CASE 000Customer zero
We didn’t build this to sell it. We built it because we needed it.
Trickle Up is a United Kingdom-based digital marketing agency running five delivery pods and more than sixty people. ClientFlow OS is the system we run on every day — and the reason every feature in it exists.
Case file
ClientFlow OS · 000
- Client
- Trickle Up
- Sector
- Digital marketing agency
- Based
- United Kingdom
- Team
- 5 delivery pods · 60+ people
- Running
- 25 modules, one workspace
- Status
- In daily use
Chapter 01 · The problem
Nothing was broken. That is what made it dangerous.
Running five pods across sixty people meant the information about any given client lived in five places at once. Chat in one tool. Tasks in another. Client details in a spreadsheet somebody owned. Reports written by hand at the end of the month.
Work still shipped. But by the time a client was unhappy enough to say so, the conversation was already about leaving.
“How is this client actually doing?”
Took asking
The account manager
The pod lead
Whoever sent the last invoice
Chapter 02 · What we changed
We rebuilt the whole delivery cycle in one place.
01
Request
A client submits the intake form.
02
Brief
AI turns it into a structured project brief.
03
Pod
The routing engine picks the pod with real capacity.
04
Plan
Tasks are generated in delivery order and assigned.
05
Watch
Health, growth, ad spend and burnout are read continuously.
The last step is the part we actually missed before — an intelligence layer watching for churn risk, growth opportunities, ad spend problems and team burnout, so the agency finds out before the client does.
Chapter 03 · The routing engine
A new client used to land on whoever answered first.
Now it is a scoring decision, made in the second the brief is approved. Room left, delivery rate, deadline pressure, team capacity and whether the pod has worked in that industry before — weighted, out of a hundred, every pod every time.
New client
Harbour Dental · SEO + Paid Ads
Deadline in 6 weeks · dental, seen before
Pod Crimson
Below the 40-point floor — not offered the work
38/100
Pod Cobalt
92/100
AssignedPod Dusk
65/100
Pod Ember
52/100
Pod Ivory
Below the 40-point floor — not offered the work
39/100
- Room left /30
- Delivery rate /25
- Deadline pressure /25
- Team capacity /10
- Industry match /10
The five factors and their weights are the ones the routing engine actually uses. The scores above illustrate one decision — a pod that has room, ships on time and has done dental before takes the work; a pod under forty is not offered it at all.
Chapter 04 · A week on it
What running on it actually looks like.
- Every morningClient health re-scoredA client going quiet shows as risk, not a resignation email.
- On approvalNew client routed to a podScored out of 100 across every pod, in the second the brief is approved.
- All dayFive tasks per person, maxThe daily limit stops one person quietly absorbing everything.
- After a callTwo-hour PM review windowAI meeting notes wait for a project manager before a client sees them.
- Friday, 09:00Weekly reports draftedWritten from what actually happened, reviewed before sending.
Chapter 05 · Before and now
The same agency, eighteen months apart.
Showing how we run now
- ToolsOne workspace the whole agency works in
- StatusA live board showing who is on what, right now
- ReportsAI-drafted reports every Friday, reviewed before sending
- InvoicesOverdue invoices chased automatically, in the right tone
- ClientsClient health scored continuously, so risk surfaces early
CASE 001Unassigned
The next case file is yours.
Fourteen days free, no card. If it works for sixty people and five pods, it will work for yours.