Traffic growth plan: Prajavani and Deccan Herald
Prepared: 19 Nov 2026 by the Product, SEO and Data team · Sites: Prajavani (PV, prajavani.net) and Deccan Herald (DH, deccanherald.com)
The goal
Grow Google traffic on both sites and make it less volatile. The plan has two parts:
- Fix Current Issues: fix the 43 issues we found that hold back the traffic we already have. → All issues, with status, evidence, impact, owner and fix
- New verticals: add pages and content that answer searches people already make. → New verticals, with Search Console and GA4 evidence
We measure both against a baseline from before Deepavali. → How we'll know it worked
Why now: the November dip
From 6 Nov 2026, PV's Google web clicks halved and DH's fell 18%. Neither was a Google penalty or a site fault, but the causes differ.
PV: Deepavali. Festival-week demand for the PV cartoon (and its satta-matka searches) fell and came back.
- PV's positions held (1.0 on brand queries), and direct visits to the cartoon fell as much as search visits, which no Google change could cause.
- PV recovered to 134,399 clicks on 16 Nov 2026, 94% of the pre-festival level.
DH: one-off traffic ending, and piracy tags slipping. Not Deepavali. DH readers did not go away: in GA4, DH organic sessions (including Discover) rose 21% and direct sessions rose in the festival week, and the "deccan herald" brand query held at position 1.
- The fightkaro satire story faded (6,858 → 598 clicks a day) and US midterm stories ended (world section 3,942 → 1,093). Without these two spikes, DH web is roughly at its normal level.
- The piracy-name tags lost rankings: all tag pages 11,047 → 7,192 clicks a day;
/tags/tamilrockers2,859 → 215, position 7.2 → 16.8. - DH web was 36,746 clicks on 16 Nov 2026 (50,351 a day before). It will not come back by itself.
Still open: PV's Google News clicks fell from about 450 to about 45 a day on 12 Nov 2026 and have not recovered. This is under investigation (P7 item 1).
The lesson: where Google traffic comes from today
Measured in a normal period before the festival: 6 Oct – 2 Nov 2026 (28 days), Search Console clicks by page group. Search Console counts clicks (visits from Google), not users. GA4 can't be used for this split because it records a large share of sessions with landing page "(not set)".
PV: the cartoon is 84% of Google web search.
Share of PV Google web search clicks, 28 days, 6 Oct – 2 Nov 2026:volatiledurable
- Cartoon and Churumuri pages: 3.86M web search clicks, about 138K a day. That is 84% of PV's web search clicks and 34% of all PV Google clicks (web, Discover and News).
- All other PV articles get only about 12K web search clicks a day (7%). This is the cost of issues #1 and #2: Google has the headline but not the body text of most PV stories.
- PV articles do get traffic, but from Discover (7.8M clicks in 28 days). Discover is 65% of all PV Google clicks, and it is volatile by nature.
DH: piracy-name tags are 28% of Google web search.
Share of DH Google web search clicks, 28 days, 6 Oct – 2 Nov 2026:volatiledurable
- Piracy-name tag pages: 396K web search clicks, about 14K a day. That is 28% of DH's web search clicks, and 4% of all DH Google clicks.
- DH articles get about 24K web search clicks a day (47%).
- Discover is 80% of all DH Google clicks (7.3M in 28 days). DH Discover has already fallen once, from 16.5M clicks in Aug 2025 to 3.8M in Nov 2025.
- The fightkaro story started after this period, so it doesn't show here.
What this means: on both sites, the traffic that holds steady is a small part of the total. Web search for articles, listings and utility pages is held back by technical problems. The plan fixes those problems (part 1) and adds durable pages (part 2), so that a festival week, a fading fad or a Google update to risky pages no longer decides the day.
Source: Search Console page-level clicks. They total 13.2M (PV) and 9.2M (DH) for the period, against property totals of 12.6M and 9.3M.
Part 1: Fix Current Issues
Every issue we found, most important first. Each row gives the problem in plain words and where its fix stands, the key evidence (full evidence behind the link), what we gain when it is fixed, who owns it, and the exact item in the fix package the tech team works from.
