IT providers fail slowly: response times stretch, proactive work stops, invoices grow teeth. One sign is a bad patch; three or more is a pattern, and patterns don't self-correct while the direct debit keeps clearing. The switch you're avoiding is mostly done by the incoming provider anyway.
IT providers rarely fail loudly. They fail slowly: response times stretch, the proactive work quietly stops, and the service you signed up for erodes into something you tolerate. Because it happens gradually, most businesses only see it clearly in hindsight, usually after they have switched.
Here are the signs we hear most often from businesses that come to us, and the standard you are entitled to expect instead.
1. Response times have quietly doubled
The SLA says four hours; tickets used to be answered in twenty minutes; now they sit until tomorrow. Nobody announced a change.
Pull your last twenty tickets and look at the actual first-response times, not the SLA. The trend line tells you everything the account manager won't.
Good looks like: a published average, not just an SLA ceiling. Ours is 3.3 minutes, and we have never missed an SLA commitment.
2. Every issue starts from zero
You explain your setup again, to a different person, every time you call. That is a documentation failure and a staffing model choice, and you are paying for it in repeated explanations. Good looks like: engineers who know your systems, and documentation good enough that the ones who don’t can find out in minutes.
3. The only proactive contact is the renewal notice
No quarterly reviews, no roadmap, no “you should know about this deadline.” Then the renewal arrives with a price increase.
If your provider hasn't mentioned the PSTN switch-off or Windows 10 end-of-life unprompted, they aren't watching your environment. They're billing it.
Good looks like: scheduled reviews and a provider who raises deadlines before you read about them elsewhere.
4. Recurring problems never get root-caused
The same printer, the same sync issue, the same server hiccup, fixed the same way every few weeks. Break-fix on repeat is cheaper for them than engineering the problem away, as long as you keep tolerating it. Good looks like: recurring issues treated as engineering failures, investigated once, and killed.
5. Invoices keep surprising you
“Out of scope” appears on invoices for things you assumed were covered. Call-outs are chargeable. A routine change becomes a mini project. Whatever the contract technically says, a pricing model that generates surprise invoices is a pricing model built on them. Good looks like: a fixed per-user monthly price, no ticket fees, and project work quoted upfront.
6. Your team has stopped reporting things
Staff live with broken things because raising a ticket feels pointless. This is the quiet killer: the ticket count looks low, the provider looks fine on paper, and your team absorbs the dysfunction as unpaid IT support.
If people are saying "don't bother, it takes forever," the relationship has already failed.
Good looks like: support your team actually uses because it is faster than working around the problem.
7. Security has stagnated
MFA half-deployed years ago and never finished. No MDM on the phones. Backups nobody has test-restored. Cyber Essentials mentioned once and dropped. The threat picture moves every year; a security posture that hasn’t changed since onboarding is not keeping up. Good looks like: a layered, current security stack and someone accountable for keeping it current.
8. Leaving feels deliberately hard
Three-year terms, 90-day notice windows, admin passwords they are cagey about, documentation that mysteriously belongs to them. A provider confident in their service does not need contractual hostages. Good looks like: rolling contracts with a short exit clause. Ours is 60 days, in every agreement, because retention should be earned monthly.
If you counted more than two
The fear that keeps businesses in bad contracts is the switch itself: the imagined chaos of moving. In practice, the incoming provider does the work:
- We audit your environment before the switch date
- We deal with the outgoing provider directly, including the awkward conversations
- Migration runs out of hours, and we own the documentation from day one
Every client who has joined WikiTech has stayed, and with a 60-day exit clause, that is a choice they re-make every month.