WatchTheStars.co.uk — Full Website Review (1 Sept 2026)

Verified against the repo and live site today. Prior audits (14 Aug, 9 Aug, 1 Jul) were cross-checked so this only lists what is still open or new.

GSC snapshot (pulled 2 Sept 2026, last 28 days, Web search)

Totals: 333 clicks / 35.6k impressions / 0.9% CTR / avg position 39.5. Last 90 days: 615 clicks / 64.5k impressions, so the last month is more than half the quarter — traffic is accelerating. 298 pages received impressions; 249 of them got zero clicks.

Where clicks come from

Page Clicks Impr. Pos.
/meteor-showers/ 172 (52% of site) 6,144 26.6
Roman launch how-to-watch post 63 1,320 5.2
Homepage 19 90 47
August eclipse guide 10 4,178 12.2
Everything else 69

By section: root pages 201 clicks, blog 115, equipment 15, skills 2, planets 0 (6.6k impr.), constellations 0 (2.7k), stars 0 (1.7k).

Top queries: almost entirely meteor-shower intent — "meteor shower 2027 uk" (p2), "next meteor shower uk 2027" (p2.3), "perseid meteor shower 2027" (614 impr., p9), "when is the next meteor shower uk" (192 impr., p11.7). Keyword themes across 1,000 queries: meteor 157 queries / 36 clicks; planet 140 / 0; telescope 56 / 8; venus 60 / 0; eclipse 55 / 0; mars 46 / 0; constellation 42 / 0; tonight 37 / 0; aurora, northern lights, ISS, geminids, orionids: 0 queries each (no page exists to rank).

What this changes in the priorities

  1. Meteor showers are the business. One page is half the site's clicks at position 26. Splitting into per-shower pages (Perseids, Geminids, Orionids, Draconids, Leonids, Quadrantids, Lyrids) plus a "next meteor shower UK" evergreen answer block is the single biggest growth lever, and it needs an equipment card yesterday. Geminids (December) is the next spike — build that page this month.
  2. Planet/constellation/star pages get 11k impressions and zero clicks at position 60–70. Queries are generic ("mercury planet", "diameter of earth", "orion's belt"). They're ranking for encyclopedic terms against Wikipedia/NASA. Either re-angle them to UK-observing intent ("how to see Mercury from the UK this month", "Orion's Belt: how to find it tonight") or accept they're link equity, not traffic.
  3. Eclipse guide: 4,178 impressions, 0.2% CTR at p12. Rewrite the title/meta to the current intent (retrospective + next UK eclipse) as well as the body.
  4. Striking-distance pages (p4–12, 100–600 impressions, near-zero CTR): fifth UFO file release (568), Perseid guide (495), planet parade (477), AARO FY2025 report (315), Bagram (262), Gulf of Oman (235), Brazil 1963 sphere (223), lunar eclipse tonight (207), Varginha (172), Puerto Rico 1964 (163). UAP deep-dives rank well but titles/metas aren't earning the click — rewrite those 10 titles for CTR first.
  5. Confirms the tool gaps: "tonight" (37 queries, 0 clicks), "dark sky" (105 impr., p84), "best telescopes" (p58) all show demand the current pages aren't converting.

0. Do these first (highest impact, lowest effort)

# Action Why Effort
1 Fix the stale August eclipse guide (blog/_posts/2026-07-solar-eclipse-august-2026-uk-guide.md) — still says "order today", "order by Monday", "forecast nearer the day". Rewrite as a retrospective, link the 13 Aug recap at the top, bump the byline date Biggest cited/impressioned page on the site; AI answers are grounding on wrong copy 30 min
2 Add equipmentCard to the 8 root pages missing one: meteor-showers, stargazing-for-beginners, stargazing-tips, uk-dark-sky-sites, astronomy-events, constellations, halleys-comet, uap-research (+ homepage) /meteor-showers/ alone is ~5.7k sessions/month with zero monetisation 1 hr
3 Pull a fresh GSC performance export (clicks/impressions by URL). gsc-indexing-audit.xlsx is from April, pre-dates 150 posts, and has no click data Everything below should be prioritised by where traffic actually lands 15 min
4 Create /eclipse-glasses/ (currently 404, ~130 impressions at position 11, commercial intent) as an equipment page with FLO/Amazon cards Free commercial traffic already knocking 1 hr
5 Backfill FAQ blocks on the 80 older posts still missing them (Feb–Jun backlog) Directly tied to the 200+ never-cited pages; new posts already do this Batchable with a script + review
6 Delete ~26 MB of junk in images/blog/ (sgl-photos.zip, sgl-photos (1).zip, sgl-tmp/) Repo hygiene / build time 5 min

1. SEO and GEO

Already solid (don't touch): clean titles/metas/canonicals across 350+ pages, hreflang, WebSite+Organization graph, FAQPage on 255+ pages, Product schema on equipment pages, per-crawler Content-Signal in robots.txt, IndexNow, security headers, immutable asset caching.