41 items (43 issues; #1 and #2, #5 and #15 shown together): #1 and the issues that wait on it first, then the rest by importance. Done 1 · Partly live 3 · Fix built 1 · Waiting 3 · Not started 26 · Investigating 2 · Decision needed 4 · Ongoing 1
Priority: Critical High Medium Low Note
| # | Problem and current status | Evidence | Impact for us if fixed | Owner | Fix package (for tech) |
|---|---|---|---|---|---|
| Fix #1 (article text in the page Google reads) and the issues that wait on it | |||||
| 1 + 2 | CriticalGoogle can't read our article text, so most PV articles are in Google without it Partly live Live on DH (19 Nov 2026). PV fix built, not released; after release Google has to re-read PV's articles. When Google first loads an article, the paragraphs are empty. The text only appears after the page's code runs in a browser. Google does that late or not at all for PV, so it has saved most PV articles with the headline only, and they can only be found by headline words. Older stories never caught up. | Same code on both sites, very different result in Google (tests of 16 Nov 2026): PV, badly affected: only about 1 in 4 articles tested had their text in Google. 482 of 634 (76%) had none (site: search test of 663 stories). DH, mostly unaffected: 14 of 15 stories that first looked "missing" were found on re-test, and 8 of 8 never-shown stories had their text in Google. Estimate: 0–7% of DH stories. Why the difference: Google renders DH far more often, with about 7× more Googlebot visits per new story. Status: DH fix live 19 Nov 2026; PV still to ship. | PV articles start appearing for the many searches that match their body text, not just the headline. This is the biggest single lever for non-branded article traffic: today PV articles get only about 12K Google web search clicks a day. | Tech (PV release); SEO team re-tests after release | P1 item 1 P1 item 8 |
| 3 | HighGoogle visits new PV articles too rarely Waiting Pending: re-analyse after #1 + #2 ship on PV. If Google then has PV's article text, low crawl matters much less; what remains is decided with the AMP decision (#26). Google visits PV far less than DH for each new article, and spends much of what it does visit on duplicate copies (AMP pages, redirects). This is the likely reason Google skips reading PV's article text (#1 + #2), so it is re-assessed once that fix is live. | Crawl per new article: PV about 360 Googlebot requests per new URL, DH about 2,700 (Search Console crawl stats, 25 Jul – 21 Oct 2026). Wasted on copies: 46% of Google's successful PV fetches are AMP copies (sample of 993 URLs, 12 Nov 2026). Slower answers: PV responds to Google in about 230 ms, DH about 150 ms. Already improved: blocking the login check (#4, 25 Oct 2026) freed about 18K requests a day; crawl of real PV pages rose about 20%. | New stories found and refreshed sooner. Measured after #1 + #2: Google's discovery crawl and time from publish to first crawl for new PV articles. | Tech; SEO team (crawl stats) | P3 |
| 26 | HighAMP copies use half of Google's visits for little traffic Waiting Waits on #1: DH decision from about 17 Dec 2026 (#1 live there 19 Nov), PV about 4 weeks after PV's #1 release. Until then the PV AMP copy is the only one Google reads with full text. Every story has a second AMP copy that Google keeps visiting. | Crawl cost: AMP copies take 45–50% of Google's page visits on both sites. Traffic, last 28 days (20 Oct – 16 Nov 2026): clicks Google served as AMP: PV 2.3% of Google clicks (7.2% of web search, 0.2% of Discover); DH 6.2% (33.2% of web, 1.9% of Discover). Trend: Discover AMP is falling on both sites since September. Health: all sampled AMP pages valid; Google's AMP check passes on PV 68 of 68 and DH 59 of 62 (17 Nov 2026). | If AMP is retired: Google spends its visits on real pages and new stories, and one copy per story is simpler to run. | Leadership decision after #1; Tech | P3 |
| 31 | HighMost PV articles are print copies that Google never shows Waiting Waits on #1: measured about 4 weeks after PV's #1 release against Sep–Nov 2026. No other change to print copies (including publishing times) until then. 79% of new PV articles are copies of print stories; most of these get no Google views at all. | Output: 79% of PV's new articles are print copies. Never shown: print copies are 96% of the 30% of PV articles Google never shows; all are indexed, they just don't rank. Source: traffic funnel, Sep–Nov 2026. | Tells us whether #1 fixes this or whether print copies need different handling. | SEO team (measure); PV editorial | P1 item 8 |
| Other issues, most important first | |||||
| 14 | HighGoogle wastes visits on redirects and server errors Partly live DH AMP link fixed 11 Nov 2026. PV fix written (branch amp-issue1, 26 Oct) but not merged or live: checked 19 Nov 2026. DH server errors being investigated. Google was sent through redirects to reach AMP pages, and DH sometimes returns server errors. | Redirects: 23% of Google's visits to DH hit redirects, caused by the AMP link; fixed 11 Nov 2026. PV: the same encoded AMP link, but it doesn't redirect (still unchanged on 19 Nov 2026). Errors: DH server errors and a host-status warning are still open; 617 server-error pages left from the Aug AMP outage. Source: Search Console crawl stats and Page indexing, Nov 2026. | More of Google's visits reach real stories, so new content is found faster. | Tech (PV amphtml); Quintype (DH 5xx) | P3 item 2 P3 item 3 P3 item 4 |
| 19 | HighListing pages exist at several addresses (PV: big; DH: small) Not started Fix proposed, not started. Full redirect list ready (19 Nov 2026); DH already has canonicals, needs them pointed at section pages. Every section and many landing pages also open at a short address like prajavani.net/tumakuru or /cartoons. On PV these copies have no canonical, so Google treats each as a separate page and splits rankings. DH's copies point Google to the wrong main page. | PV: 450 short addresses should redirect to their section page; 247 are known to Google, with 48.7M clicks in 16 months. 6 of 6 checked live have no canonical. Biggest: /cartoons (44.9M clicks) and /churumuri-op-ed (3.4M) need a careful separate step. The other 448 carry about 299K clicks.Keep: 50 hand-made PV pages such as /horoscope and /news-landing (944K clicks) need their own canonical.DH, low impact: 240 short addresses, 49 known to Google, about 38K clicks; canonicals point to themselves or another short address. | Ranking signals for each topic add up on one URL instead of being split, protecting and growing PV's biggest listing pages. DH gains little. | Tech; SEO team (URL list) | P4 item 4 P4 item 5 |
