Market Entry Case Framework 2026: Breakeven Before Entry

Two decisions live inside one market entry prompt

Entry cases are usually taught as four steps: size the market, check the right to win, quantify the investment, pick a route in. The steps are fine. The reason candidates still lose them is that the fourth step carries the decision and gets the least time.

There is also a threshold problem that the teaching skips. Every entry case compares a return to a bar, and the bar is a convention rather than a fact. Three different firms publish three different bars, and the same case flips answer depending on which one you use.

So the opening move is a question to the interviewer. Ask what hurdle the client applies: a first-year breakeven, a steady-state contribution, or a payback period with a stated limit. Oliver Wyman's own published case makes the stakes visible. Its model answer refuses a profitable entry because it "does not align with the client's investment criteria for a five-year payback period", checked 29 September 2026 at oliverwyman.com. The business made money. The answer was still no.

The prompt, and the public numbers under it

Nordwatt operates fast charging sites across Europe. It is considering the United States. The chief executive wants a recommendation on entry and on how to enter, with a number attached.

The market data below is public and dated. The Department of Energy's alternative fuels data center lists 82,509 charging station locations and 260,086 ports in the United States, of which 77,258 are DC fast ports, checked 29 September 2026 at afdc.energy.gov. California alone holds 20,079 stations and 68,577 ports.

The client's own operating assumptions, which you should label as assumptions out loud:

1. Site design. Four DC fast ports per site.
2. Capital. 600,000 dollars a site, all in.
3. Throughput. 120 kilowatt hours per port per day, the European average.
4. Price. 48 cents a kilowatt hour to the driver.
5. Energy cost. 22 cents a kilowatt hour, including demand charges.
6. Site fixed cost. 30,000 dollars a year for lease, maintenance and network.
7. Hurdle. Capital returned inside seven years.

One thing is missing from the public record, and it is the thing that decides the case. The Department of Energy publishes counts. It publishes no utilization and no delivered energy. Say that out loud, then treat throughput as the assumption to test.

The market entry case framework falls out of the payback equation

Payback equals capital divided by annual cash contribution. Annual contribution equals throughput times margin per kilowatt hour, minus site fixed cost. Write that and the branches name themselves.

• Margin per kilowatt hour. Price to driver. Energy and demand cost. Card and network fees.
• Throughput. Ports per site. Sessions per port per day. Energy per session.
• Capital per site. Hardware. Grid connection. Civil works. Grants, which are public money and reduce the denominator.
• Mode. Own the site, host on someone else's forecourt for a revenue share, or license the software and operate nothing.

The fourth branch belongs in the structure from the start, because it changes the denominator rather than the margin. That is the branch that separates an answer from an opinion.

Two branches deserve a line of their own here. Grants are the first: public capital reduces the amount at risk without touching the operating margin, so a grant of 150,000 dollars a site cuts our capital to 450,000 and the payback with it. Energy cost is the second, and it is the only input that can move against us after the contract is signed, because demand charges are set by the utility rather than by us.

Notice what the equation refuses to contain. Brand, driver loyalty and app quality all matter to a charging business, and none of them appears as a term. They enter through throughput, which is the only channel available to them, and that is where they should be discussed. A branch that cannot be connected to a term is a branch that will not survive a follow-up question.

Working the site economics to a 305 kWh breakeven

Margin first. 48 cents minus 22 cents is 26 cents a kilowatt hour.

Volume second, and watch the unit. Four ports at 120 kilowatt hours a day is 480 kilowatt hours a day per site. Then 480 times 365 days is 175,200 kilowatt hours a year.

Gross margin third. 175,200 kilowatt hours at 26 cents is 45,552 dollars a year. Take off the 30,000 dollar site fixed cost and the site returns 15,552 dollars of annual cash.

Payback fourth. 600,000 dollars divided by 15,552 dollars a year is 38.6 years. The equipment will not last that long. At the assumed throughput, an owned site consumes capital it never returns.

Now solve the equation backwards, which is the move that turns the case into a decision. A seven-year payback needs 600,000 divided by 7, so 85,714 dollars of cash a year, plus the 30,000 dollar fixed cost. That is 115,714 dollars of gross margin. Divide by 26 cents and the site needs 445,055 kilowatt hours a year. Spread over four ports and 365 days, which is 1,460 port-days, that is 305 kilowatt hours per port per day.

So the entire entry decision reduces to one testable number: 305 against an assumed 120, which is 2.5 times the European figure. Everything else in the case is commentary on that ratio.