Still open from prior audits

New gaps

GEO-specific

2. Monetisation

Current state: FLO affiliate (primary, all via products.json, rel="sponsored" correct), Amazon UK (12 products, tag live), 100% of blog posts carry a card and disclosure. No display ads. Newsletter is real plumbing (Turnstile → Supabase → Resend weekly send) with one lead magnet. The eclipse-2027 travel affiliate plan in docs/AFFILIATE_PROGRAMS.md shows no evidence of having been applied for or linked.

Gaps and quick wins

New ideas, ranked by value ÷ effort

  1. Head-to-head comparison pages (Seestar S50 vs Dwarf 3, Heritage 130P vs Evostar 72ED, Celestron vs Sky-Watcher beginner scopes). Comparison intent converts far better than roundups and the site has none.
  2. Newsletter sponsorship / classifieds. Once you know the subscriber count (Supabase), a single sponsored slot from FLO, Rother Valley or Widescreen Centre is realistic at a few thousand subscribers.
  3. Direct deal with FLO rather than the standard affiliate %: a negotiated rate or flat monthly sponsorship once traffic is provable from the fresh GSC export.
  4. Raptive (threshold now ~25k pageviews/month) is the realistic display-ad route; Ezoic's threshold rose to ~250k users, and Mediavine now wants existing ad revenue. Track monthly pageviews and apply the month you cross 25k. Test on a subset of pages; don't stack ads on top of cards and the signup on tool pages.
  5. Seasonal gift guides (Christmas telescope guide, "best astronomy gifts under £50/£100") — strong Q4 affiliate intent, low effort, annual refresh.
  6. Digital products: a paid printable UK 2027 sky calendar / eclipse planner PDF, sold via Ko-fi or Gumroad. Small but 100% margin and reuses existing content.
  7. Astronomy insurance / dark-sky holiday cottage affiliates for the dark-sky-sites pages (Awin has Sykes, Hoseasons etc.).

Risks: indexing rather than conversion is the current cap (per the April audit); ad clutter will hurt CWV and the "honest guides" positioning; the eclipse window is time-boxed.

3. Missing tools and pages that would drive traffic

The site's one interactive asset, the /tonight/ Sky Score, is good but is a composite. The queries people actually type are single-purpose, and single-purpose tools rank. Most of the maths already exists in tonight-sky.js, so these are extractions more than builds. All run client-side or via a small Cloudflare Worker cache.

Priority Tool / page Target queries Data Build
High ISS + Starlink pass predictor "ISS pass tonight UK", "Starlink train tonight" CelesTrak TLE + satellite.js, or N2YO Client-side; postcode → passes with time/direction/brightness; per-city variants
High Aurora forecast "aurora forecast UK tonight", "northern lights tonight" NOAA SWPC Kp/Ovation JSON, cached via Worker Live status page with colour alert; the existing aurora skill guide has nowhere to send readers
High Interactive UK light-pollution / dark-sky map "light pollution map UK", "dark sky sites near me" Existing getLightPollution logic + Leaflet Postcode → nearest sites sorted by distance; upgrades the static list
High Moon phase tonight / next full moon "moon phase tonight", "when is the next full moon" SunCalc.js Trivial; distinct from full-moon-names
High "What planet is that?" sky identifier "bright star in the south west tonight" Existing ephemeris in tonight-sky.js Direction/altitude picker; mobile device-orientation mode later
Med Per-city stargazing pages (30–50 UK towns) "stargazing near Manchester" Bortle + nearest dark site + sunset/moonrise + club Build-time static pages from a towns JSON
Med Individual meteor-shower pages (Perseids, Geminids, Quadrantids, Lyrids, Orionids…) "Perseids 2026 UK", "Geminids tonight" Existing events data Each has its own seasonal spike; umbrella page splits the intent
Med "Night sky in [Month]" evergreen pages (12) "night sky September" Events calendar Refresh yearly; link from weekly night-sky
Med Astrophotography exposure calculator (NPF/500 rule) "500 rule calculator" Pure JS Cross-links skills guides + camera/mount cards
Med Opposition / conjunction countdown pages "Jupiter opposition 2027", "Venus Jupiter conjunction" Ephemeris Evergreen-refreshed rather than lost in weekly posts
Med UAP sightings map "UFO sightings map UK" Pre-plotted from the 45 incident pages Static Leaflet map; becomes the UAP hub linking page. Avoid live user reports (moderation)
Low Sunrise/sunset/twilight fold into per-city pages SunCalc
Low Comet tracker event-driven Better as posts per bright comet

4. Other ways to build the site out

Suggested 30-day order

Week 1: table items 1–6, Event schema, CSP enforce, robots fixes, eclipse-travel applications. Week 2: moon-phase tool, aurora tool, /eclipse-glasses/, comparison pages ×2. Week 3: ISS/Starlink predictor, per-shower pages, Pagefind search. Week 4: dark-sky map, first 20 per-city pages, gift guide draft for Q4, review fresh GSC data and re-prioritise.