Audit date: 5 June 2026 | Data period: Feb–Jun 2026
Sources: Google Analytics 4 (28-day & 7-day), Google Search Console (3-month), site files
The site is technically healthy — fast, clean schema, proper canonical tags, AI crawlers allowed. The problem is authority. With only 4 external backlinks from 3 domains, Google has almost nothing to judge the site's credibility against. Result: 28 Google clicks in 3 months across 10,400 impressions. Average position 30. Everything else is noise until this is fixed.
The good news: Bing is working, Reddit sends real traffic, and AI engines (ChatGPT, Copilot) are starting to cite the site. Traffic is trending up (+41.4% vs prior 28 days). The content is clearly finding an audience — just not through Google yet.
| Metric | Value | vs. Prior Period |
|---|---|---|
| Active users | 1,500 | +41.4% |
| New users | 1,500 | +41.4% |
| Sessions (last 7 days) | 459 | +46.5% |
| Avg. engagement time | 31s | — |
| Event count | 13k | +76.3% |
Traffic is growing but engagement time is low. 31 seconds average is roughly half of what good content sites achieve. This signals users are landing and not finding what they expected, or the content isn't pulling them deeper into the site.
| Source / Medium | Sessions |
|---|---|
| (direct) / (none) | 1,100 |
| bing / organic | 134 |
| reddit.com / referral | 102 |
| duckduckgo / organic | 49 |
| facebook.com / referral | 47 |
| chatgpt.com / (not set) | 17 |
| ecosia.org / organic | 14 |
Google organic does not appear in the top channels. This is the site's biggest problem — see Section 2.
Flag: Singapore is the top city at 391 users — unusual for a UK astronomy site. This could be bot traffic, a single link from a Singapore-based aggregator, or a crawl spike. Worth investigating in GA4 (Audience → Geography → filter by Singapore, check engagement time and bounce rate).
| Page | Views | Bounce Rate |
|---|---|---|
| Journey Through The Cosmos (hero) | 522 | 21.9% |
| Six Federal Agents / Orange Orbs | 198 | 55.7% |
| Our Solar System | 141 | 2.1% |
| Astronomy Equipment Guides | 118 | 7.0% |
| Is It Worth Going Out Tonight? | 108 | 15.5% |
| Stars & Star Systems | 103 | 6.3% |
Good bounce rates on the hub pages (Solar System 2.1%, Equipment 7%). The UAP posts have higher bounce (55.7%) which may indicate the audience isn't finding more to read.
| Metric | Value |
|---|---|
| Total clicks | 28 |
| Total impressions | 10,400 |
| Average CTR | 0.3% |
| Average position | 30 |
28 clicks in 3 months from Google is near-zero. The 10,400 impressions confirm Google is indexing and showing the site — it's just not ranking well enough for clicks. Position 30 = page 3 on average.
| Page | Clicks | Impressions | CTR | Position |
|---|---|---|---|---|
| congress-46-uap-videos-deadline | 7 | 1,720 | 0.4% | 10.8 |
| observing-mercury-year | 5 | 446 | 1.1% | 8.5 |
| bst-starguider-8mm (eyepiece) | 4 | 45 | 8.9% | 15.8 |
| / (homepage) | 2 | 162 | 1.2% | 49.8 |
| luna-nonhuman-origin | 1 | 647 | 0.2% | 9.3 |
| western-us-orbs-federal-agents | 1 | 265 | 0.4% | 8.1 |
| nasa-gateway-cancelled-moon-base | 1 | 149 | 0.7% | 7.3 |
Note: The luna, orbs/federal agents, and NASA gateway pages are all sitting at position 7–9.3 — just off page 1. These are the fastest wins for improving click volume.
The BST Starguider page has the best CTR (8.9%) with position 15.8. Equipment review content converts well once it ranks. This validates the affiliate content strategy — it just needs more of it ranking higher.
| Query | Clicks | Impressions | CTR | Position |
|---|---|---|---|---|
| bst starguider 8mm | 2 | 10 | 20% | 7.6 |
| watch the stars | 1 | 15 | 6.7% | 26.1 |
| bst starguider eyepiece review | 1 | 1 | 100% | 21.0 |
| "proxima centauri d" minimum mass | 0 | 67 | 0% | 9.4 |
| saturn | 0 | 66 | 0% | 72.4 |
| star adventurer 2i | 0 | 31 | 0% | 45.4 |
| anna paulina luna | 0 | 28 | 0% | 8.4 |
"saturn" at position 72 is a significant problem. It's one of the most searched astronomy terms and the site has a full Saturn guide — but it's on page 7. This is almost certainly a backlink authority issue, not a content issue.
