cac-and-conversion
/cac-and-conversion
Liczy koszt pozyskania klienta, czas zwrotu i wartość klienta oraz sprawdza, czy cena pokrywa koszty, dla każdego modelu przychodów.
- Autor
- cao-partners-au ↗
- Repozytorium
- olc-agent-skills-starter ↗
- Licencja
- MIT
- Ostatnia zmiana repozytorium
- 12.11.2026
Ocena i bezpieczeństwo
- Pliki w paczce
- 2
- Skrypty
- 0
- Komentarze HTML
- 0
- Adresy URL w SKILL.md
- 0
Brak sygnałów. Repozytorium zmienione 12.11.2026, bez odwołań do wycofanych modeli i API.
MIT, otwarta: można używać i zmieniać, z zachowaniem informacji o autorze.
Co robi i kiedy
Liczy koszt pozyskania klienta, czas zwrotu i wartość klienta oraz sprawdza, czy cena pokrywa koszty, dla każdego modelu przychodów.
Wyzwalacze z opisu
Oryginalny opis (EN)
Work out what it costs to acquire a paying customer (CAC), how fast that cost comes back (payback), what a customer is actually worth, and whether the price covers it. Works for any revenue model: one-off sales, subscriptions and memberships, retainers, deposit-then-balance or milestone billing, and commission or percentage fees. Use when asked "what is our CAC", "how much does a customer cost us", "how long until a customer pays back", "is this channel paying for itself", "should we raise prices", "what is a lead worth", "what can we afford to pay per lead", or "where do deals stall".
Jak działa
- Step 1: standard CAC first
- Step 2: compare equal, mature windows
- Step 3: payback, which depends on how you get paid
- Step 4: what you can afford per lead
- Step 5: the funnel, and where it stalls
- Step 6: realised revenue per customer, never list price
- Step 7: the pricing read
- The quality trap
- Reporting it honestly
- What makes this wrong
Podgląd SKILL.md
--- name: cac-and-conversion description: >- Work out what it costs to acquire a paying customer (CAC), how fast that cost comes back (payback), what a customer is actually worth, and whether the price covers it. Works for any revenue model: one-off sales, subscriptions and memberships, retainers, deposit-then-balance or milestone billing, and commission or percentage fees. Use when asked "what is our CAC", "how much does a customer cost us", "how long until a customer pays back", "is this channel paying for itself", "should we raise prices", "what is a lead worth", "what can we afford to pay per lead", or "where do deals stall". --- # What a customer costs, how fast it comes back, and whether the price covers it Cost per lead is a marketing number. **Cost per customer, and how quickly that customer pays it back, is a business number.** It decides pricing, budget, and whether a channel is worth running. Run this after `marketing-roas`, which produces the spend and attribution inputs and reconciles revenue against the accounting system. --- ## Step 1: standard CAC first Most businesses should start here, and many never need more. **CAC = acquisition cost in a period ÷ new paying customers acquired from that spend.** - **Paid CAC (per channel):** that channel's ad spend ÷ new customers that channel produced. Never divide one channel's spend by all new customers (see `marketing-roas`). - **Blended CAC:** all marketing spend ÷ all new customers, including organic and referral. It looks better than paid CAC; report both and label which is which. - **Fully loaded CAC (for pricing decisions):** add agency/freelancer fees, tools, creative production and the sales time spent closing. State what is and isn't included. A "customer" is someone who **paid**, not a lead, a booking or a signed quote. --- ## Step 2: compare equal, mature windows Most CAC errors are window errors, not maths errors. - **Match the windows.** Spend from 1 to 30 September goes with the customers who came from leads created 1 to 30 September. Never divide a full month's spend by a partial month's sales, or this month's spend by customers who came from last month's leads. - **Only judge mature cohorts.** Measure the time from lead to first payment (median and slowest). A cohort only counts once its leads are older than that lag. Report younger cohorts as "not yet mature", never as a trend in either direction. - **Say the dates** on every figure and compare equal-length periods (full weeks or months). Example of the trap: dividing three months of spend by the customers won so far, when the newest leads haven't had time to buy, overstates CAC. Matching each month's spend to its own cohort and dropping the immature weeks can move CAC by 10 to 20%. --- ## Step 3: payback, which depends on how you get paid CAC says what a customer costs; **payback says how long that cash is tied up.** Two businesses with the same CAC can be in completely different positions. Use **cash received** (payments that landed), not invoices raised, and **gross margin** where delivery has real costs. | Revenue model | How to read payback | |---|---| | **One-off sale** (product, job, project paid on completion) | Payback is immediate if first-order gross profit ≥ CAC. If not, payback only comes from repeat purchases: measure the repeat rate and time to second order before assuming it. | | **Subscription / membership / retainer** | Payback months = CAC ÷ (monthly revenue × gross margin). Check with real cohort retention: find the month cumulative gross profit per original customer first covers CAC. Early churn makes this much longer than the simple ratio. | | **Deposit then balance, or milestone billing** | Count each payment when it lands. Payback is reached at the first payment whose cumulative cash covers CAC. Track how many customers stop after the first payment; their revenue is the deposit only. | | **Commission / percentage fee** (agents, brokers, finance) | Use the achieved fee, not an assumed average sale. Allow for long settlement lags; recent cohorts are a floor, not a result. | | **High-ticket with a sales process** | Same as the model it bills under, but measure the lead-to-paid lag first; it sets how long you wait before judging a cohort. | Report payback **per product or tier**, never blended. A cheap entry product and a premium one can share a CAC and have opposite payback. Planning reference: payback inside 3 to 12 months is comfortable for most small businesses, but the real limit is your cash runway. If a cohort hasn't paid back yet, say so; don't project a date. --- ## Step 4: what you can afford per lead **Breakeven cost per lead = average revenue per customer × lead-to-customer rate.