We Built the PMP Prep Tool We Couldn't Find
We should be upfront: we're biased. We built GanttGrind, so obviously we think it's good.
But here's why we built it. We couldn't find what we were looking for. We paid for the courses. We ground through the question banks. The content wasn't the problem — most PMP prep has plenty of content. What was missing was signal. The ability to look at how you've been studying and get an honest answer to the question every candidate is actually asking: am I ready?
Not "have you done enough questions?" Not "do you feel confident?" Ready. As in: if you walk into the testing center today, do you pass?
That question turns out to be much harder to answer than it looks. We decided to take it seriously.
What GanttGrind actually is
An adaptive PMP exam prep platform. You practice questions, and the platform tracks your performance at the domain, topic, and subtopic level — not just how many you've gotten right, but where you're strong, where you're weak, and how your mastery is changing over time.
The practice engine is adaptive by design. It pulls questions based on where you need them most: 40% from your weak areas, 30% from questions you've missed before, 20% from content you haven't seen yet, 10% review of things you've already locked in. The weighting is deliberate. The goal isn't to expose you to as many questions as possible — it's to make each study session more useful than a random pass through a question bank.
You can also drill any domain on its own, run a timed full exam simulation, or just practice freely if you want exposure without the algorithm deciding for you. The platform remembers where you've been.
We also cover CAPM — the full 1,500-question bank is free, because we think the entry-level certification should be accessible without a credit card. PMP premium is $15, once, for life. No subscription.
The part we're most serious about
Every person who uses GanttGrind and then uploads their score report — pass or not — is contributing to something bigger than their own prep.
We're building a prediction engine. It already exists and it's already learning. The question it's designed to answer is: given your preparation profile right now, what's the probability you pass?
Here's how it works. When someone practices on GanttGrind and then takes the real exam and uploads their score report, they leave behind two things: a detailed map of how they prepared (every question answered, every session, every domain score) and an outcome (pass or not). The model reads these preparation maps and learns which patterns tend to lead to passing.
It works in phases. Early on, with fewer data points, it uses weighted rules built from what good PMP preparation looks like — think of it as a seasoned instructor assessing your study log. As more people contribute their score reports, it transitions to logistic regression trained on real outcomes. At 30 quality profiles, it starts learning from actual data. At 200, it can start detecting specific patterns: not just "are you ready" but "people with your exact domain weakness combination pass at this rate."
One thing we're transparent about: a score report paired with 700 practice questions behind it teaches the model something real. A score report with 50 practice questions behind it teaches it less. Both matter. They're not the same. The model weights them accordingly — a thinner practice profile contributes less, proportional to how reliably it reflects genuine preparation.
What this means practically: if you use GanttGrind seriously and share your outcome, you're making predictions more accurate for every person who comes after you. The more the community contributes, the more specific and honest the predictions get.
What we're not
We don't do video lectures. We don't have a structured curriculum you follow week-by-week. We're not going to teach you project management from scratch.
If you've never managed a project and haven't opened the PMBOK, GanttGrind isn't where you start. A good instructor or the PMI materials are.
What we are is what you use once you've done the learning — to find out what actually stuck, where the gaps are, and whether you're actually ready.
A note on questions
Every question on GanttGrind goes through a six-stage validation pipeline. They're scenario-based, not definition recall — the same format as the real exam. We're obsessive about quality because a bad practice question doesn't just waste your study time. It teaches you the wrong thing.
We cover both active PMP content outlines: ECO 2021 (through June 2026) and ECO 2026 (effective July 2026). The 2026 outline is a significant change — Business Environment nearly triples in weight, from 8% to 26%. If your exam date is July 2026 or later, that matters a lot. We cover both so you're practicing the right content for your timeline.
Who we are
Small team. Bootstrapped. This was built by a PMP who was frustrated with exam prep that couldn't tell you what you actually needed to study.
We're not backed by a publisher. We don't have a roadmap locked in for 18 months. If you find a question error, it gets fixed — not logged and forgotten. If something is missing that would help you study better, email us. We actually build it.
[email protected]. We read everything.