London, 1858. The flush toilet had overwhelmed the cesspits and the Thames drove Parliament from its chambers. Bazalgette did not patch the old system: he rebuilt it end to end, and doubled every pipe beyond his own calculations to carry a city that did not exist yet.
A bank has committed to 200,000 new customers a year for five years, and its sign-up process cannot deliver that. Your leaders decide how far to automate it, what that costs to build and to run, who does the work that remains, and what governance they buy before something goes wrong.
Where the journey begins, and where most people give up.
Know-your-customer checks. Security against patience.
Credit and risk scoring. Who gets in, and who is wrongly turned away.
Creating the account in systems built decades apart.
Features and preferences. Compliance against flexibility.
The moment a name on a form becomes a customer.
Each step can stay as it is or move up a tier of automation, and the steps have to work together, because completion multiplies along the chain.
Illustrative arithmetic: if every step keeps 85% of the people who reach it, 38% of those who start finish.
A process that keeps most people at every single step still loses most of them by the end. That arithmetic surprises almost every room.
One process on the table, and a room that has to agree where the automation stops.
Automation pays, up to a point. Every tier gets more people through the process, and every tier costs more: once to build, and then for as long as it stays in service, in licences, infrastructure and the work of keeping it upright as the systems underneath it change. Past a point the extra cost of going further, to build and to run, outgrows the extra savings it buys, and the five-year total climbs again.
So the first question is not which tier to buy. It is whether the step is worth automating as it stands, because automating a poor process only makes it a faster poor process.
The app prices every choice across the life of the contract rather than the first quarter. Recognition goes to the teams that improve the process, make the plan pay, retrain their own people and take the longer view. Not to the teams that spend the most, and not to the teams that avoid the question.
Every figure in the simulation is sourced, dated, and shown to the room. Where we have estimated rather than measured, the app says so.
Inside a Tech Tempest day, where the same teams have already designed the service they are now being asked to deliver. Or on its own for a leadership team, shaped to the group and the time you have.
A trained facilitator runs the session. Your people need no preparation and nothing to read beforehand.
A room, a screen, and a laptop per team. Your IT team needs shift127.com allowed through the network, or the app will not open.
The simulation, the process model and the figures behind it. Nothing is installed on your network.
Tell us whether you are thinking about The Great Stink on its own or as part of a Tech Tempest day, the size of the group and who is in the room, and we will come back with dates and a price.
Book a Session