** Example: $2,000 average first-year revenue × 1 in 20 leads becoming customers = $100 per lead breaks even on revenue. Use gross profit instead of revenue for the true ceiling, then set the target below it by the margin you need. **Breakeven cost per click = target cost per lead × landing page conversion rate.** These come from your own deal maths, not platform benchmarks, and every pause or scale decision should key off them. --- ## Step 5: the funnel, and where it stalls Write the actual sequence with the money marked, in the business's own words, e.g. ``` enquiry -> call/quote booked -> call/quote held -> proposal accepted -> PAID (-> repeat / renewal) ``` Count how many of a matched cohort reach each stage. **The biggest drop is usually worth more than any advertising change.** Work out what one percentage point at that stage is worth and compare it with the cost of buying more leads. Common winners: show-up rate for booked calls (reminders and confirmations), quote follow-up speed, and first-payment collection. If payment happens in stages, treat each payment as its own event and report how many customers only ever trigger the first one. --- ## Step 6: realised revenue per customer, never list price Average **actual payments received** over every customer acquired, including the ones who cancelled early, stopped after a deposit, or bought the entry product only. The gap between list price and realised revenue per customer is usually the most useful number this skill produces. Report the distribution, not just the mean (completed at full price, entry tier, part-paid, refunded), so a cheaper tier or high drop-off doesn't hide inside an average. --- ## Step 7: the pricing read 1. **Does the first payment cover CAC?** If a customer who only ever makes the first purchase or payment returns less than CAC plus delivery cost, they are unprofitable. Know what share of customers look like that. 2. **Is a cheaper offer carrying the same CAC?** A lower price doesn't halve the cost to win the customer. If the mix shifts toward the cheaper offer, revenue per customer falls while CAC holds. 3. **Is the stall cheaper to fix than the lead?** Usually yes (Step 5). --- ## The quality trap Ad platforms optimise toward whatever you tell them a conversion is. Fed form fills, they buy cheap form fills: cost per lead falls while customers don't. Fix, in order: 1. **Judge ads and channels on CRM outcomes** (held calls, quotes accepted, payments), not on the platform's lead count. 2. **Send later stages back to the platforms** (Meta Conversions API lifecycle events, Google offline conversion import) so they learn which leads become customers. Optimise for a later stage only once an ad set reaches roughly 50 of that event a week. 3. When the platform and the CRM disagree, **the CRM wins**. --- ## Reporting it honestly - State the CAC definition and what's excluded. - Give paid and blended CAC side by side, with dates and cohort maturity. - Give payback per product/tier and say which revenue model it assumes. - Separate cohort revenue from monthly sales. - Count referrals and repeat business separately; they cost nothing to acquire and flatter any blended figure. - End with a decision: raise this price, fix this stage, hold or move this budget. ## What makes this wrong - **Invoices raised instead of cash received.** - **Unmatched or immature windows** (Step 2). - **One customer billed under several names or entities**, reading as several customers or a missing one. Reconcile per customer, not per invoice. - **Blending tiers or products** in payback. - **Ignoring labour** in a pricing decision. Advertising CAC excludes the people who sell and deliver; say so, or the price gets set against half the truth. Method notes on windows, payback and the quality trap adapted from coreyhaines31/marketingskills (MIT).
Pliki w paczce (2)
- README.md 2 KB
- SKILL.md 8 KB
Wersje
W bazie nie ma innych skilli o nazwie cac-and-conversion.
Migawka bazy z 13.11.2026 · wersja paczki e3ba78c71373 (skrót SHA-256 całej paczki)
Pobrania w ostatnich 30 dniach: 0 · ranking
Zgłoś problem z tym skillem: szkodliwa treść, błędny opis albo prośba autora o usunięcie.
Podobne skille
- 100unit-economics-analysis
Liczy koszt pozyskania klienta, marżę wkładu, LTV, stosunek LTV do CAC i czas zwrotu, z podziałem na segmenty i kanały. - 90unit-economics-cac-ltv-payback
Pomaga ocenić opłacalność pojedynczego klienta, czyli CAC, LTV, okres zwrotu i marżę kontrybucyjną, przy decyzji o zwiększeniu wydatków na wzrost. - 100biz-cac-ltv
Oblicza i analizuje koszt pozyskania klienta (CAC) i wartość klienta w czasie (LTV), by ocenić unit economics, opłacalność pozyskiwania i efektywność wydatków marketingowych. - 90unit-economics-expert
Liczy CAC, LTV, czas zwrotu i marżę klienta oraz pokazuje, która dźwignia poprawia wynik. - 100local-unit-economics
Modeluje ekonomię lokalnej firmy usługowej: koszt leada według kanału, wskaźniki rezerwacji, stawienia i zamknięcia, CAC, gotówkę z pierwszych 90 dni, LTV do CAC i werdykt samofinansowania. - 100unit-economics-check
Sprawdza CAC, LTV i czas zwrotu w biznesie subskrypcyjnym na podstawie własnych danych i wyznacza bezpieczny limit CAC.