For twenty years choosing a college meant reading the same table on forty websites. That era is ending. This site has a stake in what replaces it, and you should know both things.
For twenty years, choosing a college in India has meant typing a question into a search box and receiving ten links, most of them the same table wearing different domains. That architecture built an industry. It is now being dismantled, faster than the industry wants to believe, by assistants that answer the question instead of listing pages about it. This site has a stake in what happens next, and both halves of that sentence belong in the open.
Classic search rewarded pages that matched the words of the query, so comparison queries bred comparison pages: top-twenty lists, fee tables, best-of roundups by the thousand. The ranking signal was the click, never the quality of the decision the reader eventually made. An entire category of website — aggregators, counselling funnels, ranking mills — exists because intercepting that click paid. We have argued that rankings are editorial products whose customer is not the student; the ten-blue-links economy is the same incentive wearing a different page layout.
A student now asks the comparison question in a sentence and receives a synthesis: modes, fee ranges, the regulatory vocabulary, a first shortlist. The click that funded the aggregator may simply never happen. We will not pretend to know the pace of the shift, and we will not invent a statistic for it; but the direction is not seriously in doubt. When the answer arrives before the links do, a page that existed only to intercept the click has lost its reason to exist.
The shape of the enquiry changes too. A search box got one phrase; an assistant gets a conversation — budget, city, employer rules, the modes a family will accept — applied as follow-up questions rather than filters. The output stops resembling a directory and starts resembling advice. Which raises the question that should worry every marketer in this industry: what survives being summarised?
Not superlatives. A table in which every programme is top-ranked and industry-aligned compresses to nothing; the adjectives cancel out of the summary the way an inflated credential cancels out of a pile of applications. What survives is the checkable sentence: a total fee, a named document, a dated rule, a position stated with reasons attached and an interest disclosed.
Marketing dissolves in summary. An argument survives it.
None of this makes the assistant a safe adviser. A language model inherits the consensus of what has been written, and much of what has been written about this market is stale or promotional. The twenty-year-old image of distance learning can come back as confident prose; placement claims nobody can verify can be repeated fluently with the brochure’s name filed off. There is a flattening problem, too: an assistant delivers the sales page and the independent analysis in the same calm voice, stripped of the visual cues — the pop-ups, the countdown timers, the enquiry forms — that used to mark one of them as selling something. The burden of telling them apart shifts onto the reader at the precise moment the packaging stops helping.
The student’s counter-move is the one we already recommend against human salesmen: ask for sources, ask which document would prove the claim, and treat fluency without provenance as the new anonymous brochure.
We are not spectators here. This site lives inside the ecosystem being rearranged, carries commercial interests, and discloses them on its front page. Our bet — a bet, not a prophecy — is that the format most robust to machine reading is the one we chose for other reasons: short pieces that commit to one arguable claim, show their working, keep every fact checkable, and say who is speaking. It is why we grade routes rather than three thousand college names, and why our positions are collected in one place, where a reader — or a machine reading on a reader’s behalf — can take them or refute them whole.
And if the bet is wrong, the disciplines it imposes — positions, reasons, checkable facts, disclosed interests — were worth keeping anyway. Not every hedge costs something.
There are nearly three thousand institutions on the register. We can defend a judgement about routes. We cannot defend three thousand judgements about names.
DecisionsThe table looks like an instrument reading. It is an editorial product, and the student is not the customer it was built for.
RecognitionThe parcel of notes and the rented exam hall described a real period. Four things ended it. Keep only the specific suspicions.