kvitta new to BC
Live on Business Central, or nearly ·6 min read

The implementation covered everything except the reconciliation that actually hurts

Posting groups, dimensions, VAT, the item card: all designed properly. Then the first Adyen payout arrives, covering three hundred orders in one net amount, and somebody opens Excel. Six weeks later the unapplied entries are four figures and nobody is sure which are real.

Free to start with nothing installed, no partner hours, and nothing to design during hypercare.

Why this gap exists in almost every e-commerce implementation

Three reasons, none of them anybody's mistake.

1

Business Central assumes one customer pays one invoice

Which is true for wholesale and almost never true online. A payout is one net amount covering hundreds of orders, minus fees, minus returns, with a reference that belongs to the provider rather than to your invoice series.

2

The demo used clean data

In the demo the reference matched. In your files it is SO12356-1 against invoice 12356, and one character of difference is the whole reason a person is now doing this by hand.

3

Nobody owned the decisions

Which account fees go to, what counts as a rounding difference, what to do with a payment 7,56 short. Small policy calls that were never made, so they get made again every day, slightly differently.

Week one

A clean AR ledger by Friday, without opening a support ticket

Monday
Ten minutes of setup
Create the workspace, drop three days of files, let Kvitta read your sources and propose the first rules.
Tuesday
Make the four policy calls
Fee accounts, rounding threshold, what to exclude, how far a date may drift. Read each one as a sentence and approve it.
Wednesday
Clear the backlog
Run the weeks you are behind on. Old entries are a handful of patterns repeated, which is exactly what rules are for.
Friday
A short routine that somebody else can run
The routine is drop, decide, post. It is written down as rules, so it is no longer one person's private knowledge.
Wakakuu did exactly this and went from 55 minutes to 9. Read the case study

Safe to add during hypercare

Nothing to install to start
The free plan installs nothing. A connected plan adds one small Kvitta app, published by us and kept current across release waves. The rules stay in Kvitta.
A journal in your batch
The journal is balanced to 0,00 and lands in your batch: pasted on the free plan, written by Kvitta on a connected plan, and posted in the batches you allow.
No partner hours needed
Set it up yourself in an afternoon. If your partner wants to look first, everything is readable in plain sentences.

Questions from new BC teams

We are still in hypercare. Is it too early?

No, and earlier is usually better. No matching logic goes into Business Central, so it cannot interfere with anything being stabilised, and a clean AR ledger makes every other issue easier to see.

Do we need the API connection to Business Central?

Not to start. An open items export works fine on day one, and you can connect properly later when IT has time.

Our chart of accounts is not finished.

Fine. Account numbers sit in the rules and are edited in one place, so changing 6570 later takes seconds rather than a reimplementation.

Should our partner set this up?

They can, but the person who does the reconciliation is the better choice. The decisions are accounting policy rather than configuration, and they are the one who knows the answers.

We post one revenue journal a month from payout reports. Does this help?

Probably not. Kvitta works against open customer ledger entries, so without receivables per order there is nothing to match.