| 27 | HighHoroscope pages are split up and partly invisible to Google Partly live DH horoscope text now in the page Google reads (19 Nov 2026, with #1). Hub, headings, titles and PV duplicate hubs not started. PV's horoscope hub is spread over several URLs; DH's daily horoscope text only appears after the page's code runs. | PV hubs, web clicks in 28 days (20 Oct – 16 Nov 2026): /horoscope 33,043 (main hub) · /weekly-horoscope 3,879 · /daily-horoscope 2,118 · /astro-vastu/horoscope 508 · /astro-vastu/astrology 499 · /monthly-horoscope 369 · tag pages /tags/horoscope, /tags/horoscope-weekly, /tags/horoscope-daily … 509 together.PV searches: brand searches ("prajavani dina bhavishya", 57.5K impressions) show /horoscope, /weekly-horoscope and /daily-horoscope at once; generic searches ("ಭವಿಷ್ಯ" 75K, "dina bhavishya" 43K, "rashi bhavishya kannada" 14K impressions) sit at positions 5–8.PV totals: 51,952 search + 55,358 Discover clicks; daily articles also filed under /astro-vastu/vaastu/.DH hubs, web clicks in 28 days: /horoscope 21,765 (main hub) · /astrology 868 · /astrology/weekly-horoscope 37 · /amp/horoscope (indexed on its own) · tag pages /tags/horoscope-weekly 1,399, /tags/horoscope 622, /tags/weekly-horoscope 430, 12 sign tags 5–219 each.DH searches: "horoscope today" (28.8K impressions): hub at position 18 while that day's article is at 2–5, because the hub shows Google no horoscope text. DH totals: 34,906 search + 13,082 Discover clicks. Page check (19 Nov 2026): DH daily text now in the page Google reads; no main headings on either site; one DH article with no description. | Horoscope, a steady daily-demand page, ranks better on both sites. It is also the base for the horoscope vertical. | Tech; PV and DH editorial (horoscope copy) | P4 item 8 |
| 32 | HighGoogle can't follow links to older stories Not started Fix proposed, not started. Next-page links on listings aren't real links, and articles carry few links to other stories in the first HTML. | Articles: 25 of 30 sampled articles on each site link to no other story in the first HTML Google gets (only footer links); print and wire copies had none; no site shows an automatic related-stories block. Only editor-added "Also read" cards add links. Pagination: next-page links have no address on either site (PV 0 of 59); PV ?page=2 loads page 1.DH section pages: 8 of 8 sampled send no story list until JavaScript runs. Source: live server-HTML checks, 16 and 19 Nov 2026 (30 articles and 10 listings per site). | Google finds and keeps older stories, which brings long-tail search traffic. | Tech | P4 item 1 P1 item 2 P4 item 3 |
| 42 | HighNovember traffic dip, and PV Google News collapse Investigating Web dip explained (PV: Deepavali, recovered; DH: one-off traffic ended). PV Google News under investigation. Web search clicks fell on both sites from 6 Nov. PV's Google News traffic fell to near zero from 12 Nov and hasn't recovered. | PV web dip from 6 Nov 2026: Deepavali; positions unchanged and indexing flat; back to 134,399 clicks on 16 Nov. DH −18%: one-off traffic ended and piracy tags lost rankings (GA4 check, 19 Nov 2026). PV Google News: about 450 → 45 clicks a day from 12 Nov 2026, still open. Source: Search Console, Nov 2026. | About 400 Google News clicks a day back on PV if a setting caused the drop. | Quintype + PV editorial (Google News); SEO team (tracking) | P7 item 1 |
| 43 | HighOld DH story addresses still work, and Google prefers them Not started Fix proposed, not started. Stories moved from the old DH site still open at their old addresses, so Google often treats the old address as the main one. | What we saw: migrated DH stories still open at old addresses ( /content/<id>/<slug>.html) with their own canonical, instead of redirecting.In Google: "Google chose different canonical" on 80,379 pages; old URLs got 7,199 clicks in 4 weeks. PV: same code, 1,240 pages. Source: Search Console, 13 Nov 2026. | Each story has one address, so its ranking signals stop being split between two URLs. | Tech | P2 item 1 |
| 5 + 15 | MediumPV pages jump while loading on phones Not started Fix proposed, not started. The Taboola block below the article loads late and pushes the page around. | What fails: 67.7K PV pages fail Google's mobile layout-stability test (Core Web Vitals, CLS). Cause: the Taboola feed below the article has no reserved space. Only while scrolling: in our tests the page jumps above the limit only after scrolling past the article (58 of 59 pages were fine on load alone). Includes #15: a weather forecast article ( …-4323207, 0.54) and a feature article (…-4376516, 0.33) where readers see far more jumping than our load-only tests; both fit this scroll-time cause. | Pages pass Google's page-experience check (a light ranking signal), and readers stop losing their place. | Tech | P6 item 4 |
| 7 | MediumPV header and main image flip as the page loads Not started Fix proposed, not started. The top of the page draws one layout, then switches to another, and the main image isn't in the first version Google gets. | What we saw: the header and main image switch layout after the page loads. Cause: the main image is missing from the first HTML; the device check starts empty. Where: page tests and code on PV articles, Nov 2026. | Less page jumping, and Google sees the main image straight away (useful for image search and Discover). | Tech | P1 item 4 P6 item 6 |
| 8 | MediumAds are requested twice on PV Not started Fix proposed, not started. The same ad slot is set up twice, so the page asks for ads twice. | What we saw: the same Google Ad Manager slot is defined twice, so ads are requested twice. Where: PV code and page tests, Nov 2026. | Cleaner ad delivery and reporting, slightly faster pages. No direct search gain. | Tech (ad ops to confirm) | P6 item 7 |
