The cost of mock interviews shouldn't be the barrier that keeps you from sharpening your FAANG data skills, and it doesn't have to be. Prepfully and similar platforms charge a premium for what is essentially structured practice with a side of feedback. That's valuable, but it's not the only path. The person asking this question is sensing something important: the price tag doesn't guarantee the practice sticks. What actually sharpens your skills is repetition, honest feedback, and the ability to iterate on your weak spots without wincing at your bank account.
So what does this mean for you in practical terms? It means the alternative isn't a single platform. It's a combination of free and low-cost resources that already exist in your network and your own routine. Peer mock interviews, for instance, are a legitimate starting point. Find a study group, a colleague, or a subreddit dedicated to data science interviews. Swap roles: you interview them, they interview you. The act of asking questions is just as instructive as answering them, because you learn what a strong response sounds like from the other side. Recording yourself is another low-cost move. It's uncomfortable, but it works. You'll hear your own verbal tics, your pauses, and your over-explanation, and you'll fix them faster than any paid session might push you to.
The deeper issue here isn't the cost. It's the assumption that a paid platform is the only structured way to prepare. That assumption undersells the work you're already capable of doing. FAANG data interviews are less about knowing every algorithm cold and more about demonstrating a clear, logical approach under pressure. You can practice that with a timer, a whiteboard, and a friend who's willing to give blunt feedback. You can also use public problem sets from past interviews, which are widely shared and free. The structure you pay for is real, but it's not magic. It's just a schedule and a rubric, and you can build your own version of that with a little discipline.
Here's the concrete point: before you spend another dollar on a mock interview service, set a two-week experiment. Find two partners, schedule three sessions, and record yourself on one question each time. Review the recordings. Note where you stalled, where you rushed, and where you explained your reasoning clearly. Then adjust and do it again. If that process doesn't feel like enough, then yes, pay for a single session with a vetted interviewer to calibrate. But start with the cheap reps. Your skills will improve because you practiced, not because you paid. That's the edge that sticks.