The case study process is genuinely good. The scoring may not be up to par. I want to be clear about something before anything else — I’m not writing this because I’m angry about not getting the job. I’m writing this because I engaged seriously with this process, I think the bones of it are genuinely excellent, and I believe the specific critique I’m about to make could make it meaningfully better. Take it in that spirit. The case study itself is well designed. Asking candidates to work through a realistic operational scenario with real data, call recordings, and supporting materials is a legitimate and differentiated way to assess how someone thinks. The request for a separate working notes document showing your thinking process in real time is particularly strong. The AI disclosure policy is honest and reasonable. And the people who leave angry reviews about having to do a case study at all are probably exactly who the process is designed to filter out. If you’re not willing to engage seriously with a real problem before you get the job, that tells you something important. I have no issue with the ask itself. The issue is more specific than that — and it comes directly from reading their own published materials carefully. There is a contradiction worth naming. The rejection email stated my submission didn’t align with their expectations, and that feedback couldn’t be shared because it would reveal expected answers. The Glassdoor response article on their Substack confirms this openly — the case study problems are ones they’ve already solved, graded against known answers. Here’s where it gets interesting. Their careers page lists thirteen company values. Several of them describe precisely the behaviors a rubric with predetermined answers CAN’T evaluate. Value #1 — Noticing and Curiosity — states: “We actively look for clues to disconfirm our deepest held beliefs and are continuously updating our mental model of the world.” My working notes document had an entire section dedicated to hypotheses I formed and then systematically disproved. I spent more time disproving my own initial assumptions than building the final argument. That is value #1 demonstrated in writing. A rubric looking for a specific destination OR ‘expected’ answers cannot see that. Value #5 — Think from First Principles — states: “We don’t believe in asking how can we be 10% better because the status quo is the wrong starting point. Instead, we imagine the ideal impossible customer experience from scratch and back our way into reality.” My pain point argument reframed the entire problem from the most obvious surface-level metric to the systemic question underneath it. That is first principles thinking. It didn’t ask which problem was biggest. It asked why none of them were staying solved. A rubric with a predetermined answer would miss this entirely. Value #11 — See for Yourself — states: “We listen to support calls, not just read reports or look at metrics. We pull random ops and support tickets and read what happened, who said what, look at timestamps.” I engaged with every source material provided — cross-referencing qualitative and quantitative inputs and using both together to build the argument. That value is described almost word for word in how I approached the exercise. And then there’s the Bill Gates story from their High Quality and Fast article — the entire point being that what makes someone exceptional is caring about what’s right, not who’s right. Being willing to change your mind when presented with better reasoning. That maps directly to Value #6 — Humble. A rubric with predetermined answers is structurally incapable of rewarding that quality. You cannot demonstrate “I don’t care who’s right, I care what’s right” when the right answer is decided before you start. A candidate who reframes the problem entirely — who rules out the obvious answer and builds a more sophisticated argument from the first principles of the data — may score lower than a candidate who converged on the expected answer more cleanly. That is not a measure of intellectual honesty or first principles thinking. That is a measure of alignment with whoever wrote the rubric. There is a second smaller contradiction worth naming. The Glassdoor article says the case study is primarily a writing communication assessment. How you express and communicate ideas through writing is what’s being evaluated. But the rejection email says submissions need to align with their expectations. Those are different judgments. One evaluates how you communicate. The other evaluates whether you arrived at the right answer. It is worth being clear about which one is actually being scored because they are not the same thing and candidates deserve to know which bar they are being held to. The fix is not complicated. I understand the operational reality. Almost 40k applicants in a single month requires a fast filter and a rubric is a practical concession to scale. That is a legitimate constraint. But here is the thing, Clipboard is a technology company that openly uses AI in its own work. AI evaluation frameworks exist that can assess reasoning quality, hypothesis testing, intellectual honesty, and first principles thinking at scale without requiring predetermined answers. The same values described on the careers page could be built directly into the evaluation criteria itself. You do not have to choose between scale and alignment with your stated values. That is an operations problem with a real solution. And frankly the kind of problem a strong RSO hire would be excited to work on. The people who built this process clearly care about getting hiring right. The working notes ask alone tells me that. It is genuinely one of the most thoughtful things I have seen in any hiring process. This is not a broken process. It is a good process with one fixable misalignment between what the values say and what the scoring rewards. I would genuinely encourage them to look at it. (Hopefully they read this with that lens) What I’d tell other candidates: Engage seriously with the case study. Don’t treat it as a box to check. Treat it like day one on the job, which is exactly how they frame it. The working notes document is not optional (even though they say so). I would think it’s probably where you’re most authentically evaluated. Read their published values and operating principles before you start. And understand going in that there are expected answers, so the question isn’t just what you think. It’s whether your thinking lands close enough to what they were looking for. If you’re the kind of person who finds that process energizing rather than annoying, you’re probably their kind of person. The exercise itself is worth doing regardless of the outcome. I came out of it with sharper thinking and a clearer framework for how I approach operational problems. I genuinely hope they consider this feedback. Not because I want to relitigate my submission — but because a process this thoughtfully designed deserves scoring that is equally thoughtful.​​​