Make or Buy at Kima Industries: A Decision Inside the Margin of Error
Kima Industries makes 20,000 units a year of component X for its own assembly line. The cost card per unit shows: direct materials Ksh 110, direct labour Ksh 70, variable overhead Ksh 30, and fixed overhead absorbed at Ksh 60. Total fixed overhead is Ksh 1,200,000, of which Ksh 400,000 is head-office apportionment that will not change and Ksh 800,000 is the component's supervisor and a machine lease that both end if the component stops being made. Twenty per cent of the direct labour is a salaried technician who would be redeployed elsewhere in the factory. An outside supplier offers all 20,000 units at Ksh 250 each. If production ceases, the freed floor space can be let for Ksh 300,000 a year. In-house, 5% of units need rework at Ksh 40 each; the supplier guarantees no defects. The works manager opposes outsourcing 'because our own cost is only Ksh 270 a unit'.
1. Compute the relevant annual cost of making component X and recommend whether Kima should outsource.
Read the worked answer
Rule: include only future costs that differ between the alternatives. Exclude absorbed head-office overhead and any cost that continues regardless.
| Relevant cost of making (per unit) | Ksh |
|---|---|
| Direct materials | 110 |
| Direct labour, avoidable portion only (80% × 70) | 56 |
| Variable overhead | 30 |
| Avoidable fixed cost (800,000 ÷ 20,000) | 40 |
| Opportunity cost of freed space (300,000 ÷ 20,000) | 15 |
| Expected rework (5% × 40) | 2 |
| Total relevant cost of making | 253 |
Buying: Ksh 250 per unit, with no rework.
Annual comparison: make 253 × 20,000 = Ksh 5,060,000; buy 250 × 20,000 = Ksh 5,000,000. Outsourcing saves Ksh 60,000 a year.
Critique of the works manager: the Ksh 270 he quotes includes the Ksh 400,000 head-office apportionment and the full Ksh 70 labour, neither of which is escapable. His figure double-counts costs that stay behind and ignores the rent forgone.
Judgement: the saving is only 1.2% of the buy price - well inside the estimating error on rework rates, rental values and supplier reliability. It is not a decision to take on the arithmetic alone.
2. Test how robust the recommendation is to changes in annual volume, and state the volume at which the decision reverses.
Read the worked answer
Let Q = annual units. Making costs 196Q (materials 110 + labour 56 + variable overhead 30) + 2Q rework + 1,100,000 of avoidable fixed cost and rent forgone.
So: make = 198Q + 1,100,000, and buy = 250Q.
Set them equal: 250Q = 198Q + 1,100,000 → 52Q = 1,100,000 → Q ≈ 21,154 units.
Interpretation:
- At the budgeted 20,000 units, buying is cheaper by Ksh 60,000.
- Above about 21,154 units a year, making becomes cheaper - a volume rise of just 6% reverses the decision, because the fixed cost and rent are spread more thinly.
- If volume falls the case for buying strengthens sharply: at 15,000 units, making costs 198 + (1,100,000 ÷ 15,000) = Ksh 271.33 a unit against Ksh 250 - buying saves about Ksh 320,000.
Judgement: the recommendation is fragile. Buying wins by a margin smaller than the forecast error in volume, so signing a multi-year fixed-price contract could lock Kima into the wrong answer if demand grows. Also check whether the supplier's Ksh 250 is an introductory price - once Kima's own machine lease is gone, it has little bargaining power.
3. Identify the non-financial factors that should influence the final decision.
Read the worked answer
- Supplier dependence. The supplier becomes a single source for a critical component; any strike, quality failure or insolvency stops Kima's assembly line, whose losses dwarf Ksh 60,000.
- Price risk after exit. Once the machine lease is surrendered and the supervisor leaves, reopening production is slow and costly, giving the supplier hold-up power at the next price review.
- Quality evidence. The 'no defects' guarantee is only as good as the supplier's record; ask for defect data and penalty clauses, and note that in-house rework may be a symptom of a fixable process problem rather than a reason to exit.
- People. The supervisor's redeployment or redundancy has morale, legal and knowledge-retention consequences; the technician's redeployment should be confirmed, not assumed.
- Capacity and lead time. Buying may require higher buffer inventory and longer lead times, tying up working capital that offsets part of the saving.
- Confidentiality. Component X may embody design information Kima does not want in a supplier's hands.
- Strategic fit. Is component X core? If it is not, the thin saving plus exit costs may still justify outsourcing as a focus decision - but it should be argued on those grounds, not on the Ksh 60,000.