Content marketing has a terrible feedback loop. You publish, and then you wait weeks to find out whether it did anything. Most people fill that gap by checking their analytics daily and concluding nothing.
Here are the checks I use that give a signal within days instead of months. They're not substitutes for the outcome metric, they're leading indicators — things that are true of content that eventually works.
Check 1: are you being indexed at all?
Before anything else: is the page in the index? Search Console → Coverage (now "Pages").
If a page isn't indexed, nothing else matters. The reason is usually one of:
- Not in the sitemap, or discovered but not crawled
- Discovered, crawled, judged not worth indexing (content quality)
- Crawled, blocked by something (canonical pointing elsewhere, noindex)
I had 25 pages sitting in "Discovered — currently not indexed" for weeks. That category means Google knows the URL exists and has not bothered. That's a content signal, not a technical one.
Days to signal: 3–7 after requesting indexing.
Check 2: how deep is your page, relative to what's ranking?
Pick the query where you rank worst-but-visible. Not your best query — that one is ranking for reasons you can't diagnose. Your worst-visible query is where the deficit is legible.
Then measure body text on the top five results. Not word count of the article, actual rendered body text.
What I found on one query:
| Position | Site | Body text |
|---|---|---|
| 1 | competitor A | 21,746 |
| 2 | competitor B | 16,115 |
| 3 | competitor C | 10,971 |
| 4 | competitor D | 8,733 |
| 5 | competitor E | 7,809 |
| ~82 | mine | 3,653 |
There is no mystery here. If you're at 17% of the median result's depth and ranking on page 8, the causality is not ambiguous.
Days to signal: same day. This is a measurement, not an experiment.
Check 3: does your page answer the questions the search results imply?
Search your target query and read the "People also ask" box, plus the subheadings of the top results. Those are the questions the query is actually asking.
If your page doesn't answer them, you have content that is about the topic rather than content that answers the question. The difference shows up in dwell time and in whether Google promotes you.
Concretely, for one of my pages the implied questions were: how long should this be, what structure, can I mention this sensitive thing, what if I cry, can I read from paper. My page answered none of them and had two examples. Now it has six FAQ entries covering exactly those.
Days to signal: same day, because it's a checklist.
Check 4: are the pages actually reachable?
Crawl your own site from the homepage. Every URL in your sitemap should be reachable by following links, ideally within 2–3 hops.
I ran this and found 12 pages that existed only in the sitemap. Zero internal links pointed to them. They were orphaned — discovered only via sitemap, with no internal link equity at all.
crawled 102 / reachable 102 / orphans 12
Fixed by adding index cards linking to them. After: orphans: 0, reachable 114.
Days to signal: same day.
Check 5: is the structured data consistent with the page?
If you emit FAQPage JSON-LD but the questions aren't visible on the page, best case it's ignored, worst case it's a manual action. Generate both from the same source so they can't diverge:
{cat.faq?.length > 0 && <FaqJsonLd items={cat.faq} />}
{/* and further down in the same component */}
{cat.faq?.map((item) => <details key={item.q}>...</details>)}
Days to signal: same day.
Check 6 (the one everyone skips): do you have backlinks?
Count them. Not "domain authority," just: how many URLs link to yours.
Mine was two. Two links. Site-wide. While ranking on page 7–9 for everything and wondering whether the content was the problem.
If you're under twenty and not ranking, your content quality is not the bottleneck and no amount of editing will fix it. This is the single most common misdiagnosis I see in "I wrote 50 posts and got no traffic" posts.
Days to signal: immediate, and the fix takes weeks.
What these checks can't tell you
None of them predict clicks. You can have a deep, well-structured, indexed, internally-linked page answering the right questions and still get nothing, because the query has no commercial intent or because the market is somewhere else entirely.
For my own project the honest version is: three months of search data showed queries in French and Spanish, not Japanese, and about one impression per day. So the content work is real but the market question is upstream of it.
The order I'd do these in
- Backlink count (5 minutes, tells you if anything else matters)
- Content depth vs. competitors on your worst-visible query (30 minutes)
- FAQ / implied-question check (30 minutes)
- Orphan crawl (10 minutes with a script)
- Index status (passive, check weekly)
Steps 1 and 2 will diagnose most "my content isn't working" situations before you've written anything new.
(Project in question: Speech Kobo. The open dataset it produced: GitHub.)
Top comments (0)