"proxima centauri d" at position 9.4 with 67 impressions — this is in the current news cycle and near page 1. A targeted push here (better title, internal links, maybe a dedicated follow-up post) could land it on page 1.
| Status | Count |
|---|---|
| Indexed | 181 |
| Not indexed | 50 |
Not indexed — breakdown:
| Reason | Count | Priority |
|---|---|---|
| Not found (404) | 24 | 🔴 Critical |
| Crawled – currently not indexed | 15 | 🟠 High |
| Duplicate without user-selected canonical | 4 | 🟠 High |
| Server error (5xx) | 2 | 🔴 Critical |
| Blocked (403) | 1 | 🟠 High |
| Alternative page with proper canonical | 1 | 🟡 Low |
| Page with redirect | 3 | — |
24 x 404 errors is a regression. The previous audit (11 May) noted 6 were fixed. Either new 404s have appeared since then, or the GSC validation is still pending. These need to be pulled and resolved — each one wastes crawl budget and signals a broken site to Google.
2 x server errors (5xx) — same situation. The May audit listed the /equipment/cameras/ 5xx as resolved, but these may be different URLs or the fix didn't hold.
| Metric | Value |
|---|---|
| Total external links | 4 |
| Referring domains | 3 |
| Linking sites | 2000ad.com (×2), idcrawl.com, x.com |
This is the root cause of all ranking problems. Google's ranking algorithm is built on authority, and authority comes primarily from external links. 4 links from 3 domains gives the site near-zero authority. This explains why "saturn" ranks on page 7 — a competitor with identical content but 50 links will beat it every time. Backlinks were flagged as P1 in the May audit. Nothing has changed here.
No data — insufficient real-user traffic for the Chrome UX Report to generate a score. This is a secondary symptom of the traffic problem. PageSpeed (lab data) previously showed excellent results (TTFB 186ms, page size 67KB), so the vitals are likely fine — just unmeasurable at current traffic levels.
Most of this was resolved in the May 11 audit. Current state:
| Check | Status | Notes |
|---|---|---|
| Title tags (50–65 chars) | ✅ | buildPageTitle filter enforces this |
| Meta descriptions | ✅ | All pages |
| Canonical tags | ✅ | Self-referencing on all pages |
| H1 tags | ✅ | Present on all pages |
| Open Graph / Twitter Cards | ✅ | Complete |
| Schema markup | ✅ | Article, NewsArticle, BreadcrumbList, Person, Org |
| hreflang (en-GB) | ✅ | Added May 11 |
| Sitemap | ✅ | /sitemap.xml, 189 pages, last read 11 May |
| robots.txt | ✅ | All AI crawlers explicitly allowed |
| HTTPS | ✅ | |
| Image format (WebP) | ✅ | |
| Site speed | ✅ | Excellent (67KB page size) |
| llms.txt | ✅ | Created May 11 |
| SpeakableSpecification | ✅ | In blog schema |
| Author page | ✅ | /author/ with Person schema |
| About page | ✅ | Created May 11 |
| Contact page | ✅ | Created May 11 |
| 404 errors | 🔴 | 24 in GSC — up or unchanged from May 11 |
| 5xx errors | 🔴 | 2 in GSC — thought to be resolved May 11 |
| Duplicate canonical | 🟠 | 4 pages — outstanding from May 11 |
| Crawled not indexed | 🟠 | 15 pages — thin or low-authority content |
| Engagement time (31s) | 🟠 | Below benchmark; internal linking may help |
| GA4 conversion tracking | 🔴 | Still 0 key events configured |
| Singapore anomaly | 🟡 | 391 users — investigate for bot traffic |
The GEO setup is ahead of most sites. What's working:
No Wikipedia/Wikidata presence — This is the single highest-leverage GEO improvement available. AI models (ChatGPT, Gemini, Claude) heavily weight Wikipedia and Wikidata for entity recognition. Being cited in a relevant Wikipedia article (e.g., amateur astronomy in the UK, UAP disclosure timeline) would dramatically increase the frequency with which AI systems recommend or cite WatchTheStars. Even a Wikidata entry for Ian Clayton as an astronomy author would help.