| 20 | MediumListing pages have many main headings, or none Not started Fix proposed, not started. Each story card on a listing is marked as a main heading, and some pages have none. | Too many: 11 main headings (H1) on PV listing pages and DH tag pages, in the page Google reads and in the browser (checked 19 Nov 2026). None: DH section pages (e.g. /india) show their H1 only after the page's code runs; Google's first copy has no H1 and no story list either. Horoscope articles: no H1 on either site. Cause found in code: story cards default to H1 on both sites; the DH section page draws nothing until the browser runs its code. Source: frontend code (PV 16 Nov, DH 18 Nov 2026) and Chrome DevTools, 19 Nov 2026. | Google understands what each listing page is about. Small. | Tech | P4 item 6 |
| 21 | MediumEvery PV article has a second copy at a short address Not started Fix proposed, not started (DH already redirects). Each article also opens at prajavani.net/<story-name>, with the home page's title. | What we saw: every PV article also opens at /<story-slug> with a normal page, its own canonical and the home title.Source: every PV article tested, Nov 2026. | No duplicate copies competing with the real article; less wasted Google effort. | Tech | P2 item 2 |
| 22 | MediumWrong or empty titles on several page types Not started Fix proposed, not started. App pages, some tag pages and static pages show the home page's title, or none. Search and bookmark pages can appear in Google. | Wrong: home title and description on app pages and some tag pages. Empty: title and description on static pages. Should be hidden: search, bookmark and subscribe pages are indexable. Cause: code, not manual entry. Unknown page types fall back to the home page's title; /about-us is a hand-coded page with no title set; for 4 PV tags the tag lookup comes back empty, so the home title is used (checked 19 Nov 2026). Only the broken tag description is also a CMS gap (empty description fields). Source: page-type check on both sites, Nov 2026; PV code and Quintype API, 19 Nov 2026. | Better-looking search results, more clicks on those pages, and junk pages kept out of Google. | Tech | P5 item 1 |
| 25 | MediumMany old PV redirects go to the wrong place Not started Fix proposed, not started; full list ready. Of 303 hand-written redirects, many send readers to search or the home page, loop through several hops, or end on missing pages. | Wrong target: 84 of 303 hand-written PV redirects go to search and 18 to the home page. Broken: 14 chains, 4 end on 404 pages, 69 can never fire, 3 duplicate entries. Source: redirect check, Nov 2026. | Old links and bookmarks reach the right page, and their value passes to it. | Tech; SEO team (redirect list) | P2 item 8 |
| 29 | Medium7–9% of visits have no landing page in GA4 Not started Cause found; fix proposed, not started. Our analytics records the page view late, so some visits start without a page. | What we saw: GA4 landing page "(not set)" on 7–9% of sessions on both sites. Cause: the page-view event is sent late (1.5 s extra wait on articles), so another event starts the session. Source: GA4 and code, 15 Nov 2026. | Accurate reporting of which pages bring readers. Needed to measure this plan. | Tech; Analytics | P6 item 11 |
| 30 | MediumSponsored DH stories appear in Google Not started Fix proposed, not started; needs commercial sign-off. Paid stories on some DH sections can be found in Google, and a robots rule stops Google seeing the 'don't index' instruction on others. | Blocked but indexed: 38 of 44 sampled /dhbrandpr/ sponsored stories are in Google although robots.txt blocks them.Indexable: sponsored stories on /dhbrandspot/ and /connect/.Clear: no manual actions or security issues on either site (16 Nov 2026). Source: URL Inspection, 16 Nov 2026. | Paid content handled the way Google expects, protecting the site's standing. | Tech; DH ad sales (sponsored content) | P2 item 10 |
| 33 | MediumPV search titles start with English words Decision needed Editorial test to plan. Most PV titles shown in Google begin with an English keyword before the Kannada headline, and many are cut off. | Prefix: 79% of PV search titles start with English words before the Kannada headline. Cut off: 2,052 titles are stored cut off with "…". Source: story data, 16 Nov 2026. | Possibly more clicks from search results. Needs a test before rolling out. | PV editorial (title test); Tech | P5 item 8 |
| 34 | MediumDH live blogs and articles send Google weak signals Not started Fix proposed, not started. Live blogs don't tell Google they're live, and article data is duplicated or badly formatted. | Live blogs: no live-blog markup on DH live blogs, on AMP copies either. Articles: article data carries escaped HTML; og:type is "website" with no og:url; an SVG fallback image.Source: live check, 16–17 Nov 2026. | Live blogs can get the 'LIVE' badge in Google, and articles are understood better. | Tech | P5 item 3 P5 item 4 P5 item 5 P3 item 5 |
| 38 | MediumGoogle keeps fetching old DH image addresses Not started Fix proposed, not started. Image URLs from the old DH site return empty pages, and Google keeps asking for them. | What we saw: Google keeps requesting old Drupal image addresses ( /sites/dh/files/…), which return an empty page.Scale: 17–31K requests a day. Source: crawl stats, Nov 2026. | Google's effort goes to real pages instead. | Tech | P2 item 7 |
| 39 | MediumArticle pages are very heavy Not started Still open (re-checked 19 Nov 2026). DH's footer part is fixed; PV unchanged, including on the pending release branch. Each article page carries large amounts of menu, footer and style data that readers don't need. | PV: 2.2 MB of HTML per article (1.9 MB on 16 Nov); 1.3 MB of it is page data, mostly the full menu (651 KB) and full footer stories (379 KB) that readers don't see. DH: 1.3 MB; its footer fix (13 Nov) cut the page data to 223 KB, so the rest is inline styles (653 KB) and icons (368 KB). Code: PV still loads full footer stories; both sites still send the whole menu to the browser. Source: live check of one article per site and frontend code (PV 16 and 19 Nov, DH 18 Nov 2026), 19 Nov 2026. | Faster pages on phones, and less work for Google per story. | Tech | P6 item 1 P6 item 2 P6 item 3 |
