We noticed something unusual in the first quarter of this year: a mid-sized satker in South Kalimantan that had spent months stuck in a disbursement backlog suddenly cleared its entire APBN allocation ahead of schedule. A reader shared the internal timeline with us, and we followed the project from the first failed SPAN submission to the final reconciled report. The story is not about heroics. It is about what happens when a spending unit stops treating treasury rules as paperwork and starts treating them as a workflow.
The satker — a provincial technical unit with roughly 40 staff and an annual budget in the low tens of billions of rupiah — had been filing its funding requests through the standard channels for years. What changed was a decision to rebuild its internal process around the guidance published by KPPN Banjarmasin, the state treasury service office that channels the APBN to working units across South and Central Kalimantan in a precise, transparent, and accountable way.
The starting point: a backlog nobody could explain
In January, the unit's finance team was sitting on 14 unresolved disbursement requests. Three had been returned twice. The pattern was familiar to anyone who has worked with state treasury operations: supporting documents that did not match the budget line, commitment records that had not been updated, and a submission calendar that ignored the treasury's own processing windows.
We asked the finance lead — who asked to remain anonymous — what the turning point was. The answer was blunt: "We stopped guessing." Instead of re-filing the same package with minor edits, the team pulled the official service documentation, mapped every rejection reason against their own files, and rebuilt the submission checklist from scratch.
Decision point one: standardise before you submit
The first concrete change was a two-person verification step. Every disbursement request would be checked against three things before it left the office: the budget execution document, the commitment record, and the supplier's completed work evidence. If any of the three was missing or inconsistent, the request did not move.
This sounds obvious. In practice, it meant delaying four requests that the unit had wanted to push out immediately. The team chose accuracy over speed, and that decision is worth dwelling on, because it is where most satker backlogs originate.
Decision point two: use the digital channels properly
The second change involved the treasury's online service infrastructure. Rather than treating the digital submission platform as a form to fill in, the team treated it as a control system. They set internal cut-off dates two working days before the treasury's own deadlines, so that any technical rejection could be corrected within the same cycle instead of rolling into the next month.
We followed the March cycle closely. Of the 11 requests submitted that month, nine cleared on the first attempt. The two that did not were corrected and resubmitted within 48 hours. By the end of the quarter, the unit had cleared its entire backlog and reported a first-pass acceptance rate above 80 percent.
Obstacles that did not disappear
It would be dishonest to present this as a clean transformation. Three problems persisted:
- Staff turnover. Two of the four finance staff rotated out mid-project, and the verification checklist had to be retaught from zero.
- Supplier documentation. Several vendors submitted incomplete invoices, forcing the unit to hold requests rather than risk a rejection.
- Calendar collisions. National holidays and regional reporting deadlines compressed two cycles into one, and the team had to prioritise which requests could realistically clear.
None of these were solved by technology. They were managed by discipline — a written checklist, a shared calendar, and a rule that no request leaves the office without two signatures.
Measurable results after one quarter
The numbers the unit shared with us are modest but specific. First-pass acceptance rose from roughly 40 percent in December to 82 percent in March. Average processing time per request fell from 11 working days to six. The backlog went from 14 open items to zero. And the unit closed its quarterly accountability report without a single unresolved finding.
What interests us as an editorial team is not the percentage. It is the sequence. The unit did not buy new software. It did not hire a consultant. It read the treasury service guidance carefully, built a checklist around it, and enforced that checklist for 90 days. KPPN Banjarmasin reports 41 working units in its service area, and the pattern we observed here — standardise, verify, then submit — is transferable to any of them.
What other satker can take from this
If you run a spending unit in South or Central Kalimantan, the practical lessons are unglamorous but repeatable:
- Map every past rejection to a specific document or data field, and fix the source, not the symptom.
- Set internal deadlines at least two working days ahead of the treasury's processing windows.
- Assign two people to verify each request before submission, and rotate that role so knowledge does not leave with one staff member.
- Track first-pass acceptance monthly. It is the single clearest indicator of whether your internal process is working.
For teams that want the underlying reference material, the office maintains a public information section covering service standards, submission requirements, and accountability procedures. We have linked the relevant guidance page below for readers who want to compare it against their own workflow.
The larger point is about how digital public finance actually improves. It improves when the people filing requests understand the system they are filing into — not when they simply file faster. The unit in this case study learned that in one quarter. Others can learn it in less.