Deckard
Contact
Contact

Numbers everyone agrees on

Dashboards and BI

Every number in the Monday meeting gets one definition, a named owner and a dashboard on live data. Margin by customer and by product shows where the next sale is worth the most.

Start a conversationSee it step by step
Days to resolveIn range

3.6

Business days from a ticket opening to its resolution. Mean of 400 tickets, last 13 weeks.

Owner

Support lead

Last refresh

09:00

Target range

3 to 4 days

Days to resolve: three answers become one

TodayThree spreadsheets

Each team brings its own formula to the meeting

01Define
Support’s clockWeekends out
Ops’ clockHolidays out
Finance’s clockStarts at opening
02Build
Tests on the data6 passing
TracedTicket to dashboard
Weighted byTicket count
03Publish
Click-throughTo the ticket
AlertIf it leaves range
EmailedTo the leads, weekly
OutAgreed and ownedSigned off
Versions in use1, down from 3
Weekly deckBuilt at 07:00
DisputesSettled at the ticket

Behind each number

Built in your central database and shown in the reporting tool your team already uses.

  1. 01Metric definitions written once and reused everywhere
  2. 02Refresh as often as the work needs
  3. 03Executive views for the weekly review
  4. 04Click from any number to the records behind it
  5. 05Alerts when a number leaves its target range
  6. 06Scheduled reports by email
Yours at the end

support-metrics/

metrics.ymlThe definition, owner, clock and tests

models/Tickets and teams, joined from 2 systems

dashboards/4 views, kept as code

warehouse/Access rights and refresh times

See it work

One number, down to the ticket

How it goes live

  1. 01 · Optional

    The disputed numbers

    Every number that gets argued over, and who owns it

  2. 02 · Optional

    One number, live

    One disputed number, defined and running on live data

  3. 03 · Dashboards and BI

    Shared definitions

    Every definition tested, and the executive views built on it

  4. 04 · Ongoing

    Owner sign-off

    Definitions change only through review by their owner

Related services

Exported, watched and set as a baseline

Sent into the month-end finance pack, checked on the way.

Data engineering and pipelines

An agent pulls its tickets when it leaves its range.

Agentic AI and workflow automation

Support roadmap items are scored against this baseline.

AI strategy and assessment

All solutions

By industry

Power, plant output and margin, each defined once

Stranded capacityAs of 09:00
No cooling to draw it
Contracted in halls C and D150 kW
No matching U free
Spread across 38 cabinets110 kW
B feed unavailable, hall D
Until the second feed is energized80 kW
Reserved but not energized
Counted separately from stranded88 kW

340 kW stranded across 8 halls and 1,184 cabinets.

Data centers and colocation

OEE, 71%By line and shift
Availability
Run time over planned time83%
Performance
Actual rate over the ideal rate91%
Quality
Good parts over all parts made94%
Material starvation, 2nd shift
Three occurrences this week2.2 h

Planned downtime is excluded and stated separately.

Manufacturing and industrial

Margin after feesThis week
Pull orders and fees, 4 channels
Each fee read per order, as charged
Net returns against each sale
Shipping and goods cost joined by SKU
3One marketplace payout pending
Its fees held at the rate on the order
4Publish margin by channel
38.2% this week, down to each order

Revenue less cost of goods, channel fees, shipping and returns, over revenue.

Retail and e-commerce

Which number does nobody agree on?

You get a list of disputed numbers and a plan to settle each.

Start a conversation