Solution Generation and Curation

IN THIS LESSON

Once you have chosen a customer and a pain point, you need to identify credible ways to solve it.

A useful target is three alternative solutions. That is enough breadth to show that you considered more than one approach and to create a meaningful decision later, while still leaving enough time to explain each idea well. Coming up with three good solutions can be difficult, so this lesson also gives you a practical way to recover when the ideas do not come immediately: return to the pain and ask why it exists.

The objective is to produce a small set of real product directions that respond directly to the problem you selected. Each one should contain enough substance for the interviewer to understand what would change for the customer and why that change could help. You will define the experience in greater detail in the next lesson and choose among the alternatives after that.

What this task proves about you

Solution generation shows whether you can translate customer understanding into product possibilities.

You need to demonstrate that you can consider several ways of changing the experience, recognize which ideas have a plausible relationship to the problem, and exercise judgment about which possibilities are worth serious consideration.

The interviewer is therefore looking for more than creativity in the abstract. The customer should explain why the pain matters, and the pain should explain why each proposed solution might create value. An unusual idea that does not plausibly improve the customer’s experience demonstrates less Product Sense than a familiar mechanism that addresses the problem directly.

Aim for about three credible solutions

Three solutions is a useful target. If you can identify three credible approaches, you have usually created enough breadth to demonstrate that you considered alternatives and enough contrast to make later prioritization meaningful.

You may occasionally think of more than three strong ideas. In that case, more is not inherently better. Every additional alternative consumes time that could be used to make the strongest ideas understandable, so it is usually better to narrow the visible set and explain three well than to race through five or six thinly.

The more common problem is that three good ideas do not arrive immediately under interview conditions. Take time to think before you begin presenting them. A short pause to write down your ideas is more useful than narrating the brainstorming process while you are still trying to decide what the products are.

Suppose you have chosen a goal-based learner who struggles to use limited preparation time effectively. Three credible directions might be:

• Diagnostic assessment. Identify the learner’s most important knowledge gaps so they can spend less time reviewing material they already understand. This helps because limited study time is more valuable when it is directed toward the weaknesses most likely to affect the outcome.
• Adaptive study plan. Combine the learner’s remaining time, other commitments, exam requirements, and current knowledge into a plan that changes as they progress. This helps because the product takes on the allocation problem the learner is struggling to solve alone.
• Targeted expert help. Connect the learner with an expert when they encounter a high-value concept they cannot resolve efficiently on their own. This helps because a small amount of well-timed human assistance may prevent one difficult topic from consuming a disproportionate amount of the remaining preparation window.

These descriptions are intentionally brief for this lesson. At this stage, the interviewer needs to be able to envision the solution well enough to understand what each product direction is and why it belongs in the solution set. The next lesson will help you make the customer experience and its mechanism of action more concrete.

This is also a natural place to use Why-Anchoring. When you introduce an alternative, state the idea, provide whatever explanation is necessary for the interviewer to understand it, and close the loop on why it addresses the selected pain. Once the why is clear, that solution is a complete unit of reasoning and you can move to the next one rather than continuing to add features or commentary that do not advance the argument.

When you are stuck, investigate why the pain exists

If you cannot think of enough credible solutions, go one level deeper into the problem.

Suppose your pain point is that a learner does not have enough time to prepare effectively for an exam. That experience can arise for several different reasons. The learner may underestimate how much preparation is required, fail to account for work or other commitments when making a plan, spend substantial time reviewing material they already understand, or discover too late that the plan they made is no longer realistic.

Those different causes suggest different intervention mechanisms. If learners are surprised by how little time they actually have, a planning product might combine their study requirements with the rest of their schedule and show what preparation time is realistically available. If they waste time on material they already understand, a diagnostic assessment could identify the gaps that deserve the most attention. If they repeatedly fall behind without realizing it, a product could compare actual progress with the remaining work and continuously revise the plan.

The useful move is to ask what would have to change for the pain to become smaller. That causal question gives you somewhere productive to search for solutions instead of trying to brainstorm product categories in the abstract.

You are not performing a full Root Cause Analysis or claiming that one speculative cause has been proven. You are quietly using plausible causes as inspiration for different ways a product could intervene. The purpose is to generate better alternatives, not to add another formal stage to the Product Sense answer.

