I stop listening when a candidate opens a marketing case with a brainstorm about slogans, influencers, or "going viral." The scoring sheet does not reward creative inventiveness. It rewards whether you can turn spend into customers, customers into contribution, and contribution into a payback period the client could defend to a CFO.
That is the thesis: marketing cases are funnel-arithmetic and payback tests. CAC, conversion, retention, and contribution margin decide the answer. Brand storytelling is decoration unless it changes one of those four. After this piece you should be able to structure a budget-allocation or growth prompt in under a minute, run a clean CAC-to-payback chain, and spot the denominator traps that sink otherwise strong candidates.
What the interviewer is listening for
In a typical first round of eight, two or three candidates will treat a marketing prompt like a creativity workshop. They list channels. They debate brand versus performance as if those were tribes. They never touch the P&L of a customer.
The interviewer is listening for a different sequence:
- What is the decision (reallocate spend, raise CAC ceiling, cut a channel, fund a test)?
- What is the unit of return (incremental contribution, not last-click revenue)?
- Can you walk the funnel without losing units (impressions → clicks → conversions → retained buyers)?
- Does payback clear the client's hurdle under honest retention and margin?
If you skip step 2 and optimize last-click ROAS, you will usually recommend pouring money into branded search that harvests demand the brand already created. That is a common fail mode, not a subtlety.
The four numbers that decide the case
Most marketing prompts collapse to the same identity once you strip the jargon:
| Lever | What it means in a case | Where candidates leak points |
|---|---|---|
| Conversion | Share of traffic or clicks that become a paying customer | Mixing lead CVR with purchase CVR; counting every order as a new customer |
| CAC | Fully loaded acquisition spend ÷ new customers | Dividing by all orders or "active customers" without saying so |
| Contribution margin | Revenue minus variable COGS and variable fulfillment, per order or per month | Using gross margin that still includes marketing |
| Retention / churn | Share who stay (or leave) each period | Building LTV on revenue instead of contribution; ignoring refunds |
Payback is then CAC divided by contribution per period. If CAC is $72 and monthly contribution is $18, payback is four months. That single ratio is often enough to decide "scale," "hold," or "cut" once you have a hurdle (for example, payback under six months, or LTV/CAC above 3×).
Promotion and creative still matter. They matter as inputs that move conversion or CAC—not as a separate scoring category called "ideas."
How real filings talk about spend (and why definitions matter)
Firms do not publish a universal "good CAC." Public companies do publish how much they spend to acquire attention, and occasionally how they define acquisition cost. Those disclosures calibrate ranges and teach that denominator choice is not academic.
Angi Inc.'s 2024 Form 10-K reports domestic consumer marketing expense of $309.1 million for the year ended December 31, 2024 (down from $435.6 million in 2023), covering online marketing to search engines and other platforms plus offline TV, streaming, and radio. The filing is explicit that paid search, display, and affiliates may not be cost-effective—payback language, not a campaign brief.
Match Group's 2024 results show selling and marketing expense of $622.1 million for the year (versus $586.3 million in 2023). Its 10-K describes the mix as social, search, streaming, display, and offline, and notes that platform privacy changes (including Apple's IDFA rules) have made paid acquisition harder to measure. In a case, that maps to "your last-click dashboard is lying more than it used to."
When a company discloses CAC, read the definition. Warby Parker's S-1 defined Customer Acquisition Cost as acquisition costs (media plus Home Try-On) divided by Active Customers, and stated that CAC averaged about 15% of Average Revenue Per Customer ($218 in 2020). In a follow-up letter to the SEC, CAC attributable to new customers alone was $41, $44, and $70 for 2018–2020. Same spend pile; different denominator; different story. Interviewers notice whether you ask which definition the exhibit uses.
