If network (scheme) fees keep “surprising” you, you don’t have a fee model.
You have an invoice.
How do you turn a network fee invoice into a forecast?
Invoices explain the past. CFOs and product leaders need the next 30–90 days.
Where a 30-minute monthly review handles the tactical question ("what changed?") a fee model handles the strategic one: "what will happen?"
Because network fees hit profitability before anyone has time to react.
I put the cheat sheet in the graphic.
The only point I’ll add here: you don’t get a forecast until you baseline.
Baselining means being able to see, invoice-by-invoice, exactly what you're paying in network fees.
If you can’t consolidate historical invoices into something you can trend, you’ll always be reacting.
Even a clean baseline fails if optional services stay enabled by default.
Are network (scheme) fees a forecast in your org, or still a month-end explanation?
And modeling errors compound: for instance, issuers who forecast CEDP like Level 3 systematically underestimate exposure.