How to Set a Realistic Target Percentile From Mock Scores

PRODUCTIVITY

A candidate takes their first mock, scores badly, and does one of two things. They set a target of 99 because that is what the plan was, or they quietly decide they are not a 99 person and adjust downwards. Both decisions were made with one data point, and neither of them is a target. They are moods.

A target that does any work has to be built rather than chosen, and it has to be built on the right quantity, which is usually not the one candidates reach for. This piece explains why a first mock is close to useless for this, what a mock percentile actually is, and a sequence for arriving at a number you can plan against and revise honestly.

One Mock Is Not a Baseline

Start with what a first mock actually measures, because candidates consistently over-read it.

A first mock measures your current ability, plus your unfamiliarity with the interface, plus your lack of sectional pacing, plus whatever the paper's difficulty happened to be, plus how that particular Sunday went. Only the first of those is the thing you wanted to measure, and the rest of them are large.

The unfamiliarity component alone is substantial and it disappears within a few attempts without any improvement in ability at all. So a first-mock score will usually rise on its own, which means a target built on it is built on a number that was never stable.

A baseline needs three or four mocks under proper conditions: one sitting, sectional limits enforced, no pauses. What you are looking for is not the average of those but the shape, meaning where the scores cluster and how much they swing.

What a Mock Percentile Actually Is

This is the correction that changes most people's planning, and it is not a technicality.

A mock percentile is your position within that mock's test-taking cohort. Those cohorts are self-selected and small. The people who sit a given mock series are not the CAT field, and depending on the series they can be considerably more or less prepared than it.

So a mock percentile is useful as a relative signal across your own mocks from the same series, and it is not a prediction of a CAT percentile. Setting a CAT target by reading a mock percentile as though the two were the same quantity is the single most common error in this whole exercise.

The same caution applies to raw score. There is no fixed mapping between marks and percentile: raw score comes from your correct and incorrect responses under the marking scheme, while percentile is your position relative to everyone who sat the exam, and it shifts with paper difficulty and normalisation across the three slots.

A Sequence That Produces a Real Target

Let the School List Set the Number

Most candidates pick a percentile and then find schools to match it. That is backwards, and it produces targets that are round numbers rather than requirements.

Work the other way. Decide which schools you would genuinely accept, then read their published criteria for your cycle. Where a school publishes a qualifying minimum, that is a floor for consideration rather than a call-earning score, so the number you plan against has to sit meaningfully above it. Where a school publishes a weighting, that tells you how much of the decision the exam carries at all.

Two real examples show how differently that lands. IIFT's 2026 to 2028 shortlist weighted the entrance score at 90 percent, work experience at 5 and gender diversity at 5, which is a process almost entirely decided by the exam. IIT Bombay's SJMSOM used 75 percent entrance score, 20 percent academics and 5 percent work experience for the same batch, where an academic record carries real weight. Both need re-verifying per cycle, and a candidate targeting the second has a different exam requirement from one targeting the first.

Target What You Control

An outcome target is a wish. A process target is an instruction, and the difference shows within weeks.

"Score 99" tells you nothing about Monday morning. "Raise QA accuracy from its current level while keeping attempts steady" tells you exactly what to practise and exactly how to check. Similarly, "read every set in the DILR section before attempting one" is a target you can meet or miss on a specific Sunday, and it moves the score as a side effect.

Outcome target	The process target underneath it	How to check
Higher QA score	Accuracy up at steady attempts	Section accuracy across four mocks
Higher VARC score	More passages read, selection made on the first paragraph	Passages attempted and accuracy within them
Higher DILR score	Every set swept before attempting any	Did you sweep, yes or no, per mock
Fewer unseen questions	A reading budget per question with a default skip	Count of questions never read
Fewer careless losses	One defined check per question type	Wrong-and-fast count in review

Revising It Without Lying to Yourself

A target that never changes is not honest and a target that changes every Sunday is not a target. The resolution is to fix in advance when it gets reviewed.

Review after a set number of mocks, not after a particular result. Four is a reasonable interval. At that review you are looking at the trend and the section-wise variables, and asking whether the process targets were met rather than whether the score moved, because the score is noisier than the process is.

Raise the target when the process targets are being met comfortably and the trend supports it. Lower it when several review cycles have passed with the process targets met and the outcome still well short, which is genuine information rather than discouragement. Never revise on a single mock in either direction.

How Much Time Is Left Changes the Answer

A target is a statement about a date as much as about a score, and the same baseline supports different targets depending on when you are reading this.

With several months to CAT on Sunday 29 November 2026, a large gap between baseline and target is reasonable, because there is room for the concentrated work that closes it. With a few weeks left, the same gap is not a target but a hope, and planning around it usually means neglecting the things that could still move, which are selection, pacing and accuracy on what you already know.

Late in the cycle the honest move is to keep the ambition and change the mechanism. Stop trying to add new capability and start converting the capability you have into marks, because that is the lever that still responds.

When the Process Targets Are Met and the Gap Persists

There is one case worth handling explicitly, because it is where candidates lose the most time to guessing.

Suppose several review cycles pass, the process targets are being met, and the outcome has not moved the way it should. The instinct is to work harder on the same things, and that is usually wrong, because a process target being met while the outcome stalls means the process target was aimed at the wrong variable.

Typically one of three things is happening. The accuracy gain is real but confined to question types that are rare in the paper. The section that is limiting the score is not the one you have been working on, and the mock report would have said so. Or the attempts are up and the selection is not, so you are reaching more questions and choosing among them no better than before.

That is a diagnosis problem rather than an effort problem, and it is the kind of thing an outside read of the data settles faster than another month of work. A CAT preparation strategy review exists for exactly that case: someone reading your section-wise pattern and telling you which variable is actually binding.

The Underclaiming Failure

Most writing on this warns against targets that are too high. The opposite failure is quieter and does real damage.

A candidate who sets a conservative target after a disappointing start prepares for that target, chooses a school list for it, and never finds out what they could have done. The ceiling was set by a mock they took before they knew how the interface worked.

Baselines move a long way in a few months, particularly for candidates who have never practised under sectional timing before. Set the target from the list you actually want, keep it under honest review, and let evidence lower it if it has to rather than lowering it pre-emptively out of a bad first Sunday.

The Summary

A first mock cannot support a target, because it measures unfamiliarity and paper difficulty and that particular Sunday alongside your ability. Take three or four conditioned mocks and read the shape rather than the average.

Read the right quantity while you are at it. A mock percentile is your position within that mock's self-selected cohort rather than a CAT percentile, and there is no fixed mapping between marks and percentile in any case. The trend and the section-wise accuracy are the useful parts of the report.

Then build the target from your school list rather than picking a round number, set process targets underneath it that you can actually act on, and fix in advance when it gets reviewed. Revise on evidence across four mocks, never on a single bad one, and be as suspicious of lowering it too early as of setting it too high.

If your target is a round number you chose because it sounded right, that is worth replacing with one derived from the list you actually want. The CAT target percentile tool fixes the number, and a personalised CAT preparation plan turns it into the weekly process targets that actually move it.