Case types

Growth cases: "grow revenue" is not a structure

Growth prompts are about as open as they get, which is why candidates reach for a framework fastest. Ask for the constraint, rank the four sources of growth by evidence, and commit. Worked dialogue and a drill.

On this page

• Ask for the constraint before you structure
• The four places growth comes from
• Rank by evidence, not by size of prize
• What this sounds like
• Where candidates lose these
• Drill this
• Frequently asked questions

"Our client wants to grow. Where should they look?" is about as open as case prompts get, and that vagueness is exactly what makes candidates reach for a memorized framework. There is nothing in the prompt to push back against, so they fill the silence with a template.

The move that works is the opposite. A growth prompt is unanswerable as stated, and saying so is not a dodge. Grow by how much, by when, and at what cost? Those three answers turn an open-ended question into a decision with a right answer, and asking for them is the first thing that separates candidates in this case type.

Ask for the constraint before you structure

Three questions, and they take thirty seconds.

How much and by when. Growing 5 percent next year and doubling in three are different cases with different answers. Five percent comes from the existing business. Doubling almost never does.

With what. Is there capital available, and how much? A growth plan that needs an acquisition is not available to a client that cannot fund one.

At what quality. Growth that costs margin is easy and often worthless. Ask whether the client will accept dilution, because if not, half the obvious options disappear immediately.

Interviewers usually have answers ready, and when they do not, they will tell you to assume. Either way you now have a boundary, and a structure built inside a boundary is automatically about this client.

The four places growth comes from

Growth is existing or new customers, buying existing or new products. That is four combinations, exhaustive across the customer and product dimensions, which is why it is worth having in your head as a check on your own coverage. It is also the least interesting part of your answer, so move through it quickly.

More from existing customers with existing products. Frequency, basket size, share of their spend, churn you stop. Usually the cheapest and the first place to look, and usually the one candidates skip because it sounds unambitious.

Existing products to new customers. New segments, new geographies, new channels. The distribution question dominates.

New products to existing customers. You already have the relationship, so the question is whether you can build or buy the product and whether it fits how the customer already buys from you.

New products to new customers. The most expensive quadrant and the one with the least evidence behind it. It needs the strongest justification, and it is where a growth case becomes a market entry case.

Name all four. Then rank them, out loud, with reasons. The ranking is the analysis.

One caution, because this is a grid and grids are seductive. Like the profit equation and the market entry logic, this is a shape, not a script. You never announce it as a framework. "There are four places growth can come from" is fine as a way of showing you have covered the ground. "I will apply the Ansoff matrix" is the move that loses the room.

Rank by evidence, not by size of prize

Every quadrant can be made to look big with a generous assumption. The discipline is to rank by what you actually know about this client.

Ask what evidence exists for each. Is churn high, which makes retention a known quantity of recoverable revenue? Do existing customers already buy an adjacent product from someone else, which makes cross-sell a measured opportunity rather than a hope? Is there a channel the client has never used that competitors use profitably?

Evidence beats prize size because the prize is a guess and the evidence is a fact. A candidate who says "cross-sell first, because 40 percent of our customers already buy this category elsewhere" has said something no template produces.

What this sounds like

The case turned on one number the candidate went looking for, and on the willingness to say the target might not be reachable. Neither of those comes from a framework.

Where candidates lose these

Structuring before asking. Without a target, a budget and a margin constraint, every quadrant looks equally plausible and the structure has to be generic.

Skipping the boring quadrant. Retention and frequency in the existing base are unglamorous and usually where the answer lives, especially when capital is constrained.

Ranking by prize. "New geographies could be a 500 million opportunity" is a sentence you can say about any client. It contains no information.

Treating growth as separate from margin. Growth that dilutes margin can reduce profit. If the client asked for growth, ask whether they meant profit, because they usually did. That is the bridge into a profitability case.

No sequence. Four ideas presented in parallel is a list. A stack, with a reason for the order and a rough contribution from each, is a plan.

Drill this

Twelve minutes, out loud, on a business you can see from where you are sitting.

1. Set your own constraint before you start: a growth target, a capital limit, a margin rule. Write the three down.
2. Minutes 1 to 2: state all four quadrants in one breath. No detail.
3. Minutes 3 to 6: rank them and justify the ranking with a fact you actually believe about that business, not a guess about market size.
4. Minutes 7 to 10: take the top-ranked one three levels deep and put a rough number on what it could contribute.
5. Minutes 11 to 12: stack the top two or three, total them against the target, and say plainly whether the target is reachable.

Repeat the same business the next day with a harsher constraint, for example half the capital. If your ranking does not change, you were not really using the constraint.

Frequently asked questions

Interviewing somewhere specific?

Read next