Platform math you can borrow without inventing folklore
Platforms publish metric definitions; large campaign samples publish rate ranges. Neither Google nor Meta publishes a single official cross-industry average you should treat as a client's true CAC. Google does publish how the metrics work: a conversion is a specified action after an ad click or view (purchase, signup, install), and the Ads API exposes average CPC and conversions-from-interactions rate as first-class campaign metrics. For ballpark rates in a case, the durable published samples come from large third-party datasets of real campaigns—use them as ranges, not gospel.
LocaliQ's 2026 Search Advertising Benchmarks (13,000+ U.S. search campaigns, April 2025–March 2026) report an all-industry average CPC of about $5.42 and an average conversion rate of about 8.18%. Those are medians across mixed lead and purchase goals; apparel converts far worse than auto repair. In a case, use ~$5 CPC and ~8% CVR as a starting point unless the exhibit says otherwise, then stress-test.
WordStream's 2025 Facebook Ads benchmark PDF (campaigns April 2024–June 2025) puts Meta traffic-objective median CPC near $0.70 and CTR near 1.71%, with lead-objective median conversion near 7.72%. Paid social often looks cheap on CPC and expensive on CAC once you multiply through a low purchase conversion rate. That is the arithmetic trap, not a channel morality play.
A clean CAC identity when you only have CPC and conversion rate:
CAC ≈ CPC ÷ purchase conversion rate
At $5.40 CPC and 3% purchase CVR, CAC is $180. At $0.70 CPC and 1.5% purchase CVR, CAC is still about $47. Different platforms; same identity.
Worked example: Rivermint subscriptions
Prompt. Rivermint sells a $48/month skincare subscription. Contribution margin after product, shipping, and payment fees is 55% ($26.40 contribution per active month). Monthly retention is 92% (8% monthly churn). The CMO has $240,000 to put into either branded search or prospecting paid social for the next quarter. Branded search shows last-click ROAS of 6.0×. Paid social shows last-click ROAS of 1.8×. Should you fund social?
Clarify first: we care about new subscribers and contribution payback, not last-click revenue. Invented but consistent funnel for the social test (numbers chosen to sit near published CPC/CVR ballparks, not copied from a real advertiser):
| Step | Social prospecting | Branded search |
|---|---|---|
| Spend | $240,000 | $240,000 |
| Average CPC | $1.20 | $4.80 |
| Clicks | 200,000 | 50,000 |
| Purchase / trial-to-paid CVR | 2.0% | 12.0% |
| New subscribers | 4,000 | 6,000 |
| CAC | $60 | $40 |
Expected tenure under constant 8% monthly churn is roughly 1 / 0.08 = 12.5 months. Contribution LTV ≈ 12.5 × $26.40 = $330 (undiscounted; say so). Then:
| Metric | Social | Branded search |
|---|---|---|
| Contribution LTV | $330 | $330 |
| CAC | $60 | $40 |
| LTV / CAC | 5.5× | 8.3× |
| Payback (months) | 60 / 26.40 ≈ 2.3 | 40 / 26.40 ≈ 1.5 |
Both clear a six-month payback hurdle. Branded search wins on efficiency. That does not mean you dump the entire $240k into branded search. Branded search mostly captures people already looking for Rivermint. If you starve prospecting, the branded pipeline shrinks next quarter. A defensible recommendation: keep branded search funded enough to protect navigational demand, put the incremental dollars into social if a geo holdout shows those 4,000 customers are mostly incremental, and refuse to scale social on last-click 1.8× ROAS alone.
What a soft answer sounds like: "Social builds brand, so I'd invest there." What a scored answer sounds like: "Social CAC is $60 and payback is about 2.3 months on $26.40 monthly contribution, but I need incrementality before I take budget from branded search at $40 CAC."
For promotion levers (price, place, product messaging), treat them as conversion and margin movers—same tree the 4Ps framework uses, just with a payback gate on the promotion P.
Ned's rule. If you cannot write CAC, contribution per period, and payback on one line before you pitch a channel, you are not ready to recommend spend. Creative without a payback chain is a mood board.
