Writing good PM case studies

Case studies and take-home assignments — a how-to!

I’ve completed a number of PM case studies, and reviewed even more. I’ve seen some excellent presentations from PM candidates and I’ve witnessed some clangers. There are many variations of PM case studies, and most will straddle some combination of the following, but for brevity I’ll generalize case studies into a few broad categories:

1. Create a business plan/propose a new product
2. Solve a problem for an existing product
3. Perform a specific, bounded task
4. Do free work for a company

These are my top tips for creating a clear, strong PM case study that’s memorable for the right reasons. These are the qualities I strive for, and that I look for when assessing others. These principles apply to all types of case studies. I’ll also give some examples of real take-home assignments I’ve received as a PM candidate. If you are given an assignment of the 4th variety above, proceed at your own risk. You’ll know it when you see it. When this has happened to me in the past, I’ve felt no guilt in saying thanks, but no thanks.

Let’s jump right in!

State your assumptions.

Do this up front right off the bat. You’ll always need to make some assumptions — tell your reader or listener what they are. If any of your assumptions changed anything about your approach, this is a good time to explain that. This is a critical step and one where I see many folks fall down. Done well, this is an opportunity to set the stage for your presentation and give context for your decision-making!

Make it understandable and vivid. Use visuals and plain language.

This is pretty straightforward; make your presentation look nice! Clear visuals will make your presentation more engaging. Use fewer words and more visual tools. This could be as simple as using logos in a competitive analysis slide, but you can also create slick charts and dashboards too — just make sure all of your content is clear and easy to understand. Avoid jargon or overly technical language whenever possible. You’re going for something so straightforward that anyone in a room, from any function and from any background, will understand what you’re saying.

Write for your audience.

Maybe this feels obvious, but in the event that you’re given a prompt that isn’t directly related to the business you’re interviewing with, think about what they might be looking for. Why did they give you this specific prompt? What about this prompt is relevant for the role, org, or business? What are they looking for, and what is the one thing you want them to take away from your content? Having an outline will help immensely with this.

Acknowledge shortcomings of your solution.

This could also be where you call out edge cases. You know your solution isn’t perfect, and your audience will see that too — what holes do you see in your concept, and how could you improve them?

Be wise with your time. If you’ve given guidance to spend x number of hours on preparation, stick within a 20% margin or so of that. Many PM candidates (including myself!) are fairly Type A and will want to devote extra time — which is fine, and hopefully will come across in your final product. I would caution you, however, not to go overboard — and if you do, don’t offer that information up.

I interviewed someone once who gave an impressive presentation but let us know immediately that he’d spend several weeks and many hours on the assignment instead of the suggested 2 hours, and instead of being wowed, I was disappointed every time someone asked a question and he said he hadn’t had time to cover that particular point or address some concern. It would have been completely fine if he’d just said he hadn’t had time within the constraint of the exercise, but by tipping us off that he’d worked hard on the exercise for weeks, he inadvertently raised the bar for himself and put himself at a disadvantage.

Identify a user and market. Pretty straightforward, but who is your product for? Is there a lot of growth in this market? Who is your typical user, and why will they use your product? And to that point…

Use data. Data and facts — why are you making the decisions you’re making? Almost everyone will say they’re a “data-driven PM” but there is nothing better in a product pitch like this than persuading your audience with data and sources (and please don’t quote anything like a Quora answer as a data point!).

Write an outline and make sure it flows. It’s easy to jump all over the place with these exercises, and certainly they are about product smarts and creativity — but even more, this is the opportunity to showcase your logical thinking and crisp communication. Pare it down, eliminate as much jargon as possible, and convey your ideas in a clear and succinct way — less is more! You want to tell a story: What are you building? Why are you building it? What comes next? There should be a natural arc and your audience should stay engaged. An outline also ensures you cover all the points you intend to cover. This is a good general PM skill, too.

Keep your scope tight and focused. Coco Chanel is rumored to have said, “Before you leave the house, look in the mirror and take one thing off.” That’s how I feel about product presentations. One of the more memorable interviews I’ve been in was a candidate who just had so many good ideas they couldn’t choose — they went through maybe a dozen deeply complex features in their “MVP”. Just one of those features, done well, would have been impressive. All of them felt cluttered, like the candidate couldn’t commit. It showed that they lacked the ability to drill down to what was essential to their user — they couldn’t prioritize. More crucially, they didn’t understand their user; all of their ideas addressed different use cases for different users. They tried to boil the ocean, and then some. This is not the takeaway you want your interviewer to have. Be succinct. Be pithy.

