Exact (ERP)
Invoices, due dates, actual settlement dates, payment terms, credit notes, and the open debtor position.
Revenue
Forecast cash on when customers actually pay, not when they promised to
Your liquidity forecast assumes thirty days because the contract says thirty days. Half your customers pay in forty-five, a few pay in twenty, and one large account always pays the week after their own month-end. That pattern is sitting in your ERP invoice history. Coheed learns it per customer and forecasts the date the money actually arrives.
- Systems combined
- Exact (ERP)HubSpot (CRM)
- Runs
- Daily
Sound familiar?
Teams bring us this use case when one or more of these is true.
- The cash forecast is built from payment terms and is wrong every month in the same direction.
- A late payment from one large account changes the whole month, and nobody saw it coming.
- Dunning starts a fixed number of days after the due date, regardless of what that customer normally does.
- Finance rebuilds the same spreadsheet every week from an ageing report.
How it works
Coheed runs this daily against your open debtor position and learns from every settled invoice.
- 1
Payment history is read per customer
Every historical invoice and the date it was actually settled build a payment profile per account: typical delay, variability, and whether the behaviour is drifting.
- 2
Behaviour is separated from terms
The forecast uses the observed pattern rather than the agreed term, so a customer who reliably pays fifteen days late is forecast fifteen days late instead of on time.
- 3
Open invoices are projected forward
Each open invoice gets a predicted settlement date and a confidence, which rolls up into a weekly and monthly cash projection you can plan against.
- 4
Deviations are flagged early
When a customer starts paying later than their own norm it is raised as a signal. That is often the first sign of a credit problem, and it feeds the churn and cost-to-serve picture as well.
Which data it uses
Only the fields this use case needs are read. Nothing is copied that the insight does not use.
HubSpot (CRM)
Account ownership and relationship context, so a payment signal reaches the person who holds the relationship.
Coheed
The payment profile per account, its confidence, and the drift detection that flags behavioural change.
What changes
What a team notices once this runs. No invented percentages: the effect depends on your data and your process.
- The cash forecast reflects observed behaviour rather than contractual fiction.
- Late-paying accounts are visible as a pattern instead of as a monthly surprise.
- Dunning effort goes to the accounts actually deviating from their own norm.
- A customer starting to pay slower is caught early, while it is still a conversation.
From insight to action
Collections Agent
Available as an agent
Promoted from this insight, the agent prioritises the collections list by predicted delay and account value, drafts the reminder in the right tone for the relationship, and routes anything involving a payment arrangement to a human.
Mode: Human-in-the-loop
Every insight can stay read-only, run with human approval, or run autonomously. You decide per action type, and you can change it later.
FAQ
Frequently asked questions
How much invoice history is needed?
About a year per customer gives a stable profile. Customers with fewer invoices fall back to the segment average and are marked low confidence.
Does it handle partial payments and credit notes?
Yes. Both are read from the ERP, because a customer who habitually pays in parts has a different cash profile from one who pays late in full.
Is this a replacement for our accounting software?
No. The ERP stays the source of truth for invoices and payments. Coheed adds the prediction layer on top and writes nothing back unless you ask it to.
Can it feed our treasury or bank reporting?
The projection is available as a dated series per account, so it can be exported into whatever liquidity model you already run.
Built on these maturity levels
Other use cases
Predictive Customer Lifetime Value
Score churn risk and future value per account by combining CRM engagement decline with ERP order frequency and value trends.
Lead-to-Quote Readiness
Surface correct ERP pricing, live stock, discounts, and customer agreements on every new CRM lead so sales can quote immediately.
Order Exception Monitoring
Catch stock gaps, unrealistic delivery dates, pricing errors, and missed agreements after the order, before the customer notices.
See this on your own data
Tell us your ERP, CRM, and SCM stack and we will show what this use case looks like on your systems.