Where candidates lose the room
Revenue LTV. A $48 subscription retained for 12.5 months is $600 of revenue and only $330 of contribution at 55% margin. Using $600 against a $60 CAC invents a 10× ratio the CFO never signed.
Wrong denominator. Cost per order that includes repeat buyers is not CAC. Warby Parker's own SEC correspondence is the teaching moment: active-customer CAC and new-customer CAC can diverge sharply.
Ignoring refunds and trials. A 2% "purchase" CVR that is really a free-trial start rate with 40% paid conversion is a 0.8% paid CVR. CAC triples.
Treating last-click ROAS as causal. Search often looks brilliant because brand and social created the demand. Ask for a holdout or at least a branded versus non-branded split before you cut prospecting.
Practice this today
Take any subscription or DTC prompt and force yourself to fill this card before structure fluff: CPC, CVR, CAC, contribution per period, monthly churn, payback. Then decide scale / hold / cut against a stated hurdle.
For a timed rep with scores, start at /start (three typed turns, no account required) or run a full voice case at /interview. The case library at /consulting-cases and the math drill at /drills/math are the right surfaces for funnel arithmetic. Everything is open for seven days, no card; then $120 for a recruiting season or $49 a month.
Frequently asked questions
How do you structure a marketing case interview?
Open with the decision and the KPI (incremental contribution or payback, not clicks). Then walk funnel → CAC → contribution → retention → payback, and only then discuss channel mix, creative, or brand versus performance as levers on those numbers.
What is a good CAC in a case interview?
There is no universal good CAC. Compare CAC to contribution LTV or to months of payback under the client's hurdle. Public filings show how definitions swing the number; always state whether CAC is on new customers only.
How is marketing CAC different from ROAS?
ROAS is attributed revenue divided by spend. CAC is spend divided by new customers. You can have high last-click ROAS and still destroy value if attributed revenue is not incremental or contribution margin is thin.
Should I recommend brand or performance spend?
Neither by default. Ask for the measurement window and any incrementality evidence. Fund the channel with acceptable payback on incremental contribution; use brand spend when you can proxy lift (search lift, holdouts) or when the brief explicitly requires salience over near-term CAC.
How long should the math take in a marketing case?
Aim to land CAC and payback inside three to five minutes of focused calculation. Interviewers forgive rough arithmetic; they do not forgive never attempting the identity.
Sources
- Angi Inc. Form 10-K for the year ended December 31, 2024 — domestic consumer marketing expense ($309.1M in 2024) and description of online/offline acquisition spend. Checked 2026-09-24.
- Match Group Form 8-K exhibit (Q4/FY 2024 results) — selling and marketing expense of $622.1M for 2024. Checked 2026-09-24.
- Match Group Form 10-K for the year ended December 31, 2024 — paid acquisition mix and measurement friction (including IDFA). Checked 2026-09-24.
- Warby Parker Inc. Form S-1 — CAC definition (acquisition costs ÷ Active Customers) and customer-economics disclosure. Checked 2026-09-24.
- Warby Parker response to SEC comment on CAC — new-customer CAC of $41 / $44 / $70 for 2018–2020. Checked 2026-09-24.
- Google Ads API: Conversion management — Google's definition of a conversion (action after click/view). Checked 2026-09-24.
- Google Ads API: Metrics fields — platform fields for average CPC and conversions-from-interactions rate. Checked 2026-09-24.
- LocaliQ / WordStream Search Advertising Benchmarks 2026 — published all-industry search CPC (~$5.42) and conversion rate (~8.18%). Checked 2026-09-24.
- WordStream Facebook Ads Benchmarks 2025 (PDF) — published Meta traffic CPC (~$0.70), CTR (~1.71%), and lead CVR (~7.72%). Checked 2026-09-24.
CoachNed is independent and not affiliated with McKinsey, BCG, Bain, Deloitte, or any firm named here for interview practice.