Audience and Community — The Dossier

Sampling notes

• Universe: 60 videos in data/videos/summary.jsonl. Comment files exist for 52 of them (top 200 comments each).
• Sample drawn per the prescribed policy:
  • All comments from the top quartile of long-form (length_s > 90): [yCR38NmeCpY], [oYTJaHw4na8], [Gs38SCGxPGA], [efBn5kJ0snc], [yLEakbf5gpI], [zv2HCSdsiJw].
  • All comments from the top quartile of shorts (length_s ≤ 90): [rpk9_nXIvdM], [CDdpB7oWm0k], [PGNi1Zw-tA0], [JZAr30UgX_c], [WcIfoJvkHfQ], [wJB_hfKDnFI], [7ZU-23P09ww], [NqpRtHxfML0], [qjm31UWuMJs].
  • 8 random other videos (seeded): [Q4G_3RBO0Nk], [rajRbU2z2mw], [HCSoVCTMz8A], [wPbZ1w6XGpE], [3uk0G_LNd7A], [Flc5mIdH6Ug], [nSUiQE5zxbY], [6Ix6Tt4I47A].
• Raw comments scanned: 1,390. Pre-filter dropped 688 (642 < 8 words / single-line acclamations, 44 emoji-only, 2 @-mention-only, 0 spam links). Net kept: 702 comments, well below the 2,500 cap.
• Three videos in the sample had zero comments survive the 8-word filter (Q4G_3RBO0Nk, HCSoVCTMz8A, wJB_hfKDnFI) — almost all comments on those were one-line “Jai Shri Ram” style acclamations.
• Long-form comment density is far higher than shorts: the six long-form videos alone account for ~554 of the kept 702 comments. Shorts mostly get praise emojis and short slogans, which are exactly what the filter removes — analytically interesting in itself.

Audience segments

Six recurring segments emerged. Sizes are rough share-of-engaged-commenters in the filtered sample.

1. Mobilised Hindu nationalist core (≈45–55%). Largest cluster. Markers: framing every story as Hindu-vs-invader; routine use of “Jai Shri Ram”, “Har Har Mahadev”, “vande mataram”, saffron and trishul emoji, the recurring couplet-style invocation of Shivaji / Bhavani / Hindu Rashtra that appears verbatim on at least three videos [yCR38NmeCpY] [efBn5kJ0snc] [zv2HCSdsiJw]. Engagement skews toward Muslim-encroachment, judiciary, and conversion content. Often calls for capital punishment, “fansi”, direct Yogi-tagged petitions.
2. Bengali Hindu in-state segment (≈8–12%). Concentrated almost entirely on [efBn5kJ0snc]. Self-identifies as “I am from West Bengal” or by neighbourhood (Salt Lake Sector 5, Darjeeling). Uses Bengali script for emotional asides while keeping argumentative content in English. Asks specifically for more Bengali-language podcasts and more interviews with WB-specific politicians. Strong overlap with anti-TMC, anti-infiltration framings. A noticeable sub-thread on this video pushes back on the channel’s BJP-friendly assumptions, arguing central BJP has abandoned WB karyakartas — a rare critical voice from inside the base.
3. Ex-Muslim / counter-Islam interest audience (≈6–10%). Surfaces on [oYTJaHw4na8]. Names a circuit of creators they follow (Sahil, Zafar Heretic, Adam Seeker, Apostate Prophet, Nabi Asli, Saleem Wastik). Some self-identify as recent ex-Muslims. Treats the channel as a gateway to a wider counter-apologetics ecosystem and explicitly thanks the host for booking guests from it. Asks for donations to ex-Muslim creators.
4. Policy-curious / “data person” segment (≈8–10%). Visible on [yLEakbf5gpI] and [zv2HCSdsiJw]. Distinguished by longer, paragraph-form comments engaging with delimitation math, Article 81, fertility rates, judiciary reform mechanics. More English, fewer slogans. Frequently disagrees inside the comment thread (the long delimitation back-and-forth with one repeat commenter is the clearest example). Compares The Dossier favourably to Ajeet Bharti, Dr Ankit Shah, Nitish Rajput — i.e. uses the channel as part of a basket of “right-wing but research-led” creators.
5. Sceptical / cross-pressure commenters (≈5–8%). Heterogeneous: some are Modi-disillusioned right-wingers (“BJP becoming new Khangress”), some are caste-left voices defending Acharya Prashant or attacking the host as “GC-interest” [yLEakbf5gpI], some are TMC/AAP-aligned pushing back on the BJP framing [efBn5kJ0snc] [3uk0G_LNd7A], and a few openly Muslim defenders. They are usually downvoted within-thread but persistently re-engage, which keeps reply-trees alive.
6. Diaspora and senior-tone Hindi commenters (≈5–7%). Markers: dated signoffs with full name, district, state, country (Odisha, Bradford, Karnataka, Rajasthan); reverent address to the host as “bhaiya”/“sir”/“prabhu”; long, formally-structured Devanagari paragraphs. Skews older. Asks for the content to be published as a book or translated to national-TV format [zv2HCSdsiJw].