No LinkedIn profile — The Person schema in structured data points to Instagram and Facebook. LinkedIn is treated as a more authoritative professional entity signal by AI systems. A LinkedIn profile for Ian Clayton linked from /author/ would strengthen entity recognition.
No YouTube presence — Video content gets special treatment in AI overviews and has its own rich schema. Even one or two observing videos would open up video rich results and add a YouTube entity link to the knowledge graph.
GEO content strategy note: The Artemis mission timeline page is the strongest GEO asset on the site (39 Copilot citations). The format — chronological, factual, data-dense, sourced — is exactly what AI systems prefer to cite. Applying this format to other ongoing missions (Lunar Gateway, Artemis III, SLS, VIPER) would build more citable assets.
1. Investigate and fix the 24 x 404 errors in GSC
Go to GSC → Indexing → Pages → "Not found (404)". Export the URL list. For each: decide if it needs a redirect (to the correct page) or can be left (if it was never a real page). Add 301 redirects in the _redirects file. Push to GitHub. Ping IndexNow.
2. Investigate the 2 x 5xx server errors in GSC Go to GSC → Indexing → Pages → "Server error (5xx)". Identify the URLs. Check if they return 200 now — if yes, click "Validate Fix" in GSC. If still broken, identify and fix the cause.
3. Set up GA4 conversion tracking Currently showing 0 key events across the whole site. Can't measure list growth, tool usage, or affiliate clicks without this. At minimum, configure:
4. Backlinks — start now, not later This is the only thing that will move Google rankings materially. The May audit was right. Specific actions:
5. Fix the 4 duplicate canonical pages in GSC GSC → Indexing → Pages → "Duplicate without user-selected canonical". Find the pages and ensure each has a self-referencing canonical tag. Likely caused by URL variants (trailing slash vs none, or www vs non-www) that aren't being caught by the current canonical setup.
6. Investigate and resolve the 15 "crawled not indexed" pages
Google crawled these but decided not to index them. Causes are usually: thin content (under 400 words), duplicate content, or low perceived quality. For each: expand to 800+ words if worth keeping, or add noindex if not.
7. Investigate Singapore traffic anomaly In GA4, filter by country = Singapore, check: average engagement time (if under 5s it's bots), bounce rate, pages visited. If it's bot traffic it's inflating your user numbers and distorting your engagement rate metric.
8. Push the three page-1 near-misses over the line
These pages are at position 7–9 with lots of impressions and near-zero clicks. A push in title/meta and more internal links could get them to page 1:
9. Improve average engagement time (currently 31s) Low engagement time hurts rankings. Probable causes:
10. "proxima centauri d" — quick win Position 9.4 with 67 impressions and 0 clicks. The query includes very specific astronomical data. The page is nearly on page 1. Add a tight summary paragraph at the top of the JWST/exoplanet post that directly answers the query ("Proxima Centauri d minimum mass is X Earth masses, derived from..."). This helps both the click-through and the AI citation.
11. Wikipedia / Wikidata entry ← highest GEO impact available Create a Wikidata entry for Ian Clayton as an astronomy author. Contribute sourced edits to relevant Wikipedia articles (UK amateur astronomy, UAP disclosure, meteor shower pages) and cite WatchTheStars where genuinely appropriate. This is the single biggest GEO lever not yet pulled.
12. LinkedIn profile for Ian Clayton
Link from /author/ and from the Person schema sameAs array. Strengthens entity recognition for AI systems.
13. More Artemis/mission timeline posts The Artemis mission timeline page gets 39 Copilot citations — the most of any page. Apply the same format (chronological, data-dense, NASA-sourced) to:
14. Shopping tab setup for equipment pages GSC is showing "Display your 35 products on the Shopping tab" as a recommendation. This would put equipment pages in Google Shopping results for free — it requires Product schema and possibly a Google Merchant Center entry. Given the FLO affiliate model, this is a free traffic channel worth setting up.
Report generated: 5 June 2026