| 40 | MediumPremium stories are free to read on AMP Decision needed Decide with #26 (AMP). The AMP copy of a paid story shows the full text with no paywall. | What we saw: premium stories are readable in full, free, on their AMP copies. Scale: 958 DH (7.8%) and 658 PV (1.7%) premium stories, 7 Sep – 5 Nov 2026. Cause: AMP pages get no subscription setup. Source: live check of all 8 sampled premium stories, 17 Nov 2026. | Protects subscription revenue and keeps the paywall consistent with what we tell Google. | Product (paywall decision); Tech | P3 item 1 |
| 41 | MediumPV 'Also read' links point to a broken address Not started One-line fix proposed, not started. The desktop 'Also read' card links to a doubled address, which creates duplicate pages. | What we saw: the desktop "Also read" card links to /https://www.prajavani.net/<story>, which opens as a page.In Google: 47 such URLs indexed, 2,572 impressions (19 Oct – 15 Nov 2026). DH: same code, no effect. | No duplicates, and the links pass value to the real stories. | Tech | P1 item 3 P2 item 4 |
| 6 | LowMid-article ad block also makes the page jump Not started Fix proposed, not started; ships with #5. The Taboola block inside the article has no space kept for it, so text moves when it loads. | What we saw: the mid-article Taboola block pushes the text down when it loads. Where: page tests on PV articles, Nov 2026. Ships with #5. | Steadier reading on phones; small gain. | Tech | P6 item 5 |
| 9 | LowPremium PV stories show a loading spinner first Not started Optional fix, low priority. For paid stories the page shows a spinner before the text, including to Google. | What Google and readers see: 115 of 397 words; the rest of the page is a loading spinner. Scale: 1.7% of PV stories. DH premium stories are also near-empty (17 Nov 2026, see #40). Source: live check, 16 Nov 2026. | Small. Premium stories already perform well. | Tech | P1 item 5 |
| 10 | LowAd areas are redrawn after the page loads Not started Fix proposed, not started. Ad containers are drawn differently on the server and in the browser, so they get redrawn. | What we saw: ad areas use the screen width before it is known, so the page is redrawn after loading. Where: PV code, Nov 2026. | Slightly faster, steadier pages. | Tech | P6 item 7 |
| 11 | LowMid-article ads showed only on the first story Fix built Fixed in the PV #1 code; goes live with #1. After a reader opened a second story, the in-article ads stopped appearing. | What we saw: the mid-article ad counters are never reset, so ads appear only on the first story read. Where: PV code, Nov 2026. | Ad revenue on every story a reader opens. No search effect. | Tech (ships with #1) | P1 item 6 |
| 12 | LowThe app-view check on PV does nothing Decision needed Removed in the PV #1 code; product to confirm how app views should work. A setting meant to change pages opened inside the app works the wrong way round. | What we saw: the app-view check ( ?app=true) is inverted, so it never does anything.Where: PV code, Nov 2026. | Clean, predictable app pages. No search effect. | Tech; Product (app decision) | P1 item 7 |
| 13 | LowEvery article sends Google two indexing instructions Not started Fix proposed, not started. Each article has two 'robots' tags that can disagree. | What we saw: two robots tags on every article, on both sites. Source: live page check, Nov 2026. | Removes a risk of Google following the wrong instruction. Small. | Tech | P5 item 2 |
| 16 | LowPV font files fail to load for Google Not started Fix proposed, not started. Some font files are broken, so Google renders pages with fallback fonts. | What we saw: PV font requests show as errors ("Other error") for Google. Source: Search Console, Nov 2026. | Pages look right for Google and for readers on slow connections. Small. | Tech | P6 item 9 |
| 23 | LowPage 2, 3… of listings point Google back to page 1 Not started Fix proposed, not started. Paginated listings tell Google that page 1 is the real page, so Google may ignore later pages. | What we saw: page 2, 3… of listings carry a canonical pointing to page 1. Where: both sites, Nov 2026. | Google keeps finding older stories through listing pages. Small on its own; part of #32. | Tech | P4 item 2 |
| 24 | LowA group of small fixes, and the piracy tags Not started Small fixes not started. Piracy tags kept as risky traffic (19 Nov 2026). Several small tagging and redirect defects. Separately, DH tag pages for piracy sites bring a lot of traffic but Google can drop them at any time. | Piracy-name tag pages: 17M impressions in 28 days and 28% of DH Google web search clicks (Search Console, Oct 2026); kept and monitored. Small fixes: DH /amp/horoscope self-canonical; two conflicting robots tags on DH brandspot; two-step http:// redirect; tag description "Read stories listed under on X". | Small fixes: tidier pages. Piracy tags: tracked separately so a drop doesn't surprise us. | Tech; DH editorial (piracy tags: risky traffic) | P5 item 9 P2 item 9 P7 item 3 |
| 28 | LowMade-up DH tag addresses still show a page Not started Fix proposed: copy PV's check to DH (two lines). PV already shows "not found". Any tag URL on DH, even one that doesn't exist, returns a normal page instead of 'not found'. | What we saw: /tags/zzqq-no-such-tag-123 on DH returns a normal page with the home title, its own canonical and 12 latest stories; on PV the same address correctly returns "not found" (404).Cause: PV's tag code stops when no story has the tag; DH's copy is missing that check. Source: live checks, 12 and 19 Nov 2026; PV and DH frontend code, 16–18 Nov 2026. | Less wasted Google effort on junk URLs. Small. | Tech | P2 item 6 |
