Backups that have been tested, and a recovery plan written down first.
Evolve sets up and manages backup for servers, workstations, Microsoft 365 and the cloud services your business runs on — then tests that the data actually comes back, and documents what recovery involves before you need it.
Based in Shelton, with onsite support throughout Connecticut and remote support wherever your team works.
ILLUSTRATIVE
File server — last full backup
02:14 VERIFIED
Workstations protected
24 / 24
Microsoft 365 mailboxes
31 DAILY
Offsite copy
03:40 COMPLETE
Last test restore
14 DAYS AGO
Illustration of the information a managed backup produces. Not a live client feed, and not a promise of specific times or coverage.
★★★★★
500+ Google Reviews
Serving Connecticut Since 2017
Onsite statewide, remote anywhere your team works
A backup is a copy of data. Recovery is getting your business running again.
Many businesses that call us after something goes wrong already had backups. What they did not have was an answer to the next four questions: how long the restore takes, what order things come back in, who does it, and whether anyone has ever tried.
Those questions are cheap to answer in advance and expensive to answer during an outage. So we treat backup as two pieces of work: configuring the copies, and writing down the recovery — what gets restored first, how long each step realistically takes, and what the business does in the meantime.
- What would stop you working within an hour of being gone?
- How much re-entered work is acceptable — a day, an hour, none?
- Who is allowed to authorise a restore over live data?
- When did someone last verify a restore actually worked?
The fourth one is the one that usually goes quiet.
Four places your business data actually lives.
Gaps usually appear where responsibility is assumed rather than assigned — the laptop nobody enrolled, the mailbox everyone thinks Microsoft keeps forever, the cloud application whose data nobody has ever exported.
Servers and line-of-business systems
Whole-machine image backups so a dead server can be brought back rather than rebuilt from an install disc and memory. Databases and the applications your business runs on, backed up in a state they will actually restore from.
Image-levelDatabasesBare-metal restore
Workstations and laptops
The machines where work-in-progress lives before anyone saves it to the right folder. Coverage checked by device, including the laptops that only appear on the network every other week.
Coverage by deviceRemote laptops
Mail, OneDrive, SharePoint, Teams
Microsoft keeps the service running; retaining your data past its own limited windows is your responsibility. A separate backup means a mailbox deleted four months ago is still recoverable. See Microsoft 365 & Email.
Exchange OnlineOneDriveSharePoint
A copy that is not in your building
A local copy for speed, an offsite copy for the fire, flood or theft that takes the local one with it — and offsite copies kept where an attacker who reaches your network cannot simply delete them.
Local + offsiteSeparated credentials
Four things that go wrong, and what getting back actually involves.
These are in rough order of how bad they are to live through. Each one is survivable with preparation and painful without it — and the preparation is different for each.
Ransomware across the network
The worst realistic case for most small businesses.
Isolate first so it stops spreading, work out what was reached and when, then rebuild from a copy known to predate the encryption. Servers and the systems the business cannot work without come back first; the rest follows in a planned order. Depending on the scope, this can become days of work rather than hours, which is why the offsite copy and restore order matter
Backups an attacker with network access cannot delete, enough retention to reach back past the infection, and a restore order agreed in advance. Prevention belongs next to this — see Cybersecurity.
A server fails outright
Hardware, power event, or a drive array that gives up.
Bring the machine back from an image onto replacement hardware, or stand it up temporarily elsewhere so people can work while the permanent fix is sorted. What decides the timeline is whether the backup is a whole-machine image or a folder of files, and whether replacement hardware is available.
Image-level backups rather than file copies, a known-good restore target, and documentation of how the server was configured. Hardware lifecycle planning keeps this in the calendar instead of the emergency column.
A mailbox or account is gone
Someone left, the licence was removed, and the mailbox went with it.
Restore the mailbox, its OneDrive contents, or the specific messages needed, either to the original account or to a shared mailbox someone still working can reach. Often it is one thread from two years ago that a client or an attorney is asking about.
A Microsoft 365 backup with retention past Microsoft’s own deletion windows, and an offboarding process that archives before it removes.
A file was overwritten in March
The most common call, and the one retention decides.
Find the version from before the change and put it back without disturbing everything edited since. Usually quick — provided the backup goes back far enough and keeps more than the most recent copy.
Retention chosen deliberately rather than left at a default, and version history that survives a file being overwritten rather than deleted.
We do not quote guaranteed recovery times, because they depend on how much data there is, what failed, and what hardware is available on the day. What we will do is walk through each of these with your environment in front of us and give you a realistic estimate in writing.
A backup nobody has restored from is a hypothesis.
Backup software reports success for plenty of jobs that will not restore — an excluded folder, a database backed up while it was open, a job that has been quietly failing on one machine since a February update. The only way to know is to bring data back.
Jobs checked, not assumed
Failures and skipped machines reviewed rather than left in an inbox rule. A job that stopped running gets noticed the same week, not the same quarter.
Test restores
Real data brought back to a safe location and opened, so we know the file is intact and not just present. Recorded with the date, so “when was this last tested” has an answer.