Language and regional mix

The dominant comment language is Hinglish (Roman-script Hindi mixed with English), followed by Devanagari Hindi at roughly equal volume on long-form videos. English-only comments are a clear minority and cluster on the policy/data segment and the ex-Muslim segment. Bengali script appears almost exclusively on [efBn5kJ0snc], where it carries the most emotionally charged ground-level testimony; the same video also has the highest English share, because WB commenters code-switch to English when arguing with non-Bengalis in the thread. Regional self-identification — when present — leans North/East: West Bengal (Salt Lake, Darjeeling, Kolkata), UP (Yogi-tagged petitions), Bihar, Rajasthan, Karnataka, Maharashtra, Assam, Odisha, and a small diaspora cohort (Bradford, generic US/UK). South Indian non-diaspora voices are sparse; one Karnataka commenter explicitly thanks the host for pushing back on Arya/Dravidian framings [yLEakbf5gpI], suggesting latent appetite there that the audience itself is asking the channel to serve.

Top FAQ themes

Frequencies are counts in the filtered sample; “triggers” lists the videos most responsible.

1. “Why isn’t mainstream media covering this?” — ~22 instances. Triggers: [yCR38NmeCpY], [zv2HCSdsiJw], [3uk0G_LNd7A], [efBn5kJ0snc]. Mostly rhetorical, but the recurrence suggests a CTA opportunity around comparative-coverage explainers. Not directly addressed elsewhere on the channel based on what commenters reference.
2. “What is the actual remedy / what should we do as citizens?” — ~18 instances, including direct requests to write to PMO, file PILs, organise. Triggers: [zv2HCSdsiJw], [yLEakbf5gpI], [3uk0G_LNd7A]. Audience is asking for action-toolkits; the channel has not (per comment references) produced one.
3. “When are you doing the next episode with Ajeet Bharti / on his channel?” — ~10 instances, appearing on at least three unrelated videos [rpk9_nXIvdM], [3uk0G_LNd7A], [6Ix6Tt4I47A]. The host’s tie to Ajeet Bharti is itself a recurring audience question; cross-promo cadence matters to them.
4. “Will BJP actually do anything about X?” (judiciary, infiltration, encroachment, conversions) — ~17 instances. Triggers: [zv2HCSdsiJw], [efBn5kJ0snc], [rpk9_nXIvdM], [3uk0G_LNd7A]. A mature, disillusioned sub-strain of the core audience; not catered to with a direct answer video.
5. “Why don’t you post more often / why long gaps?” — ~9 instances. Triggers: [Gs38SCGxPGA], [3uk0G_LNd7A], [nSUiQE5zxbY], [yLEakbf5gpI]. Notable because it indicates retention and parasocial pull (people miss the host) but also a frustration vector.
6. “Who is the guru/mentor you keep referencing?” — recurring on [nSUiQE5zxbY]. Small frequency (~3) but high signal because the question keeps coming back unanswered.
7. “What about UGC / academia infiltration — when will you cover it?” — appears verbatim across at least three videos [rpk9_nXIvdM], [3uk0G_LNd7A], and another, almost certainly the same commenter copy-pasting. Even if one user, the topic is unaddressed and may be a real gap.
8. “How is the host safe? Be careful.” / “Take security seriously” — ~12 instances, mostly on [zv2HCSdsiJw], [yCR38NmeCpY]. Implicit FAQ: the audience wants reassurance the host hasn’t been silenced or sued.
9. “Where can we donate / how do we financially support you?” — ~6 explicit instances [yLEakbf5gpI], [oYTJaHw4na8], [nSUiQE5zxbY]. Conversion funnel is underbuilt.
10. “Will you publish this as a book / PDF / share the notes?” — ~5 instances, all on [zv2HCSdsiJw]. Audience wants downloadable artefacts of the heavy-research episodes.
11. “Is there a Bengali / Tamil / language version?” — appears on [efBn5kJ0snc] (Bengali) and implicitly on [yLEakbf5gpI] (Karnataka commenter). Low but pointed.
12. “Why does BJP isolate state-level scandals (Saradha, Narada, Sandeshkhali)?” — specific to WB sample [efBn5kJ0snc], asked of the guest; the host did not (per commenters) follow up.
13. “What is the legal / constitutional mechanism for X?” (collegium reform, Article 81, President’s rule) — ~10 instances [zv2HCSdsiJw], [yLEakbf5gpI], [JZAr30UgX_c]. Indicates explainer demand for procedure-level content as standalone short FAQs.
14. “Can you cover RG Kar / CBI inaction / specific local case X?” — recurring micro-requests on [efBn5kJ0snc] and [yCR38NmeCpY].
15. “Where is the host / why the absence?” — ~5 instances on long-form videos after gaps [Gs38SCGxPGA], [nSUiQE5zxbY]. Parasocial signal more than a content question.

Sentiment overview

Filtered comment distribution skews strongly positive toward the channel itself, while skewing strongly negative toward the subjects covered.