Scale sets the capital. One percent of 77,258 fast ports is 773 ports. At four ports a site that is 193 sites, rounding down to a whole site, and 193 sites at 600,000 dollars each is 115.8 million dollars of capital for roughly a one percent port share.

Three firms, three thresholds, one market entry case

The firms that publish entry cases do not agree on the bar, and the disagreement is visible in their own arithmetic. This is the part of the subject the market's guides leave untouched.

Bain charges the capital to year one. Its Coffee Shop Co. case gives a 3 pound price, a 1 pound cup cost, 163,740 pounds to run the shop for a year and 245,610 pounds to open it. Its published algebra adds both cost lines: 163,740 plus 245,610 is 409,350, over a 2 pound unit margin, so "Q = 204,675 cups to breakeven", checked 29 September 2026 at bain.com. Drop the opening cost and the same shop breaks even at 163,740 over 2, which is 81,870 cups. The convention changes the answer by a factor of 2.5.

There is a second lesson in that page, and it is about reading your own exhibit. The text says to "Assume that Cambridge has a population of 100,000 people and, on average, each drinks 1 cup of coffee per day." That is 36.5 million cups a year. The recommendation instead works from 7,000,000 cups, and the narrowing steps sit inside an image we could not read. Both readings are arithmetically sound, and they disagree: 204,675 over 7,000,000 is 2.9 percent of the market, while 204,675 over 36,500,000 is 0.56 percent. Whichever the exhibit intends, the candidate has to say which base the share is taken on.

McKinsey divides fixed cost by unit margin. In Electro-Light, the drink launches "at a price of $2 to retailers" and "each bottle would cost $1.90 to produce and deliver in the new process", against 40 million dollars in total fixed costs. Unit margin is 10 cents, so breakeven is 400 million bottles. At an eighth of a gallon a bottle that is 50 million gallons. Its third question then names "the target 12.5 percent share of the electrolyte-drinks market", checked 29 September 2026 at mckinsey.com. For 50 million gallons to be 12.5 percent, the category has to be 400 million gallons. That is our inference from the firm's own numbers rather than a figure we read.

Oliver Wyman tests payback against a stated limit. Its case runs 250,000 kiloliters a day across 300 operating days at 10 rupees, which is 750 million rupees of revenue. Variable cost at 5 rupees takes 375 million, and fixed cost is 20 crore, so annual profit before tax is 17.5 crore. Profitable, and refused, because the payback exceeds five years.

The exhibit: hosting beats owning, with no capital

Read the two rightmost columns against each other. Hosting gives away 30 percent of the margin: 70 percent of 45,552 dollars is 31,886 dollars. It also removes the lease and most of the maintenance, leaving 10,000 dollars of network cost, so annual cash to Nordwatt is 21,886 dollars. That is more cash than the owned site produces, on no capital.

The comparison is uncomfortable for the version of this case candidates usually tell. At the assumed throughput, the mode choice does not adjust the answer at the margin. It reverses it. Owning loses, hosting earns, and the difference is 6,334 dollars a site a year plus 600,000 dollars of capital left in the bank.

Three market entry modes, priced on the same site

The structure named three routes in and the exhibit priced two. Here is the third, on the same site, so the mode branch arrives at the recommendation with a number against every row rather than against two of them.

Licensing means Nordwatt supplies the software and operates nothing. Assume a fee of 3 cents a kilowatt hour, declared as an assumption because no operator publishes one. On 175,200 kilowatt hours a year that is 5,256 dollars a site, since 175,200 times 0.03 is 5,256. No capital, no lease, no maintenance, no network cost.

Read the bottom two rows against each other, because that is the real choice at today's throughput. It takes 4.2 licensed sites to produce the cash of one hosted site, since 21,886 divided by 5,256 is 4.16, which rounds to 4.2. Twenty hosted sites return 437,720 dollars a year. Twenty licensed sites return 105,120 dollars, because 20 times 5,256 is 105,120.

What licensing buys instead is the one thing the other two modes cannot offer, which is a count of sites that grows without capital or headcount. And it buys the number that is missing from the public record: a licensee reports utilization, which is the 305 figure the whole decision rides on. Said that way, the third mode is an instrument for measuring the assumption rather than a way to make money on it, and that is a defensible thing to recommend alongside hosting.

The market entry call, and the number it rides on

The recommendation, in the order a partner wants it.

Enter the United States by hosting on third-party forecourts, and do not build owned sites until measured throughput clears 305 kilowatt hours per port per day. At today's European assumption of 120, an owned site pays back in 38.6 years and a hosted site pays back immediately, because it costs nothing to open.