Coverage against reality
New server, new laptop, a folder moved to SharePoint, a new cloud application. Backup scope drifts behind the business unless somebody compares the two on purpose.
The recovery plan
What is protected, where the copies are, what comes back in what order, who to call, and a realistic estimate per step. Yours to keep, and useful to whoever is holding it at 6am.
- Every system and dataset in scope, and anything deliberately excluded
- Where each copy lives and how far back it goes
- Restore order, with a realistic estimate for each step
- The date and result of the last verified restore
If you already have backups running and simply want to know whether they would work, that is a defined piece of work and a reasonable place to start.
How far back you can go is a decision, not a default.
How much work you would redo
If backups run overnight, a failure at 4pm costs a day of work. If they run hourly, it costs an hour. The right frequency comes from what your team produces during the day, not from what the software defaults to.
How far back you can reach
Problems found the same day need days of history. A mailbox someone needs from last spring, a file overwritten in March, or ransomware that sat quietly for weeks needs months. Some businesses also have contractual or regulatory reasons to keep records for a defined period — those belong in the decision.
We will tell you what each option costs before you choose, and we will not quietly set the longest retention available and bill for it. If your obligations require a specific period, bring that to the conversation — we implement and document what you and your advisors decide, and we do not determine your compliance requirements for you.
Setting up backup is a project. Knowing it still works is not.
We are happy to design and install backup as a one-off job and hand it over. The honest caveat is that backup is the service most likely to fail silently: nothing looks wrong until the day you need it. The difference between these two columns is whether anyone is checking in the meantime.
Design and installation
Included
Included
Failed jobs
Email you whoever set up the alerts
Reviewed by us and chased
Test restores
One at handover
Performed periodically and dated
New devices and systems
Added when you ask
Picked up as part of coverage checks
Recovery plan
Written at handover
Kept current as the business changes
When something fails
Call us and we help, billed hourly
We are already familiar with the environment
Managed backup is available as an add-on across all of our Managed IT plans, so you can add ongoing backup monitoring and oversight at the level of support that fits your business. Foundation keeps the essentials managed while hands-on support is billed as used; Complete adds the support desk and broader security; Elite includes labor and onsite response. Managed backup can also be purchased on its own if the rest of your IT is already covered.
If you are not sure whether your current backups would restore, that is the thing worth finding out — before anything is bought or changed.
Recovery is the last line of three.
Ransomware is one reason restores get tested. Prevention and recovery planning belong together.
Mailboxes, OneDrive and SharePoint need their own backup and their own retention decision.
Who checks the failed job, runs the test restore, and keeps the recovery plan current.
Questions we get before the first call.
Anything not covered here, ask us on the call. There is no wrong question.
Backup for servers, workstations, Microsoft 365 and cloud data, with local and offsite copies; retention set deliberately; restore testing; and a written recovery plan covering what comes back in what order and roughly how long each step takes. We do it as a defined project or as an ongoing managed service.
Microsoft keeps the service available and holds deleted items for a limited period, which is not the same as a backup you control. Once those windows pass, a deleted mailbox or overwritten file is gone. A separate Microsoft 365 backup with your own retention is what covers the mailbox somebody needs from last spring. Anything stored only on a local machine and never synced to OneDrive, SharePoint, or another protected location may still be outside the backup scope.
It depends on what failed, how much data is involved, and what hardware is available — which is why we will not put a guaranteed number on a web page. What we will do is work it out for your environment, write a realistic estimate per step into the recovery plan, and tell you where spending more would shorten it.
Yes, and it is a good place to start. We review what is in scope, what is silently excluded, how far back it goes, and whether a restore actually produces usable data. You get the findings in writing whether or not you do anything further with us.
They do not prevent it; they determine how bad it gets. What matters is that copies exist an attacker with network access cannot delete, that retention reaches back past when the infection started, and that someone has tested a restore. Prevention is the security side of the work — see Cybersecurity.
Frequency follows how much re-entered work you can tolerate; retention follows how late you tend to discover problems, plus any contractual or regulatory period you have to meet. We will lay out the options with the cost of each rather than picking the longest one and billing for it.
Both, and most businesses need both. On-premise servers get image-level backups so the whole machine can be brought back; workstations, Microsoft 365 and cloud data are covered separately. A copy stays local for speed and a copy goes offsite for the events that take the building with them.
Call (203) 433-2042 rather than filling in a form. Tell us what happened and what has been tried, and stop writing to the affected drive or system if you can. Recovery work is normal for us and you do not need to be an existing client — we will give you a realistic answer about what is possible when we know what we are dealing with.
Tell us what is going on. We will tell you what it takes to fix it.
Have a problem, project, or upgrade in mind? Send us the basics and we’ll figure out the right next step. For something urgent, call us directly.
Need help right away?
If something is affecting your business now, call us directly.
Call (203) 433-2042Larger project or Managed IT?
Talk through the scope, goals, and the right next step directly with Evolve’s owner.
Schedule a 20-Minute IT Call356 Howe Ave, Shelton, CT • (203) 433-2042 • Mon–Fri 10–7, Sat 10–4