Unit Economics for Ads — Explained Simply
Unit economics answers the key question: “How much can we afford to pay for a customer without losing money?”Without it, advertising becomes gambling: today it “works,” tomorrow it burns budget — and you can’t explain why. 🧩 A “Unit” = One Customer (or One Order) Don’t look at “overall ad spend.” Look at the economics of one customer: "how much money they bring" VS "how much they actually cost". 📌 The simplest rule: how much you can spend to acquire a customer Your max ad spend per customer = the profit you keep from that customer. If you spend more — you’re buying customers at a loss. 🧾 Start with margin, not revenue Revenue ≠ profit.What matters is what remains after costs. Example: Order value = 1000 Cost of goods = 600→ Margin = 400 This 400 is your base for calculation. 🎯 CAC: Cost to Acquire a Customer CAC = ad spend / number of new paying customers. Not cost per lead — cost per real buyer. ⏳ Payback: how fast ads “pay for themselves” Payback = how long it takes for margin to cover CAC. If payback is long, you need a cash buffer — otherwise ads will drain your working capital. 🔁 LTV: how much a customer brings over time LTV matters when you have repeat purchases or subscriptions. But it’s easy to overestimate. Keep two numbers: - Real LTV for 60–90 days - Long-term LTV (used cautiously) 🛑 The biggest mistake: focusing on ROAS without knowing your margin ROAS can look great, but if margins are low — or returns/logistics/fees eat into it — you can still be losing money. 🧠 A quick way to calculate “max CAC” **Max CAC ≈ margin from first purchase + margin from repetitions (for the real period) − extra costs (delivery, packaging, fees, support).** 🚨If your CAC is below this — ads make sense.If above — either adjust the offer/margin, optimize ads, or increase repeat purchases. 📈 How to apply this without complex analytics 1️⃣ Define your target CAC. Anything above it = stop/fix. 2️⃣ Check the funnel: if CAC is high, find the weak spot (creative, landing, offer, retention).