TMT means telecom, media, and technology-as-distribution (streaming, ads, cable). It is not a B2B SaaS packaging case, and it is not an IT-program case.
Last full pass: 24 August 2026.
A TMT case interview (McKinsey, BCG, Bain, and media/telco practices at the Big Four) is about subscribers, ARPU, churn, and content or network cost. Telco adds spectrum, towers, and load. Media adds ads vs subscriptions and content amortization. If the client is Salesforce-like software, use the technology case interview. If the client is a retailer “going digital,” use digital transformation. TMT interviewers punish people who talk NRR when the metric on the page is ARPU and monthly churn.
What you are deciding
- Price vs churn: a $2 hike that lifts ARPU 18% and monthly churn from 3.5% to 5.0% can shrink the steady-state base.
- Mix: premium vs ad-lite vs bundle with mobile.
- Network / content spend: 5G or a sports rights package is a step-fixed cost; utilization (load factor on a cell, hours watched) has to pay for it.
- M&A: content library vs distribution. Synergies are often overlapping subscribers, not 30% opex.
Clarify prepaid vs postpaid (telco), SVOD vs AVOD vs linear, and whether ads are first-party or programmatic. Ask whether content cost is fixed minimum guarantees vs per-hour. Ask whether a price hike applies to the back book or only to new subs — telco and streaming both hide this. For a mobile-network prompt, ask coverage vs capacity: adding subscribers in a congested cell is a capex case, not an ARPU case.
Open with: (1) subs × ARPU × churn as a system, (2) mix (plan, ads, bundle), (3) content or network step-costs, (4) regulation / spectrum, (5) overlap in bundles. If you open a SaaS NRR tree, you are on the wrong page.
Exhibits you should expect
| Exhibit | Trap |
|---|---|
| ARPU and churn by cohort | Blended ARPU hides the cheap-plan mix shift. |
| Content cost / sports rights | Cost is lumpy and contracted; volume of hours does not scale it down. |
| Ad load vs completion / churn | More ads raise ads ARPU and can raise SVOD churn. |
| Network capex and utilization | Coverage vs capacity. Empty rural spectrum is not “unused revenue.” |
| Bundle overlap | 1+1 subscribers ≠ 2 if they already take both products. |
Units that trip people
ARPU is monthly; churn is often monthly. Annualizing casually (×12) without compounding churn is sloppy. Approximate annual survival ≈ (1 − monthly churn)^12.
Subscribers vs hours vs ad impressions. Ads businesses sell impressions or reach, not subs. A streaming sub who never watches is still content cost if rights are per-title minimums.
Mbps vs GB vs spectrum MHz. Telco engineering units show up in exhibits. You do not need to be an RF engineer; you do need to not multiply MHz as if it were revenue.
Practice a TMT subscriber-economics case
Run a scored media or telco case where ARPU, churn, and content or network cost have to meet.
Worked mini-case: the AVOD conversion that shrinks revenue
Prompt. HarborStream has 8.50m paid SVOD subscribers at $9.80 ARPU. Annual subscription revenue = 8.50m × 9.80 × 12 = $999.6m. Product wants to auto-convert 15% of subs to an ad tier at $5.99 plus $3.20 ads ARPU (case inputs). Content cost is a fixed $410m (sports + studio output deals). Will the ad tier help?
Math. Converted subs = 0.15 × 8.50m = 1.275m. They used to contribute 1.275m × 9.80 × 12 = $150.0m. They would contribute 1.275m × ($5.99 + $3.20) × 12 = 1.275m × $9.19 × 12 = $140.6m. −$9.4m on that cohort.
Remaining SVOD: 7.225m × 9.80 × 12 = $849.7m. Total revenue $990.3m vs $999.6m. Content cost unchanged at $410m. You made the P&L worse and you trained 1.275m households that $9.80 was optional.
Recommendation. Do not auto-convert. Offer AVOD only as a win-back or pause path (churn-save), or as the default for new price-sensitive geos, with a holdout so you can measure incremental ads vs lost sub ARPU. If ads fill rate is seasonal, the $3.20 is a peak-month number — annualize with a haircut. Next exhibit: churn by price-sensitivity segment and ads fill by geography.
What a generic profitability tree misses here
- Price and churn are a coupled system, not independent volume and price boxes.
- Content and network are step-fixed. Extra subs can be high contribution or can force a new rights tier / a new layer of capex.
- Ads ARPU is not subscription ARPU. It depends on inventory, fill, and brand-safety, and it can cause churn.
- Bundles double-count customers.
- Regulation (net neutrality, spectrum licenses, kids’ advertising) binds the feasible set.
If you answer HarborStream like a SaaS company, you will talk NRR and seats. If you answer it like a transformation program, you will talk agile squads. Neither pays the sports rights.
See where you stand on a TMT case
Practice ARPU–churn tradeoffs and the content-cost so-what.
Related guides
CoachNed is independent and unaffiliated with McKinsey, BCG, Bain, or other firms named for interview-style practice. Cases on CoachNed are AI-simulated.