Share a roadmap. You don’t have to devote an inordinate amount of time to this, but you should be able to share what you see for your product in the future, beyond the initial solution you’re delivering. Ideally, you’d explain why you prioritized the work the way you did, and what you gain from the rest of your roadmap — does it open up new markets? Does it make the product stickier? What are some tradeoffs you’d need to make, what are some opportunities? All those great ideas you culled in the previous step should make an appearance here. This, too, should be structured and cohesive, not just a jumble of things that didn’t make the cut.

Choose something original, or that ignites your passion. You don’t need to do the trendiest thing. You don’t need to do something completely out of the box (although you can!). Just choose something that you think you’ll be interested enough in to deliver well.

If you’re presenting it live, practice. You don’t want to memorize a script, but you want to be comfortable walking through your deck (or whatever you’re presenting). You want to be familiar enough with it to know what’s next and roughly how long it’ll take to get through it. You want to know what good pacing feels like. You should do it live for someone else, at least once — a friend, significant other, roommate, whoever you trust to give you honest feedback. I generally use my parents, one of whom is deeply technical and the other of whom is not at all technical. This ensures I get a good range of feedback. One parent might ask, “but how would this scale?” and the other might ask, “but why would I use this?” You need some candor — don’t be shy. If you don’t have a human to practice with, say it out loud and time yourself (or even film yourself). Your thoughts happen much more quickly than your speech, and you won’t know your timing unless you practice aloud as you would in the real thing.

Allow time for questions and discussion. Seriously.

Example: Task assignment

This type of take-home assignment is generally not going to be delivered live, but sent in for review. It is typically a more tightly constrained exercise and I’ve seen this more commonly for early-career PMs who need to be able to execute.

This is an example of a task take-home assignment I received as a newly minted PM:

As a Product Manager for (Company), please write a development ticket/story to address the scenario described below. Start by identifying and listing issues or questions that you may need to address as a PM. If necessary, make an assumption, clearly state what assumption you made, and proceed with writing the ticket. There is no single correct way of approaching the problem and there are no hidden gotchas — just think about what information a developer may need from you in order to complete the task.

The above was accompanied by a simple schema diagram and a brief explanation of a desired change to the way a booking tool would handle incoming reservations.

I remember being so uncertain that I wasn’t being thorough enough, or that I was missing some key concept. My submission ended up taking me under 2 hours and it was just over a page. I broke it down exactly as the prompt asked:

Start by identifying and listing issues or questions that you may need to address as a PM.

In this section, I outlined some ambiguities that would impact the implementation, such as:

“We’re not currently storing the [id], which is something we need to pass as a parameter to the Request. Is there a way to begin storing this in Table A? If not, what’s the best alternative way to store this (possibly in another table using the [recordID] as a foreign key)?” I also shared some thoughts about solutions, such as long polling; this was helpful to show I was thinking about this and would be able to have a worthwhile discussion with an engineer.

Then, I wrote a simple user story, which would help the developer to understand the customer motivation and give us some acceptance criteria. It was a straightforward task, ultimately — and I ended up with an offer! The key with these is to be clear and concise.

A word on “free work” assignments.

There are a couple different flavors of this. In the first, companies just want to see if you naturally take to and understand their product and solution space, which is reasonable. In the second, companies are trying to get some free consulting work out of you. If you love the company, you may go for it, but I will almost never do one of these. I’ll give a few examples, both real assignments I encountered in the wild. The way I evaluate this type of assignment is generally, do I feel that this exercise has been given in good faith? Do I think this will be helpful to the company in their evaluation of my candidacy? If either of those are questionable, then — do I think this is a fun and challenging skill-building exercise? That may sway the decision of whether or not I participate.

Example 1

For this exercise, you will be analyzing use cases and finding product opportunities for [product].

Use Cases:

1. List any possible use cases for [product] that you can think of.
2. What is likely to be a customer’s desired outcome when using [product]?
3. In terms of generating revenue, which use case would you optimize the product for (i.e. which type of user with which use case is likely to want to pay money for a product)? (There is no right or wrong answer to this one, we solely care about what you factor into your decision)
4. What would make [product] a better product, and why?

Competitive Analysis:

1. Can you guess which companies and products are our biggest competitors?
2. What kind of data would you be interested in looking at when trying to make [product] more competitive?

Pricing + Features:

1. Look at the pricing table for [product]. List any thoughts and feedback on what you are seeing.
2. Can you think of any pricing optimization experiments you would like to run? If yes, describe them here.

Product Requirements:

Pick one idea for a product or pricing improvement. Write at least three detailed user stories for the feature that you would pass on to the engineering team.