| 35 | LowPV article data misses the publisher name and logo Not started Fix proposed, not started. PV's article data points to a publisher entry that doesn't exist, plus a few smaller defects. | Publisher: PV article data points to a publisher that is never defined, so Google gets no name or logo. Also: organisation data defined three times; /zzz/<slug> returns a page; ?app=true indexable.Source: confirmed live 17 Nov 2026. | Google shows the publisher correctly. Small. | Tech | P5 item 6 P5 item 7 P2 item 3 |
| 37 | LowSome DH stories can't be opened Not started One-line fix proposed, not started. Stories with accented letters in the address (like niño) redirect to themselves forever. | What we saw: DH stories with letters like "ñ" in the address redirect to themselves forever and can't be opened. Scale: 7 of 12,262 recent stories; 111 "Redirect error" URLs in Search Console, 80 with such letters. Source: Page indexing, 18 Nov 2026. | Those stories become readable and indexable again. | Tech | P2 item 5 |
| 17 | NoteSome problems don't show up in our reports Ongoing SEO team process. Search Console says pages are healthy even when Google has no body text for them. | What we saw: Search Console showed the affected pages as healthy. How #1 and #2 were found: only a special search test showed the missing text (16 Nov 2026). | Regular re-tests catch problems early and prove fixes work. | SEO team | P7 item 4 |
| 18 | NoteAI search tools are blocked from our sites Decision needed Decision needed: is blocking AI crawlers deliberate? Our robots file blocks ChatGPT, Perplexity, Claude and other AI crawlers. | What we saw: AI crawlers are blocked in robots.txt on both sites. Source: robots.txt, Nov 2026. | If allowed: referral traffic from AI search answers. No effect on Google Search either way. | Leadership (editorial + product) decision | P7 item 5 |
| 36 | NotePV gets very little Google News traffic Investigating Investigating with Publisher Center. Even before the November drop, PV got very little from Google News. | Clicks: about 14K Google News clicks a month on PV against about 400K on DH (Oct 2026). Sitemap: PV's news sitemap is compliant, so the gap is not the sitemap. Since 12 Nov 2026: PV Google News fell further, about 450 → 45 clicks a day (see #42). | Unknown; Google News may carry few Kannada sources. Worth checking because DH shows the size of the channel. | PV editorial + Quintype (Publisher Center) | P7 item 2 |
| Already fixed | |||||
| 4 | HighGoogle hit a login error on every page Done Fixed 25 Oct 2026. A subscription check on every page returned an error to Google. | PV: about 18,000 login errors (401) a day to Google. DH: the same call failed (400) on about 4% of Google's visits. Status: both stopped by the robots.txt block, live 25 Oct 2026; verified in crawl stats on 12 Nov 2026. | Done: Google's visits now go to real pages. | Done (Tech) | — |
Release order:
- Ship first, because they are cheap:
- the "Also read" link fix (P1)
- the story 301 rule and the DH non-ASCII redirect loop fix (P2)
- the Taboola reserved space (P6)
- Biggest single item: PV's article server HTML (P1).
- Wait: the AMP decision (P3) comes about 4 weeks after P1 is live and measured.
Part 2: New verticals
These are new pages, page types and tag improvements that answer searches people already make, many of which we already rank for, badly or by accident. The evidence comes from our own Search Console and GA4 data, collected for the Content Engine research.
Data used:
- Search Console: 28-day exports from Nov 2026, covering about 12 Oct – 10 Nov 2026. Impressions and clicks are 28-day totals unless the text says "per day". Position is the average.
- GA4: organic landing sessions, 14 Oct – 10 Nov 2026.
- Google Trends: used only where marked, and only for seasonality. It is not traffic.
Summary
| Vertical | Site | Status | Evidence (Search Console unless marked) | Owner |
|---|---|---|---|---|
| 1. Horoscope hub and sign pages | PV, DH | Tickets to Quintype planned; technical fixes in P4 | PV /horoscope 406K impressions, 33.5K clicks; DH /horoscope 196K, 22K |
Quintype; PV and DH editorial |
| 2. Panchanga, calendar, holidays, festivals | PV (DH holidays) | Calendar built, unknown to Google on 11 Nov; 2027 pages planned | PV panchanga page 143K impressions, 1.7% CTR; panchanga queries ~75K | Product; platform team |
| 3. Exams and results pages | PV, DH | Idea | DH "ugc net" 192K impressions; PV "kpsc" 63K | Tech (monitor); PV and DH editorial |
| 4. Weather and ಹವಾಮಾನ hub | PV, DH | IMD key pending; hub not started | PV "weather" 580K impressions, "ಹವಾಮಾನ" 220K; PV climate stories 768K | Product (IMD); PV editorial |
| 5. Dharane crop prices | PV | Page live; article engine built, not published | 111 Dharane URLs: 66K impressions, 1.6K clicks | Product; PV desk |
| 6. AQI pages and alerts | DH, PV | Widget and engine built, not live | DH AQI 87K impressions (mostly Delhi); Karnataka case from Trends | Product and tech; PV desk |
| 7. Gold and silver rates | PV, then DH | Idea; data licence not signed | PV gold queries 111K impressions, 3.7K clicks; "ಚಿನ್ನ" position 3.9 | Product and business |
| 8. Cricket series hubs | DH, PV | Idea | DH "ind vs" 595K impressions; PV cricket 481K at ~1% CTR | DH and PV sports desks |
| 9. DH topic hubs: OTT releases, Gemini AI | DH | Idea | "ott release" 183K impressions; "gemini" 1.6M | DH entertainment and tech desks |
| 10. Government scheme explainers | PV | Idea | PV scheme queries ~43K impressions | PV editorial |
| Risky: piracy-name tags | DH | Live; growth plan on hold for editorial sign-off | /tags/filmyzilla 6.9M impressions, 117K clicks; tag pages ~3% of DH organic sessions (GA4) |
DH editorial; SEO team monitors |
| Risky: matka cartoon hub | PV | Idea | Cartoon pages 1.55M organic sessions (GA4); "matka cartoon" 241K impressions | PV editorial; SEO team monitors |
Order: 1–3 first. Demand there is proven and steady, we already rank, and most of the work is fixing or extending pages we already have. Then 4–7, which need data feeds and approvals. Then 8–10, which are newsroom habits rather than builds. The risky pages are kept and grow only with editorial sign-off; they never count as growth (see How we'll know).
