Lowest Cost per Sale Is Not the Answer
TechNook sells a single gadget for BDT 12,000. Variable cost is BDT 9,000, so each sale contributes BDT 3,000. The monthly marketing budget is BDT 150,000 and the service team can install at most 160 units per month. Last month: Facebook ads cost BDT 60,000 and produced 1,200 leads at a 4% conversion rate; an SMS campaign to 5,000 past customers cost BDT 15,000 and converted 2%, and each past customer can be contacted once per quarter; a newspaper insert cost BDT 75,000 and produced 30 sales. Total sales were 178 units. Complaints about installation delays have begun and 12 customers have cancelled. Management wants to 'spend more on whatever has the lowest cost per sale'.
1. Compute the cost per sale and the net contribution after marketing for each channel, and rank them.
Read the worked answer
- Facebook: 1,200 × 4% = 48 sales; CPA = 60,000 ÷ 48 = BDT 1,250; contribution 48 × 3,000 = 144,000; net = BDT 84,000.
- SMS: 5,000 × 2% = 100 sales; CPA = 15,000 ÷ 100 = BDT 150; contribution 300,000; net = BDT 285,000.
- Newspaper: 30 sales; CPA = 75,000 ÷ 30 = BDT 2,500; contribution 90,000; net = BDT 15,000.
Total: 178 sales, BDT 150,000 spent, net contribution BDT 384,000.
Rank by cost per sale: SMS, Facebook, newspaper. All three have a CPA below the BDT 3,000 contribution, so none is loss-making - but they add very different amounts per taka. Newspaper returns about BDT 1.20 of contribution per taka spent; SMS returns about BDT 20. The useful question is not which channel is good, but where the next taka should go.
2. Recommend how to allocate the BDT 150,000 next month, given the capacity of 160 installations. Show the arithmetic.
Read the worked answer
First, the binding constraint: 178 sales already exceed 160 installations, and the 12 cancellations are the symptom. Selling more before fixing capacity destroys the value the advertising creates.
Plan for 160 sales at least cost:
- SMS BDT 15,000 = 100 sales (cannot be repeated - 5,000 past customers, once per quarter);
- Facebook BDT 60,000 = 48 sales;
- subtotal 148 sales for BDT 75,000;
- top up with newspaper: 12 × 2,500 = BDT 30,000.
Total BDT 105,000 for 160 sales, leaving BDT 45,000. Put that BDT 45,000 into installation capacity - a second technician or a contract fitter - not into more leads. If capacity reaches about 190, restore some newspaper spend, since its CPA is still below contribution. Reduce newspaper rather than eliminate it: it is weak but profitable.
3. Why is cost per sale a misleading guide here? Name the metrics you would track instead.
Read the worked answer
Cost per sale misleads in three ways in this business.
- It ignores capacity. The useful ratio is contribution after marketing per installation slot, not per taka. Facebook's BDT 1,750 net per sale is attractive, but a sale that cannot be installed becomes a cancellation and a refund.
- It ignores lifetime value. SMS reaches people who already bought. If a past customer buys 1.4 times over two years, the effective value of an SMS sale is well above BDT 3,000, so the BDT 150 CPA understates its worth - and the 5,000-name cap becomes the real limit.
- It hides diminishing returns. Newspaper's average CPA of BDT 2,500 will worsen as spend rises, while the SMS audience is fixed at 5,000. Averages say nothing about the next taka.
Track: contribution after marketing, cancellations and installation backlog, repeat purchase rate by channel, and marginal CPA for each extra BDT 10,000 of spend.