A coffee conversation in 2021 changed how I think about operations forever.
A coffee conversation in 2021 changed how I think about operations forever.
I was catching up with a partner — one of those informal chats that always turns more honest than any boardroom meeting. We were talking about how work gets shared with their GCC.
Somewhere between the second cup and the biscuits, she said something that stuck with me.
“We just email it across. And then we wait.”
I asked her what “wait” looked like.
She paused. Then started listing things — almost reluctantly, like she'd been carrying them for a while and hadn't quite said them out loud before.
Nobody could see where a task stood without sending another email to ask. Turnaround times were guesswork. Quality issues kept surfacing — same mistakes, different weeks — but because nothing was structured, nobody could trace them back to a knowledge gap. Leadership had zero visibility into utilisation. Future resourcing was being decided in the dark. Client deliverables were slipping. And underneath all of it, quietly — sensitive client data, account references, transaction details — moving back and forth across personal inboxes with almost no governance around any of it.
None of this was negligence.
It was just what everyone had always done.
The real problem wasn't email
It was that nobody had named what email had become.
It had quietly absorbed every function it was never designed to carry. Task management. Status tracking. Quality assurance. Audit trail. Capacity planning. Reporting layer. It was doing all of these things — badly — because no one had ever stopped to ask whether it should be doing them at all.
I spent time after that conversation understanding how widespread this was.
The answer was: everywhere.
73% of firms say email is a direct cause of SLA breaches. Not a contributing factor. The cause.
I wasn't surprised.
So I looked at what the market could offer.
ServiceNow. Appian. Pega. Power Automate. We evaluated them seriously. These are mature, well-certified platforms. For most organisations, one of them will be the right answer — the ROI case is strong, and they deploy in weeks, not months.
But every conversation about the specific shape of what was actually needed ended the same way.
Work allocation to a GCC. Live resourcing visibility. Real-time knowledge documentation. SME chat support. Management reporting. Client billing back. A clean architectural path for AI when the time came.
None of it fit neatly. We would have spent more time bending the tool than running the operation.
So we built it from the ground up
With complete clarity on every problem it had to solve.
Not just work allocation — though that was where we started.
The system needed to show, in real time, where every task sat from the moment it was allocated to the moment it was completed. It needed to flag SLA risk before a breach occurred — not surface it in a spreadsheet that had already been sent to the client.
It needed to make quality patterns visible — trace errors back to knowledge gaps, close them with documentation rather than repeated escalation. It needed live resourcing visibility, not a monthly headcount model built on assumptions. Management reporting that didn't require someone to compile it the night before the review or budgeting. Client billing that flowed cleanly from the system. And it needed to be built so that AI could plug in when the time came — not retrofitted. Designed for.
When it went live, the first thing people talked about wasn't the features.
It was the feeling of knowing.
Knowing where things stood without having to ask. Knowing who was stretched and who had capacity. Knowing when something needed attention before it became a problem.
The numbers followed.
Turnaround time dropped by 68%. We recorded zero SLA breaches in the first two quarters after go-live. Client billing has been reduced from 60 days to 4 days. For the first time, leadership had a clear, real-time view of workload across the team — who was at capacity, who had bandwidth, where the bottlenecks were forming before they became problems. That visibility fed directly into resource planning and budgeting conversations that, previously, had been built on instinct and approximation. Now they were built on data.
And every single piece of work — from allocation to completion — sat inside a fully auditable flow. All data residing in a governed, secure environment. No client information in a personal inbox. No confidential client data forwarded to the wrong recipient by accident. No gaps in the trail when someone asked who did what and when. Complete auditability. Total data security. One version of the truth.
For leadership, the shift showed up in how client conversations changed. When you bring a live MI pack to a review instead of a reconstructed spreadsheet, you stop looking reactive. You start looking like exactly the kind of partner they believed they were signing with.
It is now 2026. The stakes are different.
Regulation has tightened. GDPR fines for inadequate data handling reach $20 million. The average cost of a data breach in financial services hit $6.1 million in 2025 — and over 35% of those breaches trace back to email as the entry point.
AI agents are no longer a roadmap item. They are a present operating expectation. And they cannot read an inbox. They need structured, machine-readable workflows to operate. Email is not just operationally inadequate in 2026 — it is architecturally incompatible with where work management is going.
If you are still allocating work by email, I am not here to make you feel bad about it.
I have sat across from enough capable, thoughtful people doing exactly that.
But I would ask you to do one thing.
Name what email has become in your organisation. Write down every function it is currently performing that it was never designed to perform. The task manager. The audit trail. The SLA tracker. The knowledge base. The reporting layer.
Then ask honestly — is that the architecture you want to scale?
The question in 2026 is no longer email versus workflow. It is which platform. And how fast.
First published on LinkedIn on 24 March 2026. Read the original and join the discussion

