Home

The shadow-close process

Compare the close before
you cut over.

We reproduce one period in parallel with your current process, then show what matched, what differed, and the source evidence and policy behind each BeanStack entry. You decide whether the result supports cutover.

What a shadow close is

Your close, run in parallel.

Parallel

Run alongside your current process

BeanStack closes the period on its own, in parallel with whatever you do today — nothing about your existing process changes while you evaluate.

The diff

A line-by-line comparison

You see where BeanStack agrees with your books and where it differs, with the reasoning and source evidence behind each entry.

Replay

Replayed against real history

Every decision is replayed deterministically against the historical decisions your human bookkeepers made. The replay shows reproduction and differences — you and your reviewer judge correctness.

No commitment

You cut over when convinced

Nothing is filed and no money moves during a shadow close. You adopt only once the scorecard holds up — not before.

How it works

How the shadow close runs.

01

Connect and import

Connect bank feeds, billing, and documents. Keep your existing ledger, or import your QuickBooks history — chart, transactions, and balances — onto ours in hours, not months.

02

Close in shadow

BeanStack runs the period close under your policies — reading documents, posting routine entries, and holding exceptions for review.

03

Review the scorecard

You get a diff against your current books: what matched, what differed, and the evidence and policy behind every BeanStack decision.

04

Cut over on evidence

Once the scorecard convinces you, BeanStack takes over the monthly close; a named controller reviews exceptions and owns delivery.

What the review packet shows

The month-end
variance report.

Matched

Matched entries

Where both processes reproduced the same treatment — listed line by line, not summarized away.

Differences

Differences, classified

Grouped by cause — policy, source data, timing, or unresolved judgment — with the evidence behind each side of every difference.

Held items

Held for review

Material or low-confidence items BeanStack did not post automatically, and exactly what it needs to resolve each one.

Evidence

The trail behind every entry

Source document, applied policy, action taken, reviewer, and change history — attached to every line in the packet.

Cutover

Criteria agreed up front

The pass/fail bar is set before the run — you judge the packet against criteria you agreed to, not a moving target.

Why it de-risks the switch.

You adopt on evidence, not a sales promise.
Nothing is filed and no money moves while you evaluate.
Every difference is shown and classified against your own historical books.

See the comparison
on your own books.

Run a parallel shadow close in your first month. Keep your ledger, or move onto ours — you cut over only when the scorecard convinces you.