A study planner that reads your question bank performance and produces a week-by-week schedule is genuinely useful, and it solves a real problem: most candidates are poor at allocating their own time, and default to whatever feels productive rather than whatever is needed. But it introduces a subtler failure. Handed a plan that looks authoritative, candidates either follow it slavishly through circumstances it never anticipated, or abandon it entirely the first week it becomes unrealistic. Both waste it. The plan is a hypothesis about how to spend your hours, and hypotheses are meant to be tested and revised.
Key takeaways
- A planner knows your performance data and nothing about your life, so its inputs must be honest.
- Give it your real available hours, not your aspirational ones, or every week will start in deficit.
- Protect mandatory curriculum coverage explicitly, or performance-driven allocation will neglect what you have never attempted.
- Build in catch-up days, because clinical work will take some of your study days and you know it.
- Review the plan yourself every week, and override it when your judgement and its arithmetic disagree.
What a planner can and cannot know
Be clear about the boundary, because everything else follows from it.
A planner can see what you got wrong, where your accuracy is low, which topics you have neglected, how quickly you are working, and how far away the exam is. From that it can do genuinely useful arithmetic: how many questions per day, weighted towards which domains, with what proportion of revision to new material.
It cannot see that you are on nights for a fortnight in the middle. It cannot see that you are strong in cardiology on paper because you spent six months in cardiology, and that your apparent weakness in psychiatry is because you have never worked in psychiatry and never will until the exam. It does not know that your daughter's birthday falls in week five, or that you cannot concentrate after a run of long days, or that the topic it has scheduled for Tuesday is one you find so demoralising that you will find a reason not to open the app.
The plan is arithmetic performed on data. Your life is not in the data.
Give it honest inputs
Most planning failures are input failures, and the commonest is optimism about time.
Candidates enter the hours they wish they had. They plan three hours a day, every day, including the days they are on call, and by the end of the first week they are behind, and by the end of the second they have stopped looking at the plan because looking at it makes them feel worse. A plan you have abandoned is worth nothing, however sound it was.
So enter your realistic capacity. Not your best day, your typical day. Fewer hours, honestly declared, produce a plan you will actually complete, and completing a modest plan beats abandoning an ambitious one by an enormous margin.
Enter the days you genuinely cannot study, and enter them before you start rather than discovering them as failures later.
Protect coverage explicitly
Here is the specific way performance-driven planning goes wrong, and it is worth guarding against deliberately.
A planner allocates time towards your weak areas, which sounds obviously correct and mostly is. But "weak" is inferred from your performance, and you have no performance in the domains you have never attempted. A topic with five questions attempted and four correct does not look weak. It looks fine, and the planner will leave it alone.
The result is a plan that carefully optimises the parts of the syllabus you have engaged with while ignoring the parts you have not, which is precisely the wrong emphasis. So before you let performance drive allocation, ensure the whole blueprint has been sampled, and instruct the plan to protect coverage of the domains that are still unknown. We set out that audit in using Standard mode as a syllabus audit.
Build in the slack you know you will need
Every plan that assumes nothing will go wrong is a plan that will be broken in the second week, and a broken plan tends to be an abandoned one.
So build the slack in from the start. A catch-up day each week, deliberately empty, is not laziness, it is the mechanism by which the plan survives contact with a busy job. If you do not need it, you get a bonus session. If you do, you absorb the disruption without falling behind, and, crucially, without the feeling of failure that makes people stop.
The same applies to the final week. Leave it lighter than you think you need, because you will want the room, and because cramming new material in the last few days reliably produces anxiety and very little retention.
Review it weekly, and be prepared to overrule it
The plan should be revisited every week, with your own judgement in the loop, and the review should ask three questions.
Did the last week actually happen? If not, why not, and is the reason structural, in which case the plan needs to change, or a one-off, in which case the catch-up day absorbs it.
Has the data moved? A domain you have now corrected should stop absorbing hours. A weakness that has emerged should start.
Does the plan still make sense to you? This is the important one. If the plan says spend four hours on a topic you know cold, and its reason is that your accuracy there dipped because of a bad session when you were exhausted, then the plan is wrong and you are right. Override it. If the plan says spend four hours on a topic you find unpleasant and have been quietly avoiding, and your instinct is to override it, then the plan is right and you are wrong.
Telling those two situations apart is a skill, and it is one of the few parts of this process that cannot be automated.
The failure mode in both directions
Notice that there are two ways to misuse a planner, and they are opposite.
Following it blindly means grinding through a schedule that no longer reflects reality, because the app said so, and losing the judgement that told you the schedule was wrong.
Abandoning it entirely means returning to the default state that made you want a planner in the first place: revising what feels good, neglecting what does not, and discovering in the exam that your instincts about your own weaknesses were poorly calibrated.
The productive position is in the middle. Let it do the arithmetic, which it does better than you. Keep the judgement, which you do better than it.
Where iatroX fits
iatroX's planning works from your actual performance data rather than a generic schedule, allocating practice towards the domains and concepts where you are genuinely weak, with adaptive targeting and spaced repetition doing the day-to-day scheduling of what returns and when. Because coverage is tracked separately from accuracy, the domains you have never sampled remain visible rather than being quietly treated as strengths. And because missed questions can be opened in the Socratic Tutor, a correction is understood rather than merely rescheduled. Keep the weekly review in your own hands, and try it with free sample questions at iatroX. For the plan that ties this together over a defined run-in, see building a six-week plan from your data.
Frequently asked questions
Should I follow an AI study plan exactly? No. Treat it as a hypothesis based on your performance data. It cannot see your rota, your energy, or your life, so review it weekly and override it when your judgement and its arithmetic genuinely disagree.
Why does my study plan ignore topics I have never studied? Because it infers weakness from performance, and you have no performance in domains you never attempted. A barely-sampled topic looks fine. Ensure the whole blueprint is covered before letting performance drive the allocation.
How many hours should I tell a planner I have? Your realistic typical capacity, not your best day. An optimistic input produces a plan you fall behind on in week one and abandon by week three, and an abandoned plan is worth nothing.
Should I build rest days into my revision plan? Yes, and a catch-up day each week specifically. Clinical work will take some of your study days, and a plan with no slack breaks on first contact with reality, which is when candidates give up on it entirely.