260 Visibility Audits, Zero Appearances: What Six Months of Honest Data Says About a Young SEO Site

Across 260 weekly visibility audits run over 169 days, OnyxRank never appeared in Google's results or in Perplexity's answers for any of the 20 buyer queries we tracked. Over the same stretch we pushed 10,873 URL submissions through IndexNow, and Google Search Console recorded 0 clicks in 210 days.

Most SEO data posts show a success story. This one is the opposite, and we think it is more useful. Small sites get far less published data than large ones, and the pattern in our own logs is a clear picture of what indexing signals do and do not buy you. Everything below comes from OnyxRank's own tracking, and the numbers are exactly what the scripts printed.

What we measured

We have three first party sources, each answering a different question.

1. **The visibility audit log.** Every week or so, an audit checks 20 buyer queries and records whether onyxrank.com is visible in Google, whether it is visible in Perplexity, and which competitors appear instead.

2. **Google Search Console.** The Search Console API for the domain property gives impressions, clicks, queries, pages, countries and devices.

3. **Indexing telemetry.** Our IndexNow and Bing submission logs show what we sent and what the endpoints answered.

The audit log holds 260 records: 20 queries, 13 audit runs, between April 18 and October 3, 2026. Search Console data covers April 1 to October 27, 2026, which is 210 days of daily rows. IndexNow batch runs start May 4 and the last one is October 26.

Finding 1: zero visibility in 260 audits

The headline is simple. Google visible: 0 of 260. Perplexity visible: 0 of 180 checks that could be run. The other 80 Perplexity fields are empty, because on four of the 13 audit dates Perplexity could not be queried directly and no value was recorded. We left those as missing rather than counting them as misses. No audit recorded any context where OnyxRank was mentioned (0 of 260).

Thirteen audit runs, spaced from one week to several weeks apart, show no drift. The count was 0 visible on every date, from April 18 to October 3. Continued publishing in that window did not move a single tracked query from absent to present.

That is not a verdict on quality. It is a reminder that visibility for commercial queries such as "best SEO agency" is a trust problem before it is a content problem. Our audit notes say as much: 75 of the 260 notes describe OnyxRank as absent, 37 mention listicles leading the results, and 22 mention Clutch.

Finding 2: who occupies those results

While we were absent, the same names kept appearing. Across 260 audits we logged 1,411 competitor mentions, an average of 5.43 per audit, covering 420 distinct names.

|---|---|---|---|

The surprise is how thin the concentration is outside a few queries. The top 5 names account for 13.2% of all mentions and the top 10 for 22.0%. Meanwhile 238 of the 420 names appeared in exactly one audit. In other words, the field is long and churning, which suggests that a newcomer is not blocked by a fixed oligopoly. The block is elsewhere.

Some queries are locked down, though. On "programmatic SEO agency," Omniscient Digital appeared in 13 of 13 audits. On "best SEO tools," Ahrefs appeared in 13 of 13 and Semrush in 12 of 13. On "best SEO agency," WebFX appeared in 12 of 13. If a query has a stable winner in your own log, treat it as a different, harder game than the queries where the leader shows up in only 5 or 6 of 13 audits, such as "AI search optimization" and "generative engine optimization."

Finding 3: Search Console confirms it

Search Console is the first party check on the audit log. Between April 1 and October 27, 2026, the domain recorded 199 impressions and 0 clicks.

|---|---|---|---|

April alone accounts for 131 of the 199 impressions. After that the site averaged single digits to a few dozen impressions a month. Google's own [Search Console documentation](https://developers.google.com/search/docs/monitor-debug/search-console-start) describes impressions as the number of times pages appeared on Google Search, so these are appearances in results, mostly at low positions, not visits.

Position matters here. For the 18 queries Search Console was willing to show us, the impression weighted average position was 31.3 and the median query average position was 36.4. Page two and beyond is where almost nobody clicks. The homepage is the exception, with 87 impressions at an average position of 3.4, which fits branded and site operator searches. The most frequent query was `site:onyxrank.com` with 9 impressions, which is a person checking whether we are indexed, not a prospect.

Other details from the same export:

• Those 18 queries account for 41 impressions. The remaining 158 of 199 (79.4%) sit in queries Google withholds for privacy.
• Of the 18 queries, 10 had a single impression.
• The United States produced 144 impressions, Great Britain 12, Canada 10.
• Desktop produced 194 impressions and mobile 5.
• The sitemap at onyxrank.com/sitemap.xml was last submitted October 29 with 0 warnings and 0 errors. The API reports 36 submitted URLs and 0 indexed for its web content entry, which we treat as a rough indicator, not a precise index count.

Finding 4: IndexNow did what it says, which is not what many hope

IndexNow is a protocol that tells participating engines a URL changed. The [IndexNow documentation](https://www.indexnow.org/documentation) lists support from Microsoft Bing, Naver, Seznam.cz, Yandex and Yep. Google is not on that list.

Our logs show the protocol working mechanically:

• 32 batch runs between May 4 and October 26, totaling 10,873 URL submissions. The first batch held 257 URLs and the latest held 442.
• 9 additional single URL submissions after publishing, all of which returned ok.
• Of 123 calls each to api.indexnow.org and bing.com, 122 returned 200 and 1 returned 403 on both, in the first run.
• Yandex accepted 118 calls (68 returned 202 and 50 returned 200). Five failed with a closed connection.

So delivery was near perfect. The business result in Google was still 0 clicks. IndexNow tells an engine a page exists and changed. It says nothing about whether the page deserves a position.

The Bing Webmaster URL submission log makes a related point. It holds 11 records: 10 returned HTTP 400 and 1 returned 200. The first error body reported a quota of 100 remaining for the day against a 296 URL submission, so the batch was simply too large. Daily quotas shape a submission pipeline more than most guides admit.

What this changes in how we work

Three practical conclusions follow from the data, and they apply to any young domain.

**Submission is not ranking.** A clean sitemap with no errors, near perfect IndexNow responses and