support@coffeecupsolutions.comRaises a ticket automatically and sends you the reference. Good for detail, screenshots and anything that can reasonably wait for the queue.
Not a customer yet
If you are being quoted, migrating in, or simply want to ask how we would handle something, the enquiry form goes to the same building.
Service desk hours
8am to 6pm, Monday to Friday, excluding public holidays.
Monitoring
24/7 on everything we host, with alerts going to a human rota rather than a mailbox.
Out of hours
Emergency cover for customers whose agreement includes it, with the arrangements set out in that agreement.
The lifecycle
What happens after you tell us
The same sequence whether you rang it in, emailed it, or our monitoring found it before you did.
-
1
It gets logged
Every issue becomes a ticket with a reference, whether you raised it or an alert did. If you rang, we log it for you while we are on the call. Nothing lives only in somebody’s memory of a conversation.
-
2
We agree a priority
Set with you, based on business impact rather than how the ticket was worded. If we have called it wrong, say so and we will change it — that is a thirty-second conversation, not a formal process.
-
3
An engineer picks it up
With the build documentation for your environment in front of them. For anything urgent that starts immediately; for scheduled work it goes into the queue in priority order.
-
4
You hear what is going on
Updates at a frequency that matches the priority, including when the honest update is that we are still working on it. A ticket that has gone quiet is a failure on our side.
-
5
It is closed with an explanation
What happened, what we changed and whether it can happen again. After a P1 we will put that in writing, and if the answer is that we caused it, that is what the write-up will say.
Priorities
P1 to P4, defined by impact
Priority is about what the problem is doing to your business, not about how strongly it is being felt. Getting that right is what stops the genuinely urgent things queueing behind the merely annoying ones.
| Priority | What it means | Typical examples | How we respond |
|---|---|---|---|
| P1 Critical | A hosted service is down, unreachable or unusable, and there is no workaround. Also anything we believe to be a live security incident. | The site returns an error to every visitor. The application will not load. A server is offline. Signs of a compromise. | Everything else stops. Worked continuously by whoever is needed until service is restored, with updates as we go. Ring this one in — do not email it. |
| P2 High | Serious degradation, or a business-critical function broken, while the service as a whole is still up. A workaround may exist but it hurts. | Checkout failing intermittently. Pages timing out under normal load. Backups failing. A certificate about to expire. | Picked up ahead of routine work and investigated the same working day, with a cause and a plan back to you as soon as we have one. |
| P3 Medium | A fault with limited impact, affecting one person or a non-critical function, where the business can carry on in the meantime. | One user cannot connect over SFTP. A staging site is misbehaving. A scheduled job needs adjusting. | Queued and worked in order alongside other scheduled work, with a target we agree with you when it is logged. |
| P4 Low | Requests, questions and planned changes with no current service impact. | Add a DNS record. Provision a staging environment. Advice on a PHP version. A capacity question ahead of a campaign. | Scheduled. Anything carrying risk goes into a maintenance window and is agreed with you before it happens. |
On response times. We have not printed a table of minutes here, because the honest answer is that it depends on the service and on what you have contracted for. Cover hours, response targets and any service credits are set out in your own agreement, and they are the ones that count. The framework they sit within is on our service level agreement page. If you are not sure what yours says, ask us and we will send you a copy rather than making you find it.
Scope
What support covers, and what it doesn't
Out of scope does not mean we will not help. It means the work gets quoted rather than absorbed, and you hear about that before it starts.
Included on a managed plan
- The platform itself — operating system, web server, database, runtime and services
- Configuration, tuning and hardening of anything we built
- Patching, updates and the rollback when an update misbehaves
- TLS certificates, renewals and expiry alerts
- DNS we host, and helping with DNS you host elsewhere
- Backups, retention and restores, including the ones you asked for by mistake
- Performance investigation on our side of the line
- Security incidents affecting services we host
- Capacity changes, resizing and honest advice on both
- Monitoring alerts, and the work that follows them
Quoted separately
- Writing or debugging your application code — we will tell you where the fault is, and fixing it is a project
- Design, content and editorial changes to your website
- Bespoke development, integrations and data migrations beyond onboarding
- Staff IT support — laptops, printers, the office network, Microsoft 365 and the rest
- Systems we neither host nor manage, however happy we are to point you in the right direction
- Formal penetration testing and certification work
The IT support items are not a brush-off — they are the day job of Coffee Cup Solutions, which is the same company. We can simply put you through.
Escalation
Escalation here is short, because the company is. If the engineer on your ticket cannot resolve it, it goes to a senior engineer, and after that to the service desk manager. In practice that is one person turning round and asking another, usually before you have had to ask for it.
You can also escalate on demand. Ring the desk, quote the ticket reference and say you want it escalated. You do not need a reason beyond not being happy with how it is going, and asking will not put you at the back of anything.
Anything that turns out to be our fault gets a written explanation and, where it is warranted, a change to how we do it. Any service credits that apply are the ones in your agreement.
Planned maintenance
Most patching and routine work is done without you noticing, in defined windows outside business hours. Where there is any prospect of an interruption, your named contacts get notice by email — what we are doing, when, how long we expect it to take, and what you would see if it went wrong.
Work with real risk attached, such as a major version upgrade, is agreed with you first and staged where staging is possible. Occasionally a serious vulnerability means patching faster than the usual notice period allows. When that happens we act, then tell you what we did and why we did not wait.
During any window with an impact, we will tell you when it starts and when it is finished, rather than leaving you to test the site yourself and wonder.
Reporting an issue
What helps us fix it faster
None of this is mandatory and we will never send you away to fill in a form. It just shortens the part where we are asking questions instead of working.
- What you saw
- The exact error text if there is any, and the page or screen it appeared on. A screenshot beats a description.
- Where
- The site, hostname, application or server affected. If you have several with us, this saves a round trip.
- When it started
- Roughly when it began and whether it is constant or intermittent. Intermittent faults are a different hunt entirely.
- Who is affected
- Everyone, one office, one user, or just you. It changes the priority and it changes where we look first.
- What changed
- A deployment, a plugin update, a DNS change, a new integration. Even if you are certain it is unrelated, tell us.
- How to reach you
- The best number for the person who can actually see the problem, in case we need to look at it together.
FAQs
Support questions
Already a customer? Just ring.
No portal login, no reference number needed, no explaining who you are to three people first. If you are weighing us up instead, ask how we would have handled your last outage.
Or email hello@coffeecupsolutions.com