Make sure you actually have a solution

Before you count an idea among your alternatives, make sure you are proposing something meaningfully new to build or change.

Suppose you are designing a product at Meta to help people volunteer and propose that people use WhatsApp to organize with other volunteers. That may be useful advice, but people can already do it. If the proposal requires no meaningful change to the software or experience, you have not yet described what the product team should build.

Using an existing product is perfectly reasonable when you meaningfully extend it. You might instead propose a volunteering experience within WhatsApp that identifies relevant opportunities, shows which ones people in a user’s network are joining, and makes it easy to organize participation with them. Now there is new product behavior, and you can explain why that behavior could reduce the coordination burden that prevents people from acting on an intention to volunteer.

The same problem appears when candidates name a technology or product surface instead of describing a solution. Saying that you would “use Meta AI,” “build it in Groups,” or “make it AR” may identify an enabling technology or location, but it does not tell the interviewer what new customer experience exists or why it solves the pain. Technology can be part of the answer when it helps explain why the solution is credible in the strategic context—for example, because of a company advantage or broader technology trend—but it should not become the primary definition of the solution.

A useful minimum test is to ask what the product team would actually have to build that does not exist today and why that new behavior would improve the selected customer’s experience. You do not need a detailed specification yet, but there must be enough substance to know that you have proposed a specific product rather than an existing capability, a technology label, or a vague direction.

Keep ambitious ideas credible

Credibility does not mean that every solution has to be conservative or feasible with today’s technology. It means that the proposed world should make sense if the difficult barriers can be overcome.

Imagine a parking case. A product that could predict when individual parking spaces will become available several minutes before drivers arrive may be technologically difficult today. If it worked reliably, however, the customer value is easy to understand: drivers could spend less time circling and arrive with greater confidence that parking will be available.

Now imagine solving the same problem by dispatching tow trucks to move parked cars somewhere else whenever another driver needs the space, then returning each car later. Much of the required technology already exists, but the labor, coordination, cost, liability, and inconvenience make the operating model obviously unattractive. Physical possibility is not the same thing as product credibility.

This distinction becomes particularly useful in interview environments that reward moonshot thinking. A moonshot is audacious because important technical or practical barriers remain, not because the idea ignores economics or operating reality. Building a sustained human presence on Mars is a useful analogy: the barriers are enormous, but the concept describes a coherent future that could be valuable if those barriers were solved. An idea that remains unattractive even after you generously remove its hardest constraint is not improved by calling it a moonshot.

Strategic context can also help you generate and strengthen solutions. Company assets are one input, but so are competitor behavior, substitutes, market structure, technology shifts, regulation, and other facts from the strategic environment that change what is possible or attractive. A social graph may make a coordination or recommendation product more powerful at Meta. Maps data may make a location-dependent product more plausible at Google. A competitor may demonstrate that customers already value a behavior, while a technology shift may make an experience possible that previously required expensive human work. These facts should help you solve the pain and justify why the resulting idea is credible in the case.

Present a decision set, not a brainstorm

By the end of this stage, the interviewer should have a small set of alternatives that are real enough to compare later. You do not need to narrate every idea you considered, and you do not need to choose the winner yet.

A live transition might sound like:

I took a minute to think through a few ways we could address the time-allocation problem. I see three directions worth considering. One is a diagnostic that identifies the learner’s biggest gaps, which helps concentrate the limited study window on what they do not know. Another is an adaptive planner that combines the exam requirements with the learner’s actual schedule and keeps reallocating the time they have left. The third is targeted expert help for high-value topics where they are otherwise likely to get stuck for too long. I want to make each of those a little more concrete before I compare them.

The example gives the interviewer a compact view of the solution space while still explaining why each alternative belongs there. It also creates a clean handoff to the next task: making the products concrete enough that the interviewer can picture how they work.

The point

Aim for about three credible ways to solve the customer problem. When those ideas are hard to find, investigate why the pain exists and use plausible causes to uncover different intervention mechanisms. Make sure each alternative is a real product solution, explain enough for the interviewer to evaluate it, and Why-Anchor the idea so its relevance to the pain is clear before you move on.