The Buy-Off That Took Six Weeks: Fixing Check-Off Lending End to End
Check-off loan buy-offs took six weeks and lost deals. Desk-accurate affordability, cached compliance checks and an owned clearance stage cut that to 11 days.

Composite scenario drawn from typical East African deployments. Institution details are anonymised and figures are representative rather than attributable to a single client.
Where check-off lending actually breaks
A bank with a substantial workplace banking business had a clear proposition for salaried employees under employer schemes: apply through your employer, borrow against your payslip, repay by deduction at source. New loans worked reasonably well. The team could quote an indicative amount at the desk and, on a good week, disburse in five or six days.
Buy-offs were a different business entirely.
A buy-off — taking over a customer's existing check-off loan from another financier — was the bank's main growth channel, because most of the addressable customers already had a loan somewhere. And it routinely took four to six weeks. Customers went silent. Competitors took the deal. The bank's own relationship officers had started steering customers away from buy-offs, which meant steering them away from the largest part of the market.
Why the delay was structural
Three failures compounded.
Affordability was estimated, then re-estimated. The pre-sale calculation used a rule of thumb applied to a payslip photograph. When the file reached credit, affordability was recomputed properly against the payroll record, and the number was frequently lower. The customer had already been quoted the larger figure. The file went back for renegotiation, and the clock restarted.
Compliance checks ran late and serially. IPRS identity verification, CRB credit reference, KRA tax status and AML/PEP screening were each performed by a different person at a different stage. A failure at the fourth check invalidated the work of the first three. Nothing was cached, so a customer who reapplied went through all four again.
Clearance had no owner. Obtaining a clearance letter from the existing financier was the step that defined the buy-off, and it lived in nobody's queue. It happened by email, chased by whoever remembered. Files sat for weeks in a state that no system recognised as a state.
The rebuild
The bank moved onto Creodata Workplace Banking, running on Azure with the same .NET 9 microservices it would have run on-premise — feature parity mattered because a later regulatory review might have forced repatriation.
Affordability became authoritative at the first conversation. The pre-sale calculator computes against payslip-accurate inputs, including minimum take-home protections, so the figure quoted at the desk is the figure credit will confirm. The renegotiation loop, which had been the single largest source of restarts, largely disappeared.
Compliance checks moved into the workflow and ran together. IPRS, CRB, KRA and AML/PEP screening are embedded in the pipeline rather than bolted on at the end, wired through adapters with retry logic and 24-hour result caching. A customer who returns the next day is not re-screened from scratch. IPPD payroll verification confirms employment and existing deductions before the file advances rather than after.
Buy-off became a modelled path. This was the decisive change. The platform's workflow diverges after credit approval: new loans proceed directly to check-off booking and disbursement, while buy-offs route through an explicit clearance stage with its own owner, its own SLA and its own dashboard. A clearance request that has been outstanding for four days is now visible to the person accountable for it and to their manager. It is no longer an email.
Maker-checker dual control was applied to loan booking and disbursement, and the audit trail is protected at database level — a requirement that came from internal audit after an earlier incident in a different system where a booking record had been amended after the fact.
The outcome
Buy-off turnaround fell to a median of 11 working days, with the residual time dominated by the external financier's own clearance response — the part the bank does not control. New-loan disbursement settled at two to three days.
The more interesting shift was commercial. Relationship officers started leading with buy-offs again, because they could give a customer a credible date. Scheme employers, who had been fielding complaints from staff about stalled applications, became willing to promote the bank internally.
And per-stage SLA monitoring caught something nobody had suspected: the clearance stage was not uniformly slow. Three financiers responded within days; one accounted for most of the delay. That is a negotiation the bank can now have with evidence.
For lenders running check-off books
Quote the number you can honour. An indicative affordability figure that credit will revise downward is not a sales tool, it is a rework generator. Compute properly at the desk.
Cache your compliance results. Re-running IPRS, CRB and screening on every touch is expensive and slow, and it punishes the customer for the bank's own restarts.
Model the buy-off explicitly. If clearance is not a stage with an owner and an SLA, it is a black hole — and it is where the majority of your addressable market sits.
Further reading: The Complete Workplace Banking Loan Origination Guide · Check-Off Loans in Kenya: How Payroll-Deduction Lending Works
Growing a check-off book? Book a consultation or explore Workplace Banking to see how affordability, compliance and buy-off clearance fit into one pipeline.





