Discovery
38 h
New stories took a median 38 hours to be indexed. The only sitemap was one archive file of 18,400 URLs, rebuilt overnight, and nothing told Google which stories were new.
Finding 1 of 4
A fictional business. The figures are a modelled scenario, not a client result.
Umzila Weekly · Digital community news publisher · Johannesburg
0 → 41 Top Stories appearances a week
12 months · Middle East & Africa · Signal plan
Late edition
Top Stories at last
Johannesburg paper back on page one of Google
0 → 41
Top Stories appearances a week
01The challenge
No news sitemap, a metered paywall that hid article text from crawlers and inconsistent datelines kept the paper out of Top Stories.
Umzila Weekly's 22 journalists break council, transport and sport stories across Johannesburg every day. By the time Google found those stories, a day and a half later, the national titles and aggregators already owned the carousel. The reporting was first; the plumbing was last.
One story, five datelines
The same Monday-morning article, as five different templates printed its date:
04/05/2026Day or month first? Google has to guess.
2026-05-04A date with no time: a 07:00 scoop and a 19:00 follow-up look the same.
Mon 04.05 · 09:00Human text with no year and no time zone.
2026-05-04T09:00:00ZMarked as UTC: the story looks two hours older than it is.
<time>09:00</time>A time element with no machine-readable datetime at all.
After: one format on every template
2026-05-04T11:00:00+02:00
02The audit
Discovery
38 h
New stories took a median 38 hours to be indexed. The only sitemap was one archive file of 18,400 URLs, rebuilt overnight, and nothing told Google which stories were new.
Finding 1 of 4
Paywall
14%
The metered paywall cut every article to its first two paragraphs for anyone without a cookie, Googlebot included. Crawlers saw 14% of the text, and nothing in the markup explained why.
Finding 2 of 4
Datelines
5 formats
Five dateline formats across the templates, none with a time zone. 31% of stories showed a new dateModified whenever a reader comment was approved, so the dates said nothing.
Finding 3 of 4
Structured data
0%
Not one article carried NewsArticle markup. The theme printed generic WebPage JSON-LD with no author, no image and the site name where the headline should be.
Finding 4 of 4
03The strategy
Built on our Signal plan: the technical engagement for sites where crawling, rendering and markup decide what search engines see.
A news sitemap of the last 48 hours, rebuilt the moment a story is published, alongside the archive sitemaps split by month. Google learns about a story in minutes instead of waiting for the next crawl.
NewsArticle JSON-LD generated from the CMS: real headline, author with a profile page, lead image in three ratios, publisher and section. Live blogs get LiveBlogPosting.
One date component for every template: ISO 8601 with the +02:00 offset, the same time in the byline and the markup, and dateModified only for substantive edits — never for comments.
The meter stays for readers. Verified Googlebot gets the full text, and the gated body is declared with isAccessibleForFree and a cssSelector, so Google sees a paywall, not cloaking.
04Print the poster
Turn the fixes on and off. The poster on the pole reprints with what Google can do with a story published right now.
What Googlebot reads
0 of 4 fixes on
A model, not a promise: eligibility is necessary but never enough, and Google alone decides what appears in Top Stories.
Top Stories · today's edition
Not in Top Stories
Most of the article is hidden from Google and the paywall is not declared.
0
Top Stories appearances a week (modelled)
05Month by month
Months 1–2
Crawled the 18,400-story archive, read eight weeks of server logs, tested what Googlebot sees behind the meter and catalogued every dateline template.
Months 3–4
One date component across all templates; NewsArticle and LiveBlogPosting JSON-LD generated from the CMS, with author pages for all 22 journalists.
Month 5
Paywall markup on the gated body, full text for verified Googlebot only, the meter unchanged for readers. Checked with the URL Inspection tool on 200 stories.
Month 6
A 48-hour news sitemap rebuilt on publish, listed in the sitemap index; the archive split into monthly files.
Months 7–12
Section hubs for council, transport and sport, a live-blog template for load shedding and storms, and a weekly check of Top Stories presence.
06Results
Top Stories appearances a week
0 → 41
0 → 41 a week
Organic article sessions a month
+168%
46,000 → 123,300 a month
Publish to indexed (median)
38 h → 18 min
38 h → 18 min
Subscriptions from organic search
+74%
310 → 540 subscriptions a month
Modelled scenario for a fictional business — not a client result. Rankings and traffic are never guaranteed; see our Terms.
Publish to indexed, median
The news sitemap is fetched within minutes of publishing; the 18,400-URL archive now sits in monthly files.
| Query | Before | Month 12 |
|---|---|---|
| joburg water outage today | not shown | slot 1 |
| load shedding schedule gauteng | not shown | slot 3 |
| soweto derby result | not shown | slot 2 |
| johannesburg council budget vote | not shown | slot 1 |
| braamfontein road closure | not shown | slot 2 |
| jukskei river flooding | not shown | slot 1 |
Carousel slot on a weekday morning in Johannesburg, mobile. Queries shown as readers typed them.
Running a newsroom whose stories reach Google too late? Start with the Signal plan, or tell us what your templates print.