# Relevant, sunk and opportunity costs > Which numbers belong in a decision and which only feel like they do. - Course: [Finance for Product People: From Unit Economics to Business Cases](https://productovi.com/courses/finance-for-product.md) - Module: Which costs matter for this decision? - Lesson 29/32, ~27 min - Source: https://productovi.com/courses/finance-for-product/relevant-and-sunk-costs ## 960,000 already spent **Alex Tutor:** You're still on Klasna's product team, and Leo and Hanna are arguing about an unfinished project. **Leo:** The exam simulator for maths: we've put 960,000 UAH into it over four months. Two more months and it's done. We can't just throw that money away. **Hanna:** And I say stop. It will never earn back 960,000 plus whatever finishing costs. **Alex Tutor:** You're both arguing with the same number, and it shouldn't be in the room. The 960,000 is a **sunk cost**: spent and gone, whatever you decide today. **Alex Tutor:** So how should the question be put? Pick the version that keeps the 960,000 out of the room. > **Exercise:** Klasna has spent 960,000 UAH over four months on a maths exam simulator. Finishing it takes two more months, 150,000 UAH of new costs (an editor and a licence) and should bring about 380,000 UAH of contribution in the first year. Which sentence frames the decision correctly? **Sofia:** So the question isn't "was it worth it?" but "is finishing worth it from here?" **Alex Tutor:** Exactly. Hanna, what does finishing still cost in cash? **Hanna:** The two developers stay on salary whatever we decide. On top: an exam-content editor for 90,000 and a licence for the official question bank, 60,000. **Leo:** And product expects about 380,000 of contribution margin from it in the first year. > **Exercise:** Klasna is deciding whether to finish the exam simulator or stop now. Which amounts belong in that decision? Select all that apply. ## What else could the developers build? **Sofia:** 380,000 − 90,000 − 60,000 = 230,000. That's positive, so we finish it? **Alex Tutor:** Before you decide: whose time is it, and is it really free? > **Exercise:** Finishing Klasna's exam simulator brings 230,000 UAH of contribution after its new costs and takes two developers two months. Their salaries are paid whatever Klasna decides. Leo says: "Since the developers are paid anyway, their two months on the simulator cost us nothing." What is the right reply? **Alex Tutor:** Not so fast. The developers' salaries aren't extra cash, but their time isn't free either. What's the best thing they would build instead? **Hanna:** Automatic reminders to renew packages before they run out. Product estimates 260,000 of contribution over the same year, with no outside costs. **Alex Tutor:** Then 260,000 is the **opportunity cost** of the simulator: what you give up by spending those two months on it. Compare the two uses of the same people. > **Exercise:** The developers can spend the next two months either finishing the simulator (380,000 UAH of contribution, 150,000 of extra costs) or building renewal reminders (260,000 of contribution, no extra costs). By how much does the **better** option beat the other? **Leo:** So the simulator loses, but not because of the 960,000. Because something better needs the same people. **Alex Tutor:** Right. Future costs and revenues that differ between the options are **relevant costs**. Everything else stays out, however big. If the reminder estimate falls below 230,000, the answer flips; the 960,000 never moves it. **Hanna:** And neither does the office rent or my salary. They're the same in both options. > **Exercise:** **True or false:** Klasna has spent 960,000 UAH on its exam simulator; finishing it would leave 230,000 UAH after new costs, while the same developers could build renewal reminders worth 260,000. If Klasna had already spent 2 million UAH on the simulator instead of 960,000, the right choice today would be the same. ## The same lesson in May and in July **Leo:** Different question. A company wants 200 English lessons for its team at 520 UAH each, paid by bank transfer, so no card fee. That's below our 620. Yes or no? **Sofia:** The tutor gets 390 and video costs 30. 520 − 390 − 30 = 100 per lesson. It's positive, so yes? **Alex Tutor:** In July, yes. Tutors have empty slots, so nothing else is lost: 200 × 100 = 20,000 UAH. Now ask the same question in May. **Hanna:** In May our tutors are fully booked. Every corporate lesson takes a slot from a family's lesson in the 16-pack, which leaves 184.50. > **Exercise:** In May every tutor slot is booked, and each corporate lesson displaces a family lesson that leaves 184.50 UAH. What do the 200 corporate lessons at 520 UAH add to Klasna's contribution in May (payout 390, video 30, no fee)? Enter a loss as a negative number. **Alex Tutor:** The cost of a tutor's hour depends on what else it could do. In July the slot is free; in May it costs the 184.50 it would have earned. **Alex Tutor:** Real months are rarely all free or all full. Try one in between. > **Exercise:** In June Klasna's tutors are busy for part of the month. Of the 200 corporate lessons (520 UAH each by bank transfer, no fee; payout 390, video 30), 100 fill free slots and 100 displace family lessons that would each leave 184.50 UAH. What does the deal add to Klasna's contribution in June? Enter a loss as a negative number. **Leo:** So I can say yes for July and name a different price for May. **Sofia:** Which price, though? The lowest one that doesn't lose money in May. > **Exercise:** A company wants 200 English lessons, paid by bank transfer (no fee). Each lesson costs Klasna 390 UAH of tutor payout and 30 UAH of video. In May every tutor slot is booked, and each corporate lesson displaces a family lesson that leaves 184.50 UAH. What is the **lowest price per lesson** at which the deal doesn't lose money in May? ## Your turn: settle the simulator **Leo:** Hanna and I are both still a bit attached to our arguments. Settle it for us. **Hanna:** Tell us which of our arguments was wrong, and what the right comparison is. **Sofia:** Both, in a way. Let me put it in a few sentences. > **Exercise:** Settle the simulator argument in 3–5 sentences. The facts: Klasna has spent 960,000 UAH on a maths exam simulator. Leo says it must be finished because "we can't throw that money away"; Hanna says stop because "it will never earn back 960,000 plus finishing". Finishing takes the two developers two more months (salaries paid either way), an editor (90,000) and a licence (60,000), and should bring about 380,000 UAH of contribution in the first year. The same developers could build renewal reminders worth about 260,000 with no extra costs. Say what was wrong with Leo's argument, what was wrong with Hanna's, and what the right comparison shows. > **Key ideas** > > - A **sunk cost** is spent whatever you decide (the simulator's 960,000). It belongs in neither argument: not "we can't waste it", not "it will never pay back". > - **Relevant costs** are future amounts that differ between the options: here the editor (90,000), the licence (60,000) and the expected contribution (380,000). Salaries paid either way stay out. > - **Opportunity cost** is the best alternative you give up: the reminders' 260,000 beat the simulator's 230,000 for the same developers. > - The cost of capacity depends on the season: 200 corporate lessons at 520 earn +20,000 in July but −16,900 in May, when each one displaces a family lesson worth 184.50. > - The lowest acceptable price in a full month = variable costs + the contribution displaced: 390 + 30 + 184.50 = 604.50 UAH. **Alex Tutor:** Next time a whole subject looks loss-making on a report. Keep this lesson's question ready: which costs would actually go away? - Previous lesson: [Case: plan vs actual by price, volume and mix](https://productovi.com/courses/finance-for-product/plan-vs-actual-case.md) - Next lesson: [The allocation trap: should we drop programming?](https://productovi.com/courses/finance-for-product/allocation-trap.md) --- Productovi Learning. Free, no sign-up needed to read.