• Positive toward host/channel: ~70%. Praise for research depth, clarity, and bravery dominates. Recurring framing words: “brilliant”, “eye-opener”, “kadwa sach”, “research”, “fearless”.
• Mixed: ~15%. Combines praise with feature requests, criticism of editing/thumbnails [3uk0G_LNd7A], or asks the host to let guests speak without interrupting (recurs at least four times across [efBn5kJ0snc]).
• Negative toward channel: ~8%. Two sub-flavours: (a) accusations of being “BJP propaganda” or “paid video”, concentrated on [yLEakbf5gpI] and [3uk0G_LNd7A]; (b) caste-coded pushback accusing the host of GC/savarna interest framing, also [yLEakbf5gpI]. A small number of pro-Islam defenders, mostly on [oYTJaHw4na8] and [yCR38NmeCpY].
• Neutral / procedural: ~7%. Timestamps, “subscribed”, short factual additions.

The negative cluster is drawn most heavily by delimitation / population content (which activates both south-vs-north resentment and caste-left critique) and by electoral-roll content ([3uk0G_LNd7A], which draws TMC and AAP defenders). Operation Sindoor / war content [Gs38SCGxPGA] [6Ix6Tt4I47A] draws almost no negative — the wartime jingoist register suppresses dissent in the comments.

What the audience asks for

Ordered by frequency in the filtered sample.

• Action toolkits: a “what do I do as a citizen” segment or pinned comment on heavy investigations.
• More frequent uploads; visible frustration about week-plus gaps.
• More long-form podcasts with state-level politicians (the WB episode is the model).
• Cross-collaboration cadence with Ajeet Bharti restored / made predictable.
• Downloadable artefacts of the dense research videos (PDF, book form, citations list).
• More content in regional languages, especially Bengali — and Tamil/Kannada implicitly.
• Better editing alignment between thumbnail / B-roll and the actual content [3uk0G_LNd7A].
• Host should let guests finish their sentences — appears as a polite but firm note on the WB and ex-Muslim podcasts.
• A clearer donation / membership path.
• Short-form follow-ups breaking long videos into chapter shorts [zv2HCSdsiJw].
• A standing response to the “how are you safe / take care” thread — a meta-update on legal/safety status would land well.

Parasocial markers

The bond strength is high and unusually first-name-coded for an Indian political channel. The host is addressed as “Kaushlesh bhai” or “Kaushlesh ji” rather than by handle in roughly a third of the filtered comments on long-form videos; “bhaiya” and “sir” are the next most common, and “prabhu” appears in the senior-tone segment. Commenters routinely greet the host’s family (blessings invoked for long life, safety prayers), and after gaps between uploads commenters explicitly ask where the host has been, paraphrased as “we waited for you” or “haven’t seen you in days” [Gs38SCGxPGA] [nSUiQE5zxbY]. Reference to prior episodes is frequent (“after watching your Israel episode” / “your judiciary video was the best ever”) — a strong indicator of repeat viewership rather than algorithmic drive-bys. Commenters quote the host’s catchphrases back at him (the “supreme kotha” coinage from the judiciary episode reappears spontaneously in later comments). There is recurring concern-trolling-as-affection about legal exposure — paraphrased: be careful, contempt notice may come. Several commenters introduce themselves with city/state/diaspora location as if writing to a known person rather than a creator. All of this — first-name use, family-blessing register, episode-recall, safety worry, location-handshake — is textbook parasocial conversion fuel. The conversion ceiling on this channel is being throttled less by audience bond and more by upload cadence and absence of a structured membership ladder.

Five community engagement experiments

1. Pinned “action ladder” comment on every investigation episode. A 4-step ladder (RTI template / PIL precedent / MP-to-email link / share script). Tests: does it reduce the volume of “what should we do?” comments while raising click-through to an off-YouTube destination. Success metric: ≥25% click-through from pinned comment on next 5 investigation episodes; ≥40% drop in unaddressed “what to do” comments.
2. Monthly state-focused podcast slot. Lock one episode/month to a state-politician interview in the WB-episode mould, rotating WB → TN → Kerala → Assam → Karnataka. Tests whether the regional-segment latent demand converts to subscriptions. Success metric: per-state subscriber-add delta vs. baseline episode in same month; ≥1 new identifiable regional segment in comments within 90 days.
3. “Sources drop” companion artefact. For each heavy-research long-form, release a 1-page PDF of citations + a 30-second shorts-cut. Tests whether the senior/data segments will trade an email or donation for the artefact. Success metric: email-list signups per episode; donations attributable to a tracked link.
4. Scheduled cross-promo cadence with Ajeet Bharti. Public commitment to one joint or guest appearance per month, announced in pinned comments and end-cards. Tests whether resolving the recurring “when next with Ajeet?” question moves retention and comment-sentiment toward the channel. Success metric: average view-duration on subsequent solo episodes; reduction in cross-channel FAQ to <2/episode.
5. “You asked, we answered” community-tab Q&A every fortnight. Pull the top 5 unaddressed audience questions from comments (use the FAQ list above as seed), answer in a 2–3 minute community-tab video or post. Tests parasocial conversion: do replies skew toward sustained name-recall, do subscribers from comment threads become repeat commenters. Success metric: % of Q&A askers who comment on a follow-up regular episode within 30 days; community-tab engagement rate vs. channel baseline.