Reading the ledger reports

Each report is a different lens on the same ledger. This page is a tour of the self-service pages and how to read them; the calculations behind the numbers live in the concept docs, linked as we go.

The reports overview

The overview is the landing grid: a card per report, each with a readiness or evidence badge. "Ready" means the report has the data it needs; "Available with limited evidence" means it renders but something's missing (unlinked people, absent rates); "Unavailable" means a required input isn't there yet. The panel at the top flips to "ready for finance review" once close readiness passes 90, and prompts you toward classification confidence whenever the high-confidence share drops below 80%. A board-style PDF pack exports once there's data to summarise.

Projects and people

Project portfolio breaks cost down per project: hours, total cost, capitalised (CapEx), expensed (OpEx), and the cap rate. Projects below a 45% cap rate are flagged as at-risk, and the "at-risk mix" counts how many fall under that line — a prompt to check classification, not a verdict.

People does the same per person: hours, cost, cap rate and average rate. Its framing is deliberately about concentration, not individual performance — it flags when the top three people carry 70% or more of labour cost, or one person carries over 40%, because that's a delivery-risk signal worth knowing.

The cost maths behind both is in the cost report; the capitalisation split is in the R&D capitalisation guide.

Data quality

The Data Quality report (its page is reports/unlogged) cross-checks logged time against code activity and surfaces two gaps:

• Missing time logs — a ticket with commits behind it but no worklogs at all.
• Missing commits — a ticket with logged time but no commits linked to it.

A match-rate bar shows how much of your activity reconciles. Each gap is a question an auditor could ask, and period-close readiness counts them. This report needs logged time to compare against, so on commit-session effort it reads as not applicable and points you to unclassified-commit review instead.

Insights

Insights are auto-generated cards that surface a change worth attention — classification confidence slipping, effort concentrating on one person, delivery dragging, the work mix shifting, missing rates, stale syncs. They're behind an opt-in setting, so if the page is empty your organisation hasn't enabled them.

Each card states the metric, the comparison, and why it fired. You can dismiss one (a reason is required for critical, cost or data-quality insights, so the dismissal trail stays meaningful), and dismissals are per-person. When two related insights describe the same underlying story — say a work-mix shift and delivery drag in the same period — they're merged into a single compound card that explains the link, and the two originals are hidden so you read the correlation once rather than twice.

My work

My Work is the one report your engineers see. It lists their own unclassified commits and tickets and lets them set the classification — capitalisable or not, and the type of work. It's scoped strictly to their own work, shows no cost or org-wide numbers, and it can't override a classification an admin has already made. It's how classification stays a shared responsibility rather than an admin-only chore.

Onboarding

The Onboarding report tracks how new hires are ramping: days since start, a status (On track at 75+, Watch at 50+, Check in below), a ramp score built from activity signals, and an activity sparkline. It carries an explicit banner that these are activity signals to guide conversations, not performance decisions — and it needs a baseline (a couple of people with a fortnight of activity) before the comparison means anything.

The dashboard

The dashboard is home: headline KPIs (spend, capitalisable spend, active people, hours), a spend-over-time chart, and two things that drive action —

• Alert banners, ranked high to low: a paused sync, a connected integration with nothing selected to sync, tickets needing classification, missing rates, idle capital, unlinked identities, commits with no ticket.
• The setup checklist (admins only), which tracks how close your organisation is to a complete, trustworthy ledger and lights each stage as its steps finish.