Size it so the answer stays reversible. Twenty hosted sites return 20 times 21,886 dollars, which is 437,720 dollars a year and reveals the one number that is not public. If the measured figure comes in above 305, convert the best sites to owned, where the full 45,552 dollars of margin is retained.

The risks, with their arithmetic. Margin per kilowatt hour is 26 cents at the stated prices, so a 4 cent rise in energy cost cuts annual site cash from 15,552 to 8,544 dollars and pushes owned payback past 70 years. A host that renegotiates its share from 30 to 45 percent takes our hosted cash from 21,886 to 15,054 dollars, at which point the two modes converge. And the throughput assumption is European: the United States has 77,258 fast ports for a fleet of a different size and a different average trip, so the number is directional at best.

Where the fourth branch costs points

The weights below are ours, read from the rubric code in src/lib/grading, and no consulting firm publishes a rubric of its own. A structure is scored on Coverage 45, Separation 25, Depth 20 and Concision 10, each reported as its own line with the correction named. The reasoning behind each weight is on how an answer is assessed.

1. A structure with no mode branch. Coverage 45 is the largest single line for exactly this reason. Attractiveness, capability and investment without a route in leaves a quarter of the prompt unanswered.
2. A threshold taken for granted. Deciding on breakeven when the client underwrites on payback answers a different question correctly. That shows up in the synthesis rubric, where Answer first 35 depends on having answered the question asked.
3. Sizing in the wrong unit. Stations, ports and kilowatt hours are three different bases here. Our math rubric puts 15 points on Units, and the reason is that a unit error is invisible in the total and fatal in the conclusion.

The habit that fixes all three: write the hurdle, the unit and the mode on your page before you compute anything.

Six moves in a market entry case interview, and the thirty minutes

One firm publishes the clock. Accenture's case workbook states that "Case interviews typically last from 45 to 60 minutes, with a focus on the case itself lasting 30 to 45 minutes", checked 29 September 2026 at accenture.com. Take the lower bound and budget the case above against thirty minutes. The allocation is ours. The minutes sum to 30.

1. Hurdle, unit and mode, out loud. Two minutes. Ask which bar the client underwrites on, say which unit you are counting, and name the three routes in. All three are free to fix here and expensive to fix later.
2. Branches from the payback equation. Four minutes. Write capital over annual contribution, then let the terms become the branches. A structure derived from an equation cannot contain a branch with nowhere to attach.
3. Margin per unit. Two minutes. 48 cents less 22 cents is 26 cents. One subtraction, said aloud, and it is the input every later number multiplies.
4. Volume and annual contribution. Five minutes. Four ports at 120 kilowatt hours, times 365 days, times 26 cents, less the 30,000 dollar fixed cost. Four operations, each written on its own line.
5. Solve the equation backwards. Six minutes. This is the move that produces 305 against 120, and it is the most valuable six minutes in the case. A forward answer gives a payback. A backward answer gives the client a threshold to go and measure.
6. Exhibit, mode comparison, close. Eleven minutes. The largest block, because it holds the comparison and the recommendation, and because the close is the part that runs out of clock.

The order has one hard rule inside it. Moves five and six are the only two that produce a decision, and together they take 17 of the 30 minutes, which is 57 percent. A case where structure and setup have eaten twenty minutes has already lost the half that gets scored.

Where the entry question is really about the structure of the industry rather than the arithmetic of entering it, Porter's five forces decides which branch pays.

Drilling the entry arithmetic

Rework the case with the sensitivities changed. Set throughput to 250 kilowatt hours per port per day: energy sold becomes 4 times 250 times 365, which is 365,000 kilowatt hours, and gross margin becomes 94,900 dollars. Annual cash is 64,900 dollars and payback is 9.2 years, which still misses the seven-year hurdle. Now you know the bar sits between 250 and 305, and you have bracketed it by hand.

Then swap the mode and redo it. Hosted cash at 250 kilowatt hours is 70 percent of 94,900 minus 10,000, which is 56,430 dollars a year on no capital.

Our breakeven calculator does the algebra above, and the math drill times this kind of chain and returns Answer, Working, Units and Interpretation separately. Both are free and need no sign-in. Neither knows anything about charging economics, and neither will choose your entry mode, which is the judgment the client is paying for. Our assessment simulators cover firm built formats, and that page says itself that none of the six is live.

For the case where the client is already in the market and the profit moved, see the profitability framework. For entry into a store format rather than a network, see retail cases.