The Operating Model · How We Build
Everything is changing.
We're building the right thing (the wedge). This is about building it the right way — when intelligence is abundant, continuous and cheap, the 40-year rules of how software gets made no longer hold.
Code
is no longer scarce
The bottleneck
moved to judgment
Value
flees code
Almost everything we assumed about building software must be rethought.
It will take years and the endgame isn't visible. Better to keep up than sit it out. The shifts that actually bind how we build — tap each.
Value flees code — and lands exactly on Coniq's moat.
From a completely different starting line — software economics, not the loyalty market — the agent era arrives at the same moat our build strategy already chose.
- Code itself
- Software as code-someone-else-wrote-for-you
- Software that encodes workflows
- The "you can't build this yourself" moat
- Data
- Permissions
- Distribution
- Trust
- Compliance & regulatory position
- Physical assets
Squeezed will be vendors whose moat was “customers cannot justify building this themselves.”
Much of traditional loyalty SaaS is exactly that — “we wrote the loyalty engine so you don't have to.” When an operator's agents can stand one up in an afternoon, that moat evaporates. Coniq's defensibility must come from the data, the physical venues, the relationships and the trust — never the code.
That's the wedgeConiq needs a camp at the frontier.
Large organisations can't learn this through committee. The answer is a small autonomous team built around the models — whose job is to discover the new way of working and pull the company toward it.
A small autonomous team, built around the models
Not to add AI to the old way of working — to discover the new way and pull Coniq toward it. Close enough to touch real systems, real data, real constraints and real consequences.
The lighthouse is the vehicle
One Value Retail / MAG-class operator: real data, real money, real stakes. The frontier camp learns the new way where it actually counts.
The unit of work is the delegated task
Not the code to be written. Let agents run long, in parallel, with less supervision than feels comfortable. An agent forced to work like a human is a wasted agent.
Output is software, people and practices
The camp's product is also the working methods the rest of Coniq adopts next — and the engineers who become system thinkers and operators focused on business value.
Therefore
The strategy doesn't change. The way we build it does.
Build for a world where intelligence is abundant, continuous and cheap — not for today's model ceiling.
Optimise for learning what that world looks like.
Don't preserve old software processes out of habit.
Don't wait for the end state to become obvious — play the game.
Go to the frontier first — that's where the future is learned.
Go to the frontier first — because that's where the future is learned, and where Coniq's moat is proven against real consequences.
Next chapter
And the company itself — how does it operate?
Connected systems, self-improving operations, humans vs agents, and the team we need.