Story one · Media and publishing
SEO for News Publishers That Keeps Up With the News
SEO for news publishers is judged in hours. A story that is not fetched, parsed and filed under the right section by the evening has already lost its readers to whoever published ten minutes earlier. We fix the plumbing that decides which of those two sites you are.
- Crawl and index triage for breaking stories
- Section, tag and pagination architecture
- Author pages, editorial policy, trust signals
- News sitemaps, canonical rules, syndication control
- Archive review and evergreen story refresh
The problem on the desk
Why a fast newsroom still loses the story
None of the failures below are writing failures. They are failures of plumbing, and they compound every single day the site publishes.
Index lag on the story that matters
An article can be perfect and still sit in a queue behind forty thousand tag combinations, a stale sitemap and a template that ships nine megabytes of advertising script. Crawl budget is finite. When archives and faceted URLs eat it, the breaking story waits.
We have seen a regional daily where the crawler averaged nine hours to first fetch. By then the traffic had moved on, and the story was indexed as history rather than news.
Sections that multiply without a rule
Every new topic spawns a tag. Every tag spawns an archive. Every archive spawns pages two through forty, each one indexable, each one near-duplicate, none of them owned by an editor. Six months later the site has four thousand URLs that exist because a CMS default allowed them to.
The fix is a written rule, not a plugin: one canonical path per story, a hard cap on tag pages, and a decision about what deep pagination is allowed to do.
An archive nobody owns
Six thousand stories, half of them about a product line that no longer exists, a third linking to images that 404. The evergreen explainers that still earn links are buried three clicks deep and were last updated in 2019 — while the homepage promotes yesterday’s copy of a wire story.
Archives are an asset with a maintenance cost. Treated as one, they return traffic for years.
What we actually change
Above the fold and below it
The visible half is the page. The invisible half decides whether the page is ever read by anyone but your subscribers. Both get the same review.
Find out where the requests go
Thirty days of server logs, crawl stats and template-level response codes show which URLs absorb the budget. Most engagements start here, with our SEO Content Writing Services review of logs, status codes, render path and Core Web Vitals on the article template rather than the homepage.
Typical findings: a tag archive answering 200 with sixty links, a section page loading ads before text, a canonical that points to a paginated URL.
One story, one address
Each article keeps a single canonical path under its home section. Tags stop being an indexable infinite tree: they are capped, merged where they overlap, and kept as hubs only where real readers search for the topic. Faceted combinations from author, date and tag filters are handled deliberately instead of by accident.
Freshness needs a paper trail
A news sitemap limited to the last 48 hours, section sitemaps for evergreen hubs, accurate lastmod values, and an automated check that no removed story stays listed. If the sitemap lies, the crawl pattern adapts to the lie.
Show the byline’s evidence
Author pages with beat, biography, credentials and consistent profile links; article markup that names the author and the publisher; a visible corrections policy and an editorial standards page. Trust is read by humans first and parsed second — the markup only makes the existing work legible.
Wire copy is not your copy
Partner republishing, agency feeds and leftover legacy URLs create near-duplicates that split signals. We set the canonical discipline, remove outdated parallel versions, and give syndicated partners a clean attribution rule the desk can follow without a meeting.
Retire, merge, or update on a schedule
Overlapping explainers are merged into one strong page. Evergreen guides that still attract links get a quarterly refresh with an honest update date. Retired topics are redirected or left to fade, decided by a rule rather than by whoever notices first.
Architecture
A section tree built for a busy news day
A news site is a tree that grows a branch every morning. If the branches are not pruned with the same regularity, the trunk stops carrying water.
We draw the tree before we touch it: home, four to eight top sections, a shallow layer of subsections, and articles three levels at most. Every article is reachable in three clicks from the front page and one click from its section hub. The most-read module, the related-stories block and the breadcrumb trail all point inward to the same structure, which is how a section hub accumulates authority instead of leaking it.
Tag pages: cap, merge, or ignore
A tag page earns its place when readers actually browse it: a championship, an election, a long-running investigation. Everything else is a CMS convenience. Keep the browsable ones as editor-owned hubs with a paragraph of context and a curated lead story; merge overlapping tags; leave the rest to internal search rather than the index.
Pagination that protects the archive
Pages two and three of a section can be useful. Page thirty-seven is not. We let early pages stand on their own, keep self-referencing canonicals, and stop deep pagination from competing with the hubs it belongs to. The goal is simple: when a reader wants last Tuesday’s story, the archive serves it; when a crawler arrives, it does not walk the entire vault first.
Breadcrumbs and the section hub
Every article carries a visible breadcrumb back to its section, and structured breadcrumb markup that matches it. Section hubs get a stable intro paragraph, a defined story order and links to their subsections. That single change often does more for a section’s standing than another hundred articles published into it.
