Conversion balances — design concepts
PRD-98 / CAS-1778. Three layout & flow directions for the same job: set the day this cashbook
started, enter one opening balance per balance sheet account (or upload them), see where the
difference goes, and give any bank account connected later its own opening figure. All rendered in
the real Cashbook styling and opening in the true first-run state — nothing set up yet. Compare
information architecture, hierarchy and the number of decisions; the design system already decides
the visuals. Click a thumbnail to open the full concept.
Decided — a blend of A + B.
The single page (A) is the settings screen and the surface every return visit lands on. The
guided setup (B) stays as an optional path the user can choose on first run, ending in its
read-only review. Both produce the same conversion-balance set. Concept C is not taken forward,
but its per-bank card pattern informs how the Bank accounts section renders inside the single
page. Full reasoning in
DESIGN.md.
The idea: everything on one settings screen — start
date, the account grid, and every connected bank account — with a sticky save bar that always
shows the difference going to 840 Historical Adjustments. The XLSX upload is a quiet secondary
button that reveals a file field in place.
- 👍 Closest to Xero, so it is instantly familiar; editing later is the same screen with figures in it
- 👎 The heaviest first screen — a first-timer meets a 12-row grid before they have settled the date
Critic (1–5): Xero 4 · Hierarchy 3 · States 5
The idea: a three-step wizard — start date, then
balances and bank accounts, then a read-only review before anything is posted. The date (and any
lock or warning conflict) is settled before a single figure is typed.
- 👍 One decision per screen; the review step shows the exact entries before they are posted
- 👎 A wizard is the wrong shape for the returning user who wants to fix one figure
Critic (1–5): Xero 3 · Hierarchy 4 · States 3
The idea: the screen is a hub of opening entries — one
card for the whole-cashbook set, one card per connected bank account showing Cashbook vs the
bank — and each card opens its own focused editor.
- 👍 Late-connected bank accounts are first-class, and the bank-vs-Cashbook gap is on the front page
- 👎 The common job (enter the whole trial balance) costs one extra hop into a sub-screen
Critic (1–5): Xero 5 · Hierarchy 5 · States 4
How to use this board
Walk each concept through the same lenses: information hierarchy (what's first / second / third),
edge & empty states, click-vs-think cost, and the “simpler than Xero, same power” test.
The chosen direction (or a blend) gets written up in
DESIGN.md, then built high-fidelity by
/cb-prototype.
All three drop Xero's Debit / Credit columns for one signed amount, name the plug account plainly
(840 Historical Adjustments rather than Xero's two labels for the same money), skip the unpaid
invoice / bill prerequisite entirely, and separate the two dates in copy — the start date you pick
versus the entry date the day before.
About the scores: they are the fresh-eyes critic's
second-round read (see prototypes/_kit/design-critique.md).
Concept B's second-round findings — wrong step-1 copy, no live plug row, missing grid controls — were
then fixed and are not reflected in its score, because the critique caps at two rounds.
