The problem

Why could Midland Microfin's field staff not see their incentive as they earned it?

Midland Microfin is a microfinance institution that pays 4,000+ field staff on incentives, from Centre Officers to Branch Managers and the levels above them. Every payout was calculated in spreadsheets. No one in the field could see what they had earned until the payout arrived.

The business was growing faster than the sheets. Midland runs about 500 branches in 10+ states and has planned to double its branch count. Each new branch added payees, schemes and rows to files that were already hard to check. When a number looked wrong, the question went back to the team that built the sheet, and payout queries reached that team every cycle.

Why spreadsheets broke

What made microfinance incentives hard to run in spreadsheets at Midland?

The structure of the plans. Midland runs a 7-level hierarchy, and across its roles there are 50+ incentive schemes. Many vary by geography, so the same role can earn on different rules in different states.

Each loan type carries its own incentive structure: Joint Liability Group, Individual, Ajeevika, Education and Wash loans. Teams are paid on different measures, including new disbursements, on-time collection, recovery of due amounts and team performance. Payouts also move with each branch's audit grade and its customer grievances.

On top of the regular schemes, Midland runs short-term contests. Each one had to be built, and then explained to the sales team, before it could change what they did.

The buying concern

What did Midland Microfin need to see before choosing an incentive platform?

Midland needed a system that could keep pace with its growth. Calculations could no longer run in spreadsheets, and every new branch and scheme had to fit without a rebuild.

Midland ran a proof of concept with Kennect before deciding. Its team could configure schemes on the platform itself, and Kennect turned the setup around quickly. That self-serve setup, and the speed of the proof of concept, set Kennect apart from the other options Midland evaluated.

How Kennect solved it

How does Kennect run 50+ microfinance incentive schemes on daily data?

The principle Kennect built around: every scheme, contest and exception runs in the system, so nothing goes back to a spreadsheet.

• We take the datasets the Midland team integrated with Kennect and refresh them every day, so payees see current numbers.
• We configure all 50+ schemes in the calculation engine, across the 7-level hierarchy, every loan type and each geography variant.
• We let the Midland team design and launch short-term contests on the platform, and announce each one to the sales team in the Kennect app.
• We hold exceptions in the system, including excluded transactions, mid-month changes to a collection agent's portfolio and overrides to sales attribution.

Manual effort in incentive calculation has fallen by 95%. The only manual step left is uploading a few datasets. That freed 200+ man-days once spent on incentive work, and Midland now runs incentive payouts with no spreadsheets at all.

Edge cases handled

How are mid-month portfolio changes handled for collection incentives?

A collection agent's portfolio can change partway through a month. The incentive then has to reflect that change, not just the portfolio the agent ends the month with. Kennect captures the change in the system, along with excluded transactions and sales attribution overrides, so the payout follows the rule Midland set instead of a manual fix in a sheet.

Across roles:

• Field staff receive AI-powered nudges on what they have achieved and where to focus to reach their goals and earn more.
• The incentive team configures schemes and contests on the platform and uploads only a small set of datasets each cycle.
• Senior leaders, including the CFO, COO and the heads of sales, credit and operations, follow payout trends in ready incentive analytics.