# The Weight of It
London does not let you think small.
I was standing outside our Pall Mall office at half past seven on a Tuesday morning, watching the city assemble itself before the first calendar invite had fired. Black cabs cutting lines that have existed since horse-drawn carriages. The Lloyd’s building gleaming in the kind of autumn light that makes everything look consequential. And somewhere behind me, inside a building that has witnessed more financial history than most textbooks capture, a team was preparing to present a data architecture that would — if it worked — quietly rewire how one of the world’s largest banks makes decisions.
The contrast was not lost on me. The city was built on institutions that endured. We were asking one of them to become something it had never needed to be before.
There is a particular kind of pressure that comes with working inside a two-hundred-year-old institution on something that did not exist twenty years ago. It is not the pressure of urgency — though that exists. It is the pressure of proportion. Every decision you make sits inside a structure so large and so historically layered that you can feel the weight of it before you open your laptop in the morning. London, more than most cities, makes that weight visible. It is in the skyline, yes, but more than that, it is in the specific way people here discuss change — with respect for what the thing cost to build, and measured scepticism about whether it needs rebuilding at all.
—
## What Nobody Tells You About Transforming an Institution With a Memory
In 2022, we were eighteen months into a significant enterprise data programme. The technical foundations were, by any reasonable measure, sound. The architecture was clean. The governance model had been reviewed, refined, and — I say this without pride — praised in an industry forum that shall remain nameless.
Adoption was at roughly forty percent of the intended user base.
I remember sitting with the programme lead at the time, both of us staring at a dashboard that told us everything was working, while the evidence from the business told us something quite different. The expected response would have been to redouble communication efforts, run more training sessions, escalate to senior sponsors. We did some of that.
What actually shifted things was a conversation I nearly cancelled. A senior operations lead — someone with thirty years in the institution, the kind of person who could tell you why a particular process was designed the way it was in 1997 and be correct — sat across from me and said, with no particular drama: *”We built something very similar in 2009. It was excellent. Then the sponsorship moved on and it became shelfware within a year. Why is this different?”*
I did not have a clean answer. And that — the fact that I did not have one — was the most useful thing that happened to us that quarter.
—
## Three Things London Reminded Me Are True
**The resistance is not the problem. It is the map.**
Organisations that have existed long enough to accumulate genuine history — multiple market cycles, leadership generations, technology waves — know exactly where change has broken before. The institutional memory of failure is not an obstacle to transformation. It is the most accurate diagnostic tool you have access to. When a thirty-year veteran tells you why something will not work, they are not obstructing the programme. They are telling you where the bodies are buried. The question is whether your programme is designed to listen or to override.
Most transformation programmes are designed to override. That is why the failure rate remains embarrassingly consistent, regardless of how sophisticated the technology has become. I have written before about the gap between technically excellent AI deployments and actual adoption — the pattern is the same whether you are building models or data platforms — and [the real issues tend to surface well after the proof of concept](https://lakshvaswani.com/post-of-ai-deployment-issues-as-senior-banking-executive-in-grc-space-how-we-overcame-them-going-beyond-pocs-real-issues-and-solutions-humor-engagement-and-end-with-laksh-vaswani-so-it-will-com/).
**Weight creates precision, not paralysis.**
The common framing is that legacy institutions are slow because they carry too much history. The actual problem is usually that they carry history imprecisely — they know the organisation resists, but not specifically where or why. Precision about resistance is what allows you to target your energy. Organisations that treat resistance as a general condition exhaust themselves trying to change everything. Organisations that treat it as a specific map move faster than anyone expects. We ran pilots across four regions with seventeen distinct business units not because we were being thorough for its own sake, but because we needed to know exactly which seams would hold and which ones would not before we applied any serious load.
**The human infrastructure matters more than the technical infrastructure.**
This sounds obvious. It is not acted upon as though it is obvious. [Cyber risk offers the clearest case study](https://lakshvaswani.com/when-firewalls-fail-the-human-side-of-cyber-risk/): the most sophisticated firewalls in the world are neutralised by a single person clicking the wrong link. Data transformation has the same dynamic. You can build pipelines that are technically irreproachable and still find that the actual decisions in the organisation are being made from a spreadsheet someone built in 2019 and has never shared with anyone. The technology is almost never the failure point. The trust — between the data and the person who needs to act on it — is where the work actually lives.
—
## What This Means If You Are Running Something Similar
If your transformation programme is stalling, the first question is not a technical one. It is: do you know exactly where the memory of previous failure lives in your organisation, and have you sat with it deliberately?
Not in a workshop. Not in a change management deck. In a conversation with the person who was there when the last version of this broke. That conversation is uncomfortable. It is also the only one that gives you information nobody has written down.
London reminded me of this because cities, like institutions, carry their failures in plain sight if you know how to read the architecture. The new is always being built on the specific ruins of the old. The question is whether you are reading the ground before you pour the foundations, or discovering what was buried there after something unexpected gives way.
—
The city has always known how to hold the old and build the new in the same breath.
The institutions that last — really last, not just survive — are the ones that learned to treat their own history not as ballast to shed, but as the most detailed brief they will ever receive.
—