1. Horoscope hub and sign pages (PV, DH)
What: PV: one hub at /horoscope with today's 12 signs, 12 permanent sign pages, dated titles, one H1, JSON-LD, and links from the calendar. DH: build on its 12 existing sign tag pages, and possibly add city-wise daily pages.
Evidence:
- PV:
/horoscope406,390 impressions, 33,469 clicks, position 6.2./daily-horoscope85,693 and 2,191. The query "dina bhavishya" 43,013 impressions at position 5–6. - PV, GA4:
/horoscope19.8K organic sessions, 89% engaged, 120 s per session. - DH:
/horoscope196,471 impressions, 21,975 clicks, position 9.6. "horoscope today" 30,410 impressions at position 11. - DH, GA4: horoscope pages 85–92% engaged, 136–210 s per session, but under 1K sessions each.
- Trends (India): horoscope searches are down 26% year on year. Growth has to come from ranking better, not from demand.
Blocked by: fix package P4 item 8 (#27). PV's hub is split across four URLs, and DH's daily text is JavaScript-only.
Risk: shrinking demand. Who owns the astrologer content is open.
2. Panchanga, calendar, holidays and festivals (PV; DH holidays)
What:
- Get the PV calendar indexed.
- One panchanga page. Today the calendar's panchanga page redirects to
/astro-vastu/panchanga, so two pages compete. - A 2027 holiday list, a 2027 panchanga page and PDF, and festival pages.
Evidence:
- PV
/astro-vastu/panchanga: 143,277 impressions and 2,473 clicks, only 1.7% CTR at position 5.7. Panchanga queries total about 75K, for example "today panchangam in kannada" 12,647. - "kannada rajyotsava": 131,496 impressions on PV and 93,187 on DH.
- PV
/calendarhad no Search Console data, and Google didn't know the page on 11 Nov 2026. - Trends: "2027 holiday calendar" is rising.
Risk: low. This is evergreen, yearly demand.
3. Exams and results pages (PV, DH)
What: one permanent page per exam, updated every cycle: SSLC, PUC and KCET first, then KPSC, NEET and JEE. Result day gets a live article plus the permanent page. DH adds UGC NET.
Evidence:
- DH: "ugc net" 191,847 impressions and 5,709 clicks. One UGC NET story had 211,589 impressions and 6,070 clicks at position 4.5. "kea" 113,582 impressions but only 93 clicks.
- PV: "kpsc" 63,367 impressions, 3,149 clicks, position 5.4. The SSLC/PUC timetable story had 10,050 impressions and 1,022 clicks at position 2.7.
- Trends: KCET is the biggest exam search and peaks in May; SSLC peaks in May; PUC in April.
Risk: the official sites have no API, and nic.in timed out from the office network. Each page needs a desk owner every cycle.
4. Weather and the ಹವಾಮಾನ hub (PV, DH)
What: a Kannada weather hub on PV and IMD alert articles. The data source depends on IMD API approval; Open-Meteo, a paid service, is the fallback.
Evidence:
- PV: "weather" 580,458 impressions at position 5.2, and "ಹವಾಮಾನ" 220,152 at position ~4.4. PV climate stories (92 URLs): 768,294 impressions and 11,047 clicks (1.4%).
- DH: "weather bengaluru" 152,517 impressions at position 4.5, but only 0.46% CTR.
- Trends: weather searches peak in May and stay high through November.
Risk: Google's own weather box takes most clicks (CTR under 0.5%), so the target is alerts and Kannada explainers rather than forecasts. IMD's terms are pending.
5. Dharane crop prices (PV)
What:
- Commodity × market pages and weekly crop roundups (paddy, maize, arecanut and others).
- Sharp price-move alerts.
- Titles in the words people search (ಬೆಲೆ, ರೇಟ್).
- A sitemap, and links between the articles and the pages.
Evidence:
- 111 Dharane URLs had 66,079 impressions and 1,611 clicks.
- Price-word queries: maize 12,100 impressions, arecanut 1,172 (24,081 including all arecanut queries), cotton 4,376, tur 1,682, onion 1,420.
- No GA4 data yet.
Status: the Dharane page is live and the article engine is built. Publishing is blocked: pasting into the CMS doesn't save the article (tested 17 Nov 2026).
Risk: PV "cotton" queries (122K impressions) look like matka searches, so they are excluded from crop demand. Roundups published unedited would be thin pages.
6. AQI pages and alerts (DH, PV)
What: district air-quality pages with a nearest-station fallback, spike and record articles, and a Monday roundup. Data comes from CPCB through OpenAQ.
Evidence:
- DH: AQI queries 87,265 impressions and 445 clicks, but mostly Delhi ("delhi aqi" 40,779). Karnataka place names give only 590 impressions.
- PV: AQI 368 impressions.
- The Karnataka case rests on Trends: AQI searches peak in November at 3.9× an average month, and "bangalore aqi" is the top AQI search in Karnataka.
Status: the widget and the article engine are built, not live.
Risk: thin city pages. Early drafts read like data dumps. Health text needs PV desk sign-off.
