Energy retailer journey twin · independent assembl concept

Four journeys through one power company

A generic New Zealand electricity retailer, not a real one, seen from eight homes on one street: a household switching in, one moving house, one opening a bill 40% higher than last month's, and one plugging in a new EV. Each journey runs the same nine stages. The part that matters most is the wait, and what the customer sees while they wait.

assembljourney twin
Snapshot: Auckland 17.3 °C · wholesale 11.4 c/kWh · grid 98% renewable
SIMULATED

What the customer sees

9:41●●●
Simulated · a phone, not a product

Right now on the streetSimulated

What power costs here

What power costs in Auckland

Three public figures and one empty slot. The wholesale price is what the grid charges by the half hour; the residential price is MBIE's modelled average for the area across every retailer; the market share says who sells power on this street. No public dataset carries a retailer's plan, so that slot stays empty until the retailer connects its own.

Snapshot
The lanes

Nine stages, four journeys, and where the time goes

Every journey runs entry, intent, context, recommendation, commitment, action, wait, fulfilment and resolution. The striped block is a wait the customer can see. Below, the hours each stage takes as the retailer is (hollow) and with its levers on (filled).

Live, on the street

Days from entry to resolutionEstimates

The year

Two thousand households, twelve months

The cohort is billed every month with Auckland's seasons in the usage. Switch-ins, moves, bill shocks and plan changes happen at the rates on the sources list, and each one costs the retailer contacts, back-office work, and sometimes the customer.

Bills that shock, by month

What each journey costs the retailer a yearSimulated

Show the table volumes, contacts, customers lost, cost
What if

Five changes to try

All PROPOSED. Switch any on and the year runs again for the whole cohort. The dollar figure on each is that change on its own against the retailer as it is: cost to serve saved, plus the margin kept when a customer stays, less any wholesale saving.

The settingsEstimates

None of these three are public for any NZ retailer. Move them to your own and the ledger follows.

Customers lost through the journeysSimulated

Contacts a year, by journeySimulated

Proof ledger

What the changes are worth

Show the table
    AI and automation

    Where an agent helps, and how far it may act

    Five levels of authority, set per job by the retailer. Nothing here is running today: each row is a PROPOSED assembl concept. The evidence column says who measured the result, and whether anyone checked it independently. Credit decisions, disconnection and any change to a customer's price stay with people.

    Seven jobs, how far each may actPROPOSED

    Show the table what each job observes, drafts, recommends and may do
    Connect your data

    Switch the twin from simulated to yours

    Start with exports the retailer already has: a billing extract, a CRM contact log, and the switch and move event log. Files are read inside this browser tab. Nothing is uploaded or sent anywhere, and columns that look like personal information are ignored on load.

    No files yet. Every figure is simulated, a source or an estimate.
    Sample CSVs to copy (made-up numbers, no real customers)

    Generated in this page from the simulated cohort, so a demo never touches real data. The billing file carries a customer name column and two broken rows, and the contacts file an agent name column, to show the mapping, validation and privacy steps.

    The data contract: three tables
    Sources

    Figures used, and how far to trust them

    Show the table