Market sizing walkthrough: EV charging ports in Texas
A market sizing case worked end to end: two chains, every assumption stated, arithmetic run in a script, a sanity check, and what moves the answer most.
Updated October 28, 2026
What is the question actually asking?
“How many public EV charging ports will Texas need in 2030?” is a market sizing prompt. The interviewer does not have the number; they have a view of a reasonable path to one. Before you size anything, ask the two questions that change the chain: what counts as a public port (any plug, or a fast charger?), and what year. On this walkthrough: public ports of any speed, assume a 50 kW average, year 2030.
Which way in: demand or supply?
Two chains reach the same number. Demand side: how much energy EVs will draw in public, divided by what one port can deliver. Supply side: ports per vehicle today, scaled to the future fleet. Say which you are taking and why. The demand chain is used here because every link is a number you can defend; the supply chain needs a ratio nobody in the room knows.
The chain, one assumption at a time
- People. Texas has about 31,000,000 people. Round aggressively; the precision is false anyway.
- Vehicles. About 0.8 vehicles per person gives 24,800,000 vehicles. State the ratio and move on; the interviewer will correct it if they care.
- Electric share. Assume one in ten vehicles is electric by 2030: 2,480,000 EVs. This is the assumption the answer is most sensitive to, so flag it as such out loud.
- Energy per EV. 12,000 miles a year at 0.3 kWh a mile is 3,600 kWh. Then the public share: most charging happens at home, so take 20% in public.
- Public energy. 2,480,000 x 3,600 x 0.2 is about 1786M kWh a year. Say “about 1.8 billion kilowatt hours” and keep going.
- What one port delivers. A 50 kW port, used 10% of the hours in a year: 50 x 8,760 x 0.1 is 43,800 kWh. Utilization is the second big assumption; name it.
- Ports. 1786M / 43,800 is about 40,767, call it 41,000 public ports.
Does the number pass a sanity check?
Two checks, both from the same script. 41,000 ports for 2,480,000 EVs is about 60.8 EVs per public port. And 1.32 ports per thousand people. Say the first one out loud: “sixty EVs per public port, which sounds like a busy but workable network.” If a check comes out at six EVs per port or six hundred, an assumption is wrong and you say which one you would revisit first (the electric share, then the utilization).
What changes the answer most?
| Assumption | Used | If instead | Ports become |
|---|---|---|---|
| Electric share of the fleet | 10% | 20% | about 82,000, double |
| Port utilization | 10% | 20% | about 20,500, half |
| Public share of charging | 20% | 30% | about 62,000 |
| Average port power | 50 kW | 150 kW fast chargers only | about 14,000 |
Only after the number and its sensitivities are on the table does a recommendation belong: where to put the ports, which assumption to buy data on first. The interviewer is grading the chain, the checks and the order, not the final digit.
How do I say it in two minutes?
Objective and definitions (15 seconds). The chain, named before you compute it (20 seconds). The arithmetic, out loud, rounding at every step (60 seconds). The sanity check and the two assumptions that move it most (25 seconds). Then stop and ask what the interviewer would like to pressure-test. In practice, run a sizing prompt in a browser tab and let Casepartner show the math check after each number; if its sanity ratio and yours disagree, one of you rounded wrong, and it is worth finding out which.
Questions
How do you approach a market sizing question?
Define the thing being counted and the year, pick the demand or supply chain and say why, state each assumption before you use it, round at every step, then sanity check the result with a ratio before you say anything about what to do.
Should I use the demand side or the supply side?
Whichever has links you can defend. The demand chain here uses population, vehicles, electric share, energy per vehicle and port throughput; each is a public figure or a stated assumption.
What if my number is different from the interviewer’s?
Expected. The interviewer grades the chain and the checks. Say which assumption you would revisit first; here it is the electric share, then utilization.
Does Casepartner do the math for me?
It shows a math check after a number lands: the setup, the answer, one sanity ratio. The chain and the assumptions are yours, and it never recommends before the math is checked.