Publishing rhythm
Cadence decides your index queue
Editors think in deadlines. Search engines think in revisit patterns. The two only agree when the publishing schedule is legible from outside — and that is a technical decision as much as an editorial one.
| Story type | Window | Technical action | Signal we watch |
|---|---|---|---|
| Breaking / live | 0–60 min | News sitemap ping, stripped template, no lazy hero, inline critical text | Time to first fetch and first impression |
| Daily report | Same day | Section placement, two internal links from related stories, canonical set at publish | Impressions in the first 24 hours |
| Weekly explainer | 3–7 days | Author markup, question block, related-links module, one hub link | Queries won, not just clicks |
| Evergreen guide | 30–90 days | Refresh cycle, honest update date, merged duplicates, consolidated canonicals | Ranking stability across quarters |
| Retired topic | 90+ days | Redirect, merge or leave out of the index after review | Crawl share freed for news |
Distribution now happens on several surfaces at once: classic results, feed-style recommendations and the section hubs inside your own site. Each reads structure differently, and none of them can be bought into. What we can do is make the structure unambiguous, so a story is easy to classify, easy to attribute and easy to update. Our Technical SEO Services work treats those surfaces as one system rather than three campaigns — without promising placement on any of them.
The engagement
The first six weeks, step by step
No redesign, no migration, no six-month discovery phase. We start where the loss is measurable and work outward from there.
Thirty days of requests mapped by template. We fix 5xx and soft-404 patterns, canonical conflicts and the worst render-blocking offenders before anything else.
A one-page rule set the desk can apply without asking us: what gets a section, what gets a tag, what stays out of the index.
Author pages, profile links, article and breadcrumb markup, corrections and standards pages published in the CMS.
News sitemap automation, stale-URL checks, index alerts by section so a drop is caught the morning it happens.
Refresh queue for evergreen pages, monthly review with the editor and the dev lead, written handover for the desk.
Evidence
Three publishers, three bottlenecks
5 months: time to first crawl 9h → 40 min after the template and sitemap work; organic sessions to the markets section up 41%, with no change in publishing volume.
7 months: 1,900 tag pages consolidated to 140; crawl requests to article templates down 28%; organic inquiries 6 → 19 per month from the same archive.
4 months: 12 author pages and 60 refreshed guides now account for 3 of its top 10 landing pages; newsletter signups from search up 22%.
Figures are from client accounts under NDA, measured against a pre-engagement baseline, and reported with the caveat that every newsroom has a different season. We do not publish a magic multiple, because we have never seen one hold across two publishers.
Stop the clock
Send us the URL of the story that never got traffic
We will tell you what the crawler saw, which template sent it there, and what we would change first. If we are not the right fit, that answer is still useful to you.
Scope
How a quote is actually built
There is no price list, because a fifty-story-a-week newsroom and a monthly journal are not the same job. A quote follows a short discovery call and four numbers: how many sections you run, how many articles you publish, how many templates your CMS renders, and how much developer time we can realistically book.
Project sprint or monthly retainer
A four-to-six week sprint sets the architecture and hands it over. A retainer keeps the crawl clean, runs the refresh queue and reports monthly. Most publishers start with the sprint.
Tickets your developers can use
We write the specification, your team ships it, or we ship it with you. Every item arrives with acceptance criteria, so “fix the canonicals” becomes a ticket someone can close.
The deliverable is not a slide deck. It is a written playbook that survives our departure: rules, templates, dashboards and a refresh calendar your desk can run on its own.
Questions from the desk
What editors ask before signing
How quickly does anything move?
Crawl behaviour can shift within days of a template or sitemap fix, and we monitor it weekly. Ranking movement is slower and noisier in news verticals, so we judge the first eight to twelve weeks on indexation, fetch rate and section-level impressions before talking about positions.
Is this white hat?
Yes. Log analysis, architecture, markup, editorial signals and content quality. No link networks, no scraped syndication, no doorway pages posing as city editions. If a tactic only works until the next update, it is not in scope.
Do you need CMS and log access?
Log access and read-only CMS or template access make the first two weeks far more precise. If your host cannot share raw logs, we work from crawl stats, rendering tests and sitemap behaviour instead, with a wider margin of error we state up front.
Will you work with our developers?
That is the normal arrangement. We join the sprint board, write tickets in your format and review pull requests that touch templates, canonical logic or sitemap generation. We do not ask a newsroom to hand over its CMS.
What do you need from our team each month?
One editor contact for section and tag decisions, one developer contact for tickets, and a monthly thirty-minute review. Roughly two hours of combined time. Most of the work is ours, not yours.
How do you measure success?
Four layers: fetch and index rate by template, impressions and clicks by section, queries won per evergreen page, and inquiries or subscriptions attributed to organic entry. We publish the baseline in week one so the argument later is about data, not opinion.
Can you promise coverage or placement?
No, and nobody credible can. Distribution surfaces decide for themselves. Our job is to remove the reasons a story is passed over and to make the site legible; the decision stays with the platform.
Do you write the journalism?
No. We write structure, briefs, evergreen refreshes and the rules the desk follows. The reporting stays with your newsroom — that is the part that cannot be outsourced and should not be.