Kaveri Retail: A Scorecard That Is Green Everywhere While Cash Conversion Fails
Kaveri Retail operates 84 stores across six regions. Regional bonuses are paid on a scorecard weighted 40% return on net assets (RONA), 20% net promoter score (NPS), 20% sales growth and 20% shrinkage. In the Pune region, profit was ₹107.4 crore on net assets of ₹820 crore before the year's main transaction: the regional head sold the distribution centre, carried at ₹235 crore, and leased it back at ₹16.4 crore a year, removing ₹4.8 crore of depreciation. Reported RONA came in at 16.4%, against 13.1% the prior year, and the bonus was paid. Elsewhere on the scorecard, on-time delivery is measured against the promise date, which the customer-service team resets whenever a delay becomes visible; NPS rose 9 points while repeat purchase rate fell 6 percentage points; and 78% of returned customer surveys came from shoppers who were offered a ₹200 voucher at checkout. Footfall and cash conversion both deteriorated. The board has asked why every measure is green.
1. Recompute the Pune RONA after the sale-and-leaseback and explain how much of the 330 basis point improvement is real. What does Ind AS 116 do to the denominator, and why does the original measure fail the controllability test?
Read the worked answer
Recompute the reported ratio. Profit after the transaction = 107.4 − 16.4 (rent) + 4.8 (depreciation no longer charged) = ₹95.8 crore Net assets = 820 − 235 = ₹585 crore RONA = 95.8 ÷ 585 = 16.4% (confirms the reported figure)
How much is real? The region has converted ₹235 crore of owned operating assets into a fixed rental obligation. Enterprise value is unchanged; the asset base is smaller because the financing structure changed, not because the region became more productive. The asset-turnover denominator fell faster than the margin numerator, so the ratio rises mechanically.
The Ind AS 116 point. A sale-and-leaseback that qualifies as a sale does not remove the asset from the balance sheet entirely: the seller-lessee recognises a right-of-use asset measured at the proportion of the previous carrying amount relating to the retained right of use, and recognises gain only to the extent of rights transferred to the buyer-lessor. For a long leaseback on a long-life distribution centre, most of the carrying amount returns as a ROU asset, and the lease liability is added. On the correct presentation the denominator barely falls and the RONA uplift largely disappears.
The controllability failure. The regional head controls operating performance and operating assets; he does not control the group's decision to own or lease, and he certainly should not be able to book a financing gain as operating performance. Three design faults follow:
- Controllability: reward only what the manager influences.
- Definition: RONA on net book value rewards asset disposals regardless of economics.
- Risk blindness: the region now carries a 10-year fixed rent, which may be why cash conversion is deteriorating - a cost the scorecard does not see.
2. Redesign RONA, on-time delivery and NPS so that each resists gaming. For each, state the revised definition, the data source, the owner and one analytics test that would expose manipulation.
Read the worked answer
RONA → Return on invested capital on a lease-adjusted basis, with a disposal gate
- Definition: NOPAT ÷ (net operating assets + capitalised lease liabilities), measured on a rolling four-quarter average to remove year-end window dressing. Exclude gains or losses on asset disposals from the numerator; any sale-and-leaseback requires board approval before it can affect a bonus year.
- Source: general ledger, fixed-asset register, lease schedule; reconciled monthly.
- Owner: regional head for operating assets; group CFO for ownership and lease policy.
- Test: compare the change in RONA with the change in free cash flow and net operating assets - a rising ratio with falling cash flow and a falling asset base is a disposal, not a performance, story.
On-time delivery → First-promised-date compliance
- Definition: the percentage of order lines delivered on or before the date quoted at order acceptance. The promise date is locked at order entry; any change requires a logged reason code and system approval and does not reset the performance measure.
- Source: order-management system timestamp audit trail.
- Owner: operations director, not customer service (which currently sets and resets the date).
- Test: distribution of promise-date changes by user and by week; a spike in resets in the last week of the quarter exposes manipulation.
NPS → Relationship NPS with a controlled sampling frame
- Definition: NPS from a randomly drawn, quota-sampled panel of transacted customers, administered independently of store staff, reported on a rolling three months. Incentivised responses are tagged and excluded.
- Source: third-party survey platform, directly integrated with the transaction file.
- Owner: customer insights, reporting to the board rather than to regions.
- Test: survey response rate and score distribution by store and by cashier; a store whose response rate is three standard deviations above the mean while repeat purchase falls is being gamed.
Add a non-financial guardrail: any bonus payout requires the same region's cash conversion and repeat-purchase rate to be at or above prior year.
3. Explain to the board why the scorecard is green while cash conversion and footfall deteriorate, and set out how to introduce the revised scorecard without destroying trust.
Read the worked answer
Why green and deteriorating at the same time
- Measures reward what is easy to move, not what matters. RONA moved because an asset left the balance sheet; NPS moved because vouchers bought responses; delivery moved because the promise date moved. None of these required a better customer experience.
- No outcome check. Sales growth, NPS and on-time delivery were never paired with a customer-outcome measure (repeat purchase, footfall, cash conversion), so the scorecard could not detect that the inputs were improving while the outputs decayed.
- Definitional drift is invisible without a data owner. Each measure had a manager with both the incentive and the ability to change its definition - the classic condition for gaming in a responsibility-accounting system.
- Single-period, high-weight bonuses. A 40% weight on a ratio, subject to a 12-month test and no clawback, invites exactly this behaviour.
- Cash is nobody's KPI. No bonus weight attached to working capital, so the rent, the inventory build and the voucher spend did not hurt anyone.
Introducing the revised scorecard without destroying trust
- Be explicit about the reason. Present the Pune arithmetic to the regional heads; frame the change as closing loopholes, not punishing individuals. Publish the definitions in full.
- Ring-fence the past. Do not claw back this year's paid bonus unless the board concludes there was misrepresentation; apply the new rules prospectively and say so.
- Pilot for one quarter in shadow mode: report both old and new measures, calibrate targets, and let managers challenge the data before money depends on it.
- Separate controllable from non-controllable. Show each region its own scorecard with a clearly labelled "group/ownership" line that it does not control.
- Independent data ownership and audit. Definitions, sampling frames and system-generated data sit with a function that has no bonus at stake; internal audit tests one measure per region per quarter.
- Add a clawback and a multi-year element (for example, 30% of the bonus deferred over two years) so that a gain booked by shifting an asset or a cost into the following year does not survive.
- Review annually with the compensation committee - a scorecard that cannot be gamed has never been designed, only one that is harder to game and easier to detect.