Why didn't I get IIM admission with 97 percentile CAT score in EWS category despite good interview performance?

Getting a 97 percentile in CAT is a genuine achievement — but as we've seen repeatedly across student conversations on this topic, it doesn't automatically translate into an IIM conversion, especially within the EWS category. The frustration is real, and it's worth unpacking exactly why.

From profiles we've tracked that mirror this situation, the patterns are consistent. One candidate with a 97.61 percentile, 9/9/9 academics, and GNEM category reported zero conversions — with their part-time work experience not being counted toward the profile evaluation. Another with 98.74 percentile and EWS category faced the same outcome. Yet a candidate with a comparatively lower 96.5 percentile but a similar academic background did convert tier-2 IIMs — Trichy, Sirmaur, Sambalpur, Nagpur, Bodhgaya, and Jammu — finishing with a 40 CAT-PI composite score. The difference wasn't the percentile. It was everything around it.

Here's what the picture that emerges consistently points to:

Class XII marks carry more weight than most candidates expect. A 73% in Class XII is a meaningful flag in IIM shortlisting models. Peers who converted — even with lower CAT percentiles — frequently had 75% or above. This isn't a minor rounding issue; IIMs use Academic Diversity Scores that penalise sub-75% board marks in ways that can't easily be offset by CAT performance alone.

Gender and academic diversity actively shape the final merit list. IIMs are balancing cohort composition — engineering male candidates face the most competitive environment regardless of category. Even within EWS, the diversity adjustment can shift where a candidate lands on the final list.

Interview perception and score normalisation are not fully in your control. In our experience, candidates often leave interviews feeling they performed well — and many genuinely do. But interview scores are normalised across panels, and a strong personal performance can get moderated downward depending on panel calibration. As we've seen experienced professionals note: even if your individual interviewer scored you well, it may get normalised lower in the final composite.

Work experience matters significantly at tier-1 IIMs, particularly for freshers. For candidates applying straight out of college or with experience that doesn't meet counting criteria, the WAT-PI composite tends to disadvantage them relative to candidates with two or more years of full-time, countable work.

EWS seats are reserved but fewer in absolute numbers. This creates a counterintuitive situation — the reservation exists, but competition within the category can be sharper than it appears from the outside, particularly at older IIMs where total intake is smaller.

So what are the realistic paths forward? A few perspectives worth weighing:

Some experienced professionals suggest considering IIT MBA programmes — IIT Bombay, Delhi, Madras — which offer strong placements and lower fees than newer IIMs, and where the profile evaluation can be more holistic. For candidates with a 97-98 percentile profile being blocked by academics or category dynamics at IIMs, this is a route worth taking seriously rather than as a fallback.

If the goal remains the older IIMs specifically, the honest answer from what we've tracked is that 99.5 percentile and above meaningfully improves tier-1 IIM conversion probability — it gives enough buffer for academics, category adjustments, and interview normalisation to still land you in range. At 97-98 percentile in EWS with a Class XII score under 75%, the margins are simply too tight across too many variables.

CAT percentile is the entry ticket. The final seat is decided by academics profile, interview composite after normalisation, gender and academic diversity weightage, work experience standing, and seat availability within your category — all simultaneously. A 97 percentile is not a weak score. It's just not sufficient on its own when several other parameters are working against conversion.