7. Gold and silver rates (PV first, then DH)
What: a daily gold and silver rate page in Kannada, using a licensed rate feed (IBJA or a licensed city-rate provider). Never scraped.
Evidence:
- PV: gold queries 111K impressions and 3,707 clicks. "ಚಿನ್ನ" 89,362 impressions at position 3.9.
- DH: "silver price" 13,581 at position 4.5. DH gold demand is small: "gold rate" 163 impressions.
- Trends: gold-rate searches are about 19× "horoscope" and up 93% year on year.
Risk: no licence is signed yet. One page per city would be thin, because rates are the same statewide. Some "golden satta / golden matka" queries are gambling, not gold.
8. Cricket series hubs (DH, PV)
What: one permanent page per series (schedule, squads, results, stories) instead of one-off stories.
Evidence:
- DH: "ind vs" queries 594,750 impressions ("ind vs wi" 258,551) at position ~8, with about 0.6% CTR.
- PV: cricket queries 481,360 impressions, 4,691 clicks (~1%). PV's one series-schedule story got 17,578 impressions and 1,176 clicks at position 3.4.
Risk: Google's scorecard takes most clicks. Value comes from schedule and squad searches, not live scores.
9. DH topic hubs: OTT releases and Gemini AI
What: a permanent "new on OTT this week" hub instead of weekly listicles that rank at 3.7–4.6 and then fade, and a Gemini AI hub in /technology.
Evidence:
- "ott release" 182,545 impressions over 76 queries and 821 clicks. Release-intent OTT queries total ~196K.
- "gemini" 1,603,499 impressions and "gemini ai" 972,655, though much of it comes from one story.
- No GA4 data yet.
Risk: OTT and Gemini searches follow the news cycle. Ramayana (263K impressions) is a one-time film spike and is not planned as a hub.
10. Government scheme explainers (PV)
What: evergreen Kannada explainers for schemes and processes, kept up to date.
Evidence: PV scheme queries ~43K impressions, for example "ration card ekyc karnataka" 16,239 at position 5.8.
Risk: low. Each page needs updating when the rules change.
Risky traffic: kept, never counted as growth
These pages bring a lot of traffic, but Google can stop showing them at any time, through a core update, a spam update or a manual action. We keep them, report them separately, and grow them only with editorial sign-off. Details are in fix package P7 item 3.
DH piracy-name tag pages
- Evidence:
/tags/filmyzilla6,865,398 impressions and 116,843 clicks./tags/movierulz9,090,524 impressions and 47,281 clicks (0.5% CTR)./tags/moviezwap5.6M and 40,395./tags/bolly4u1.5M and 33,330./tags/tamilrockers509K and 12,504.- GA4: all
/tags/pages had 162K organic sessions, about 3% of DH organic, with 33–47% engaged against a site average of 51%.
- Already volatile:
/tags/tamilrockersfell from 2,859 to 215 clicks a day between 3–5 Nov and 12–16 Nov 2026. - Planned: title and description tests on movierulz and moviezwap, and better tag pages (current releases, where to watch legally).
- Not worth building: the proposed new test tags (ibomma, vegamovies and others) have almost no DH search demand.
PV matka cartoon hub
- Evidence:
- Cartoon pages had about 1.55M organic sessions, 94–96% engaged, the best engagement of any group (GA4).
- "matka cartoon" queries 241K impressions and 104,819 clicks at position 2.5.
/tags/matka124,514 impressions but 68 clicks.- "kalyani chart" 242,343 impressions at position 9.3.
- Planned: rebuild
/tags/matkaas a cartoon hub (today's cartoon, then the archive). No result-chart pages, ever. - Risk: gambling association for the brand. Cartoon demand dropped by more than half in the Deepavali week and came back, which shows how volatile it is.
Gaps in the evidence
- No GA4 data yet for AQI, Dharane, gold, exams, weather or the calendar. Each one gets a GA4 page group at launch.
- No revenue or ad-yield data for any vertical.
- Some figures in the Content Engine planning docs were corrected in this page:
- UGC NET demand is DH, not PV.
- DH AQI demand is Delhi, not Karnataka.
- The DH OTT total is about 366K, not 416K.
Decisions needed
| Decision | Who | When |
|---|---|---|
| Keep or retire AMP (#26), and how premium stories appear on AMP (#40) | Leadership, product | ~4 weeks after P1 ships on PV |
| Grow the risky tags further, or only keep them as they are | DH and PV editorial | Before any new piracy or matka tag work |
| Allow AI search crawlers (#18) | Leadership | Any time; no effect on Google |
| Data licences: gold rates, IMD weather | Product, business | Before verticals 4 and 7 |
| PV SEO title format: English prefix test (#33) | PV editorial | Next month |
What each team does
- Tech (website): P1 → P2 → P4 → P5 → P6, in the release order above. Every item is linked from Part 1.
- Quintype:
- What changed on 11–12 Nov for PV Google News.
- DH server errors (P3).
- Horoscope hub tickets.
- Calendar redirect.
- PV editorial:
- Publisher Center.
- SEO title test.
- Horoscope, panchanga and scheme pages.
- The matka cartoon hub decision.
- DH editorial:
- Piracy-tag decision.
- Sponsored content labels.
- Exam, OTT and cricket hubs.
- SEO team:
- Weekly scorecard and monthly body-text re-test.
- Watching the risky pages.
- Evidence for each new vertical.
- Product:
- Content engine: Dharane publishing, AQI, calendar, gold and weather data.
- Decisions on the app view (#12) and the paywall on AMP (#40).