I wasn’t sure when I received this assignment whether I should bother completing it. It was a role I wasn’t super interested in, and it did ring some “free work” alarm bells, but I decided it would be a useful PM cross-training exercise. I also implemented some guardrails. I time-boxed my effort, kept my answers short and sweet, and wasn’t afraid to say “sorry, not in scope” in my response.

My guidelines for whether something would be reasonable to include in the assignment were, 1) is this likely something they already know? and 2) is this in the spirit of helping them to see my potential role fit, and show them that I’ll be able to ramp up and have quick impact? This helped to stop me from being a perfectionist and trying to create dev-ready requirements. Ultimately they did want to move me forward in the process, but I withdrew before going on-site.

Here’s how I handled some of the above questions:

Competitive Analysis:

Can you guess which companies and products are our biggest competitors?

At a glance, it looks like [competitor 1], [competitor 2], [competitor 3], [competitor 4] and [competitor 5] are competitive — there seem like a ton of [product providers] but many of them are lightweight and not comparable to [product].

What kind of data would you be interested in looking at when trying to make our product more competitive?

I would do a basic market analysis and create a functionality grid to see what competitors are offering to identify both gaps to fill, and existing market differentiators we could highlight. I would then look at the competing pricing plans to ensure we’re correct in determining our market value. Finally, I would look in depth at our current customers. I would look at:

1. Who are our current users, and what are their use cases?
2. Are the current use cases successful? For example, if customers are trying to convert visitors to subscribers, what is the current conversion rate? How can it be improved?
3. What are customers not using — is there any functionality that doesn’t currently add value? Can we fix it? Do we need to do some customer education?

Pricing + Features:

Look at the pricing table for [product]. List any thoughts and feedback on what you are seeing.

I think the descriptors of the functionality available in the plans could be clearer or more engaging. I love the interactive examples — I would like these to be available for most of the features. I think it would be great if we could allow users to have a

clear understanding of each paid feature; for example, I know what [technical term] means, but I’m not sure all visitors would.

I like the structure of the pricing plans. [Company] isn’t capping the number of uses, which makes it potentially a cost-effective plan, and it’s very customizable. We could highlight this more by calling it out explicitly.

Pick one idea for a product or pricing improvement. Write at least three detailed user stories for the feature that you would pass on to the engineering team.

I’d like to add A/B testing as an upgrade feature. I’m not sure how rigidly the team sticks to a user story structure so I’ve kept mine in traditional user story format. These are not engineering-ready, obviously, I’d want to have a little more context and flesh them out more before passing them off to engineering — I don’t think I have room to do that in the scope of this exercise unfortunately!

“As a customer, I want to be able to create two different popups for the same use case, so I can deploy both.”

• Allow customer to link two popups
• We should allow the customer to customize, for both popups:
• Controls/Display options (frequency, delay, etc.)
• Content
• Design

Example 2

In contrast to the above, this is an assignment I did not complete. This is an excerpt from a product assignment I was sent, after being headhunted and having a call with one of the founders — emphasis is theirs:

We would like you to visit [companywebsite] to better understand our product. As you use the product, take notes on what you think we’ve done well and where we can improve. While you won’t have data to go on, you should use your product intuition and use competitor sites as a comparison.

Once you’ve completed your analysis of the website, please complete a detailed report of your findings. This should be a long form document (at least one full page but there is no maximum number of pages). Tables and figures (i.e. screenshots) are encouraged; however, formatting is up to you.

In addition to your long form report, please summarize your findings into an executive summary using Google Slides. Your executive summary should be 3–5 slides, though you can also have appendix slides with more detailed information.

You will be evaluated on the following:

• Attention to detail and thoroughness
• Insightfulness
• Presentation

This was accompanied by an email: “Due to the amount of interest for the role, we will not be able to invite all candidates onsite for in-person interviews. As such, the top performers from this take-home module will be chosen. While speed is not the primary measure of performance, it may be used to break ties.”

After reading this email to my partner (and having a bit of a giggle over it), I sent a cordial email withdrawing from the interview process. If I’d been more excited about the role, I might have reacted differently; however, this is a real time commitment they’re asking for, and I thought it was frankly disrespectful to proactively reach out to PMs and then essentially ask them to do free consulting work. I could tell right off the bat that this would not be a company I was interested in working for and I no longer wanted to pursue the opportunity. This type of open-ended consulting assignment is something you may decide to do or not as you wish (heck, it could have been a great job) — but make sure you go into it with eyes open and know that withdrawing is always a real option.

Best of luck with your assignments out there!