Design partners
No logo wall. Not yet.
We are in the design-partner phase. Below is the exact structure each case study will take, populated with illustrative content so you can see what we intend to publish and hold us to it. Every number marked to be measured will be a real measurement or it will not appear.
The two studies on this page are illustrative. The environments and before-state figures are composites from design-partner discovery calls. No customer is named, no logo is shown, and no after-state result is claimed. In this category one discovered exaggeration ends the deal, so we would rather publish an empty column.
MSP · 41 clients
A regional MSP with eleven helpdesk instances it does not own
Environment
- Fleet
- 6,400 endpoints across 41 client tenants
- Helpdesk
- ConnectWise, plus four client-owned Zendesk instances
- RMM
- NinjaOne
- Worst recurring ticket
- Printer and mapped-drive faults, ~31% of Tier 1
Before
- Endpoint tickets / month
- 3,900
- Median time to first touch
- 46 min
- Median time to resolution
- 3h 20m
- Tier 1 techs
- 9
What was turned on, and in what order
- 01Weeks 1–4 — Hooded on two client tenants. No execution. Diagnosis quality reviewed weekly against what the tech actually did.
- 02Weeks 5–8 — Approve-each on Tier 0 for those two tenants. Print spooler and SMB credential actions only.
- 03Weeks 9–12 — Unattended Tier 0 during business hours for the two tenants. Approve-each extended to the rest.
After
- Endpoint tickets / month
- 3,900 — unchanged, this is not a ticket-reduction product
- Resolved without a human
- to be measured
- Median time to first touch
- to be measured
- Bate rate
- to be measured
What they were nervous about
“The thing I was worried about was a tech losing the ability to explain to a client what happened on their machine. I wanted the audit page before I wanted the automation.”
Head of service delivery — design partner, name withheld until go-live
In-house IT · single tenant
A mid-market manufacturer with 1,800 laptops and three technicians
Environment
- Fleet
- 1,800 endpoints, 74% Windows, 22% macOS, 4% Linux
- Helpdesk
- Jira Service Management
- RMM
- Intune
- Worst recurring ticket
- VPN adapter faults after monthly patch Tuesday
Before
- Endpoint tickets / month
- 780
- Post-patch spike
- ~140 tickets in 48 hours
- Median time to resolution
- 5h 10m during a spike
- Tier 1 techs
- 3
What was turned on, and in what order
- 01Weeks 1–4 — Hooded across the whole fleet. Correlation match rate measured and tuned from 61% to 94% by adding the asset-tag field.
- 02Weeks 5–8 — Approve-each, everything. Bulk approval used for the post-patch VPN spike.
- 03Weeks 9–12 — Unattended Tier 0 and Tier 1 outside the executive device group, business hours only.
After
- Endpoint tickets / month
- to be measured
- Post-patch spike handled unattended
- to be measured
- Median time to resolution
- to be measured
- Bate rate
- to be measured
What they were nervous about
“My question was what happens at 2am when it gets it wrong and there is nobody watching. The answer was that it rolls back and puts it in the queue, and I could see that happen in a sandbox before I signed anything.”
IT manager — design partner, name withheld until go-live
Ten fleets. One quarter. No cost during Hooded.
You get the diagnoses and the audit record from week one. We get to find out where our correlation is wrong on a real fleet. If it does not work you have lost a month of Hooded telemetry and nothing else.