A common market-sizing mistake is to count everything that exists as if it were bought this year. A city may have thousands of EV-owning households, but only some will need a new home charger during the next twelve months.
This framework guide isolates that stock-to-flow problem. The general market-sizing method explains how to define a market, choose an equation and defend an estimate. Here, you will use an EV-charger example to move from an installed base to annual new installations, then check what belongs in the client's addressable market.
Say exactly what you are estimating
“Size the home-charging market” could mean chargers already installed, installations this year, revenue from hardware or revenue from installation services. Those are different quantities.
For this exercise, estimate the number of new home installations needed by households adopting an EV for the first time in a city of two million people. Exclude commercial chargers, replacement chargers and existing EV households that have not yet installed one.
Those exclusions keep the first estimate manageable. If the client needs the total annual market, add the excluded sources separately afterwards.
Confirm the geographic boundary and period too. A city, its commuter region and the whole country will have different housing and parking characteristics. State “installations per year” beside the answer so you keep a flow distinct from a stock.
Build the equation before adding numbers
A demand-side route could be:
New EV households per year × share needing a home charger × installations per adopting household
The last factor is one in this exercise. Some households may share a charger or already have suitable equipment, which is why “owns an EV” and “needs a new installation” are different conditions.
If the case provides annual first-time EV-household adoption, use it directly. If it only provides population and an ownership assumption, work towards that annual flow in stages.
A supply-side estimate can also help when you know more about installer activity than household demand. Choose the route with the clearest available inputs. Starting with population is useful in many cases, but it is not a rule that every estimate must obey.
Estimate the relevant household base
Assume 2.5 people per household and that 45% of households have parking suitable for the proposed home installation.
Two million people divided by 2.5 gives 800,000 households. Multiplying by 45% gives 360,000 parking-eligible households.
Now assume 8% of those eligible households own an EV. That gives 28,800 existing EV households within the eligible group.
| Quantity | Calculation | Result |
|---|---|---|
| Households | 2,000,000 ÷ 2.5 | 800,000 |
| Parking-eligible households | 800,000 × 45% | 360,000 |
| Eligible households already owning an EV | 360,000 × 8% | 28,800 |
Each percentage applies to a defined denominator. The 8% is an assumption about parking-eligible households. Applying a national ownership average without considering that group could produce a misleading estimate.
In an interview, explain why an assumption is reasonable for the context, then proceed. You do not need a speech defending every rounded input.
Convert the stock into annual new adopters
Suppose the number of EV-owning households in this eligible group grows by 25% over the next year. That implies 7,200 net additional EV households.
For a first estimate, assume no household leaves the EV-owning group during the year. Then net additions equal first-time adopting households. If 80% need a dedicated installation, demand is 7,200 × 80% = 5,760 installations per year, or roughly 5,800.
Now remove the no-exit assumption. If 800 households leave the group during the year, 8,000 households must enter it to produce the same net increase of 7,200. At the same 80% installation rate, demand from new adopters becomes 6,400.
That is the stock-flow relationship:
Closing stock = opening stock + entrants − exits
A replacement EV bought by a household that already owns one is not a new EV household. Nor does every additional car require a separate home charger. Keep the unit as households until you have a reason to change it.
Keep annual vehicle sales separate from the stock of vehicles already in use. The city-level numbers in this exercise are assumptions, but the distinction still changes the estimate.
Quick Guesstimate Drill
A hot-tub retailer is planning stock for next year. Its local market gains new owners, while some existing owners sell or scrap their tubs.
- Owning households
- 50,000 now, 54,000 in a year
- Households giving up
- 3,000 during the year
- First-timers buying new
- 60%, the rest buy second-hand
How many new tubs will first-time owners buy next year?
Add other demand only when it belongs in scope
If the question expands to total home installations, consider existing EV households installing for the first time and replacement or additional chargers. Estimate those separately and check the groups do not overlap.
Do not add annual EV sales to the 7,200 net additions. Vehicle sales include several kinds of purchase and would overlap with the first-time adoption already counted.
For revenue, define whose revenue you are estimating. A customer may pay for equipment and installation together, while the client earns only the installation fee. Multiplying all installations by the full customer bill would overstate an installation-only business's addressable revenue.
This is where the estimate connects to the market-entry decision: total demand, reachable demand and the client's likely sales are separate steps.
Cross-check with something genuinely different
Assume the city has 40 installation crews. Each can complete six home installations per week for 48 working weeks. Their annual capacity would be 11,520 installations.
That capacity could accommodate the initial estimate of roughly 5,800. It does not prove the demand exists. Crews might have other work or spare time, and actual throughput may be lower.
Still, the comparison is useful. An estimate of 100,000 installations would require a much larger supply response under those assumptions. Conversely, a demand estimate far below observed completed installations would suggest missing buyers or an incorrect input.
Use observed sales, deliveries or completed jobs when available; use capacity as a plausibility boundary when that is all you have. Repeating the same household assumptions in a slightly different order is not an independent check.
Show the assumption that matters most
Keeping the no-exit model and other assumptions fixed, 20% annual growth gives 4,608 installations; 30% gives 6,912. You can communicate that as roughly 4,600 to 6,900.
This range is tied to one explicit uncertainty. It is more useful than saying the answer “could be anywhere” because estimation is imperfect.
If an investment requires 10,000 annual installations to work, even the high case is below the threshold. The next task is to find a missing demand source or reconsider the investment, not refine 2.5 people per household to an extra decimal place.
Use precision where it changes the decision. Keep the final estimate proportionate to the uncertainty in its inputs, and check whether its scale makes sense.
For a related structure exercise, try EV Charging Highway Expansion. Highway charging needs a different demand equation from home installations. Explain the relevant unit and customer behaviour before choosing the numbers.
Structure drill
Would your case structure convince an MBB interviewer?
EV Charging Highway Expansion
Put a demand estimate inside a highway charging structure. Use the units and cross-check habit when explaining that demand branch. Present your approach to Casey, Case Room’s AI interviewer, then review feedback on your structure and starting point.
Quick Guesstimate Drill
A start-up plans a grease-trap cleaning service for restaurants in one city. Early interviews suggest at most 15% would switch from their current provider.
- Restaurants in city
- 8,000
- Price and frequency
- $150 a visit, four visits a year
- Revenue needed
- $1.2 million a year
What customer share would the service need to reach its required revenue?
Frequently asked questions
Pinning down what you are sizing
What does “size the home-charging market” actually mean?
It can mean several different things, so state which one you are estimating. Chargers already installed, installations this year, hardware revenue and installation service revenue are separate quantities. This guide estimates new home installations needed by first-time EV households in a city of two million people.
Should I count existing EV owners who haven't installed a charger yet?
Not under the narrow definition used here. The guide excludes existing EV households that have not yet installed, along with commercial and replacement chargers. If the question widens to total home installations, estimate that group separately and check it does not overlap with first-time adopters.
Why isn't owning an EV the same as needing a charger installation?
Some households share a charger or already have suitable equipment. The equation therefore keeps new EV households separate from the share needing a home charger. In this exercise each adopting household needs one installation, but keeping that factor visible lets you change it if the case suggests otherwise.
What if the case gives me first-time adoption figures directly?
Use them and skip the steps that derive adopters from population. The demand-side equation only needs new EV households per year, the share needing a home charger and installations per adopting household. Building adoption up from a household base is the fallback when the case does not supply it.
Turning a population into annual demand
How do I get from city population to eligible households?
Divide population by people per household, then apply the share with suitable parking. Two million people at 2.5 per household gives 800,000 households, and 45% with suitable parking leaves 360,000. Applying 8% EV ownership gives 28,800 existing EV households within that eligible group.
Why can't I treat the number of existing EV households as annual demand?
That figure is a stock, and the question asks for a yearly flow. The guide converts it using growth: 25% growth on 28,800 implies 7,200 net additional EV households. If 80% need a dedicated installation, demand is 5,760, or roughly 5,800, installations per year.
What does the “no household leaves” assumption do?
It lets net growth stand in for first-time adopters. If no household leaves the EV-owning group during the year, net additions equal new adopting households. If some households did leave, net growth would understate gross adoption, so treat this as a simplification for a first estimate.
Can I add annual EV sales to capture more demand?
No, that would double count. Vehicle sales include several kinds of purchase and overlap with the first-time adoption already counted in the net additions. If the scope widens, estimate genuinely separate groups, such as replacement or additional chargers, and confirm they do not overlap.
How should I convert installations into revenue?
First define whose revenue you mean. Hardware revenue and installation service revenue are different quantities, as the scoping step makes clear. Decide which one the client needs before attaching a price, rather than multiplying installations by a single blended figure that mixes both.
Checking and presenting the estimate
What makes a good cross-check for a demand estimate?
One built from genuinely different information, such as supply capacity. Here 40 crews completing six installations a week for 48 weeks could deliver 11,520 a year. That comfortably accommodates roughly 5,800, which is reassuring without being proof, since it measures what could be supplied rather than what households will buy.
My capacity figure is far above my demand estimate. Is my estimate wrong?
Not necessarily. Capacity shows what the market could serve, not what customers want, and crews might have other work, spare time or lower actual throughput. The comparison is most useful for flagging an estimate that available capacity could not plausibly deliver, rather than for confirming a number.
Which assumption should I flex to give a range?
The one most likely to move the answer, here annual EV growth. Holding everything else fixed, 20% growth gives 4,608 installations and 30% gives 6,912, so you can quote roughly 4,600 to 6,900. A range tied to one named uncertainty is more useful than a vague caveat.
How does a range help with an actual decision?
Compare it with the threshold the decision needs. If an investment requires 10,000 annual installations, even the high case falls below that, so the conclusion holds despite uncertainty about growth. The range matters most when the threshold sits inside it, because then the uncertain assumption decides the outcome.