Playbook
Order to cash
Quote to collected cash, and the places it stalls. Most order-to-cash problems are not selling problems — they are data and hand-off problems that surface as late invoices and disputed balances.
- Odoo Sales & Invoicing
- SAP SD
- SAP Business One
The process
What the standard flow looks like.
- 01
Quote
Pricing is derived from a condition structure rather than typed in. If a discount cannot be explained by a rule, it will not be reproducible next quarter.
- 02
Order
The sales order is the contract: what, when, at what price, on what terms. Everything downstream inherits from it, which is why order entry quality determines close quality.
- 03
Availability and credit
Availability check confirms the promise can be met; the credit check confirms it should be. Both are automatic decisions with a human queue for exceptions.
- 04
Fulfil
Delivery, picking, packing and goods issue. Stock moves and the cost of sale is posted — the first point the general ledger learns about the order.
- 05
Invoice
Billing is generated from the delivery, not re-keyed. Tax, incoterms and currency are determined by configuration rather than by the person raising it.
- 06
Collect
Cash application matches remittances to open items. What cannot be matched becomes a dispute with an owner and an age, not an unexplained balance.
Failure modes
Where we expect to find it broken.
- Orders arrive as email and spreadsheets
- A customer sends a PDF purchase order, someone re-keys it. The error rate is low but not zero, and every error is discovered weeks later at invoicing or in a dispute.
- Pricing has outgrown its structure
- Years of exceptions accumulate as manual overrides. Nobody can answer what a customer should be paying without opening the last invoice, and margin analysis stops being trustworthy.
- Credit blocks sit in nobody's queue
- The block works correctly; the release process does not exist. Orders sit for days, and the team learns to call finance directly, which is not a process.
- Cash application is a person with a bank statement
- Partial payments, consolidated remittances and deductions each need judgement. Done manually, it is slow, and the AR ageing is wrong until it is finished.
Scope
What is standard, and what is the actual work.
The distinction matters commercially. Paying to build what the platform already does is the most common way an ERP budget is wasted.
Out of the box
- Quote, order, delivery and billing document flow
- Condition-based pricing with audit of how each price was reached
- Availability check against stock and planned receipts
- Credit limits, blocks and a release workflow
- Automatic billing from delivery, with tax determination
- Open-item accounting and AR ageing
Configured or built
- The pricing procedure: which conditions apply, in what order, and what may be overridden by whom
- Credit policy — limits, exposure, who releases and within what tolerance
- Delivery scheduling and route or shipping-point determination
- Output: what the customer receives, in what format, triggered by what
- Tax and incoterms per country sold into
- Revenue recognition where delivery and billing do not coincide
The agent layer
Where an agent earns its place.
At the edges, where documents arrive unstructured and matching needs judgement. The transaction flow itself stays deterministic.
- Order intake
- Reads an emailed PDF or spreadsheet order, extracts the lines, matches them to materials and customers, and drafts the sales order for a human to confirm. The agent proposes; it does not post.
- Credit-block triage
- Classifies each blocked order by why it blocked and by payment history, so the queue is ordered by what finance would release first rather than by date.
- Cash application
- Matches remittance advice to open items, including partial and consolidated payments, and routes genuine deductions to dispute with a proposed reason code.
Outcomes
What changes.
Before
Orders are re-keyed from email, and errors surface at invoicing.
After
Orders are drafted from the customer's own document and confirmed by a person before posting.
Before
Nobody can explain a price without opening the last invoice.
After
Every price traces to a condition record, and overrides are the exception rather than the method.
Before
AR ageing is only correct once someone has finished matching the bank statement.
After
Matching runs continuously and the ageing is current; the queue holds genuine disputes only.
Questions
What people ask.
- Can this run on Odoo, or does it need SAP?
- Odoo runs order to cash well for most businesses up to significant scale. The usual reasons to need SAP are multi-entity consolidation, intercompany flows and statutory requirements across many countries — not order volume by itself.
- Do we need to fix pricing before implementing?
- No, but you do need to decide it. Migrating an undocumented pricing structure reproduces the problem in a new system. We work it out during the design phase, which is usually the most valuable part of the project.
- Where does the AI actually help here?
- At the edges where documents arrive unstructured and where matching requires judgement — order intake and cash application. The core transaction flow is deterministic and should stay that way.

