If your dashboard stops at something like "55% in Fixed Income", it's telling you how you've done on those questions but not why. A topic score can't separate a part of the topic you haven't studied yet from an error you keep making, and that's how "study your weak areas" turns into hours of unfocused work.

When I was a candidate, it frustrated me that every platform could tell me I was at 55% in a topic and none of them could tell me what to do about it on a Tuesday night with two hours to study.

Four levels of detail

A dashboard can give you roughly four levels of detail.

A single overall score ("you're at 68%") doesn't help you study. It tells you how things are going, not what to do next.

Topic accuracy ("55% in Fixed Income, 80% in Ethics") is better, since it points you at an area. But a topic is big, and you're still guessing where in it the problem is.

Learning Outcome accuracy ("within Fixed Income, you're strong on duration and credit and weak on yield measures") narrows it down a lot.

The likely error behind your misses ("your yield-measure misses keep landing on the choice built for using current yield where the question wants yield to maturity") gives you a specific guess about what to fix. You still have to check it by redoing a couple of those misses, but if it holds up, the fix is small.

To see where your dashboard stops, open a topic you're missing and check which of these levels it shows. The last level gives you the narrowest place to start, as long as you treat a tag as a possible explanation to check against your working.

A worked example

Say your dashboard shows 55% in Fixed Income.

With only topic accuracy, you might decide you're bad at Fixed Income and block off weeks to reread the whole topic. In this example, most of that time goes to duration and credit, which you were already fine on.

With the mistake-level view, you see your Fixed Income misses aren't spread across the topic at all. They cluster in two places: your wrong answers keep landing on choices built around mixing up clean and dirty price, and around annualizing a bond-equivalent yield incorrectly. Redo a few of those misses and check whether that's what you actually did, then try a couple of fresh questions on each convention. If the errors hold up, you've got two specific conventions to fix instead of a whole topic to reread, and your later practice will show whether the fix stuck.

The CFA Institute Learning Ecosystem that comes with your registration has topic-level assessments and progress tracking, and you should use it. What's harder to build is a view that groups your errors by the misunderstanding each wrong choice was written around, so you can see a pattern across questions. It takes work, since the questions have to be tagged at the concept level and the platform has to group your errors by those tags.

Two cautions

Small samples mislead. Going 0 for 3 on a Learning Outcome doesn't prove you're weak on it. It could be three hard questions or a bad night. Wait until you have enough answers before you treat a number as real. Even then, more answers won't tell you why you're missing. Check your working on a few misses, try a fresh question, and sort out what you haven't studied yet from errors you keep making on material you have.

Don't chase the percentage. The goal is to fix the misunderstanding, not to grind a topic until the number goes up. Once you think you've fixed the error, check it on questions you haven't seen before. If you just re-drill for the number, you're back to volume for its own sake.

FreeFellow is a CFA Institute Prep Provider, independent of CFA Institute, and does not promise a passing score. On FreeFellow, the readiness score is free. Fellow adds analytics by topic, Learning Outcome and difficulty, and Fellow Plus adds Common traps, which shows the misunderstandings your wrong answers keep matching and lets you drill them. You can start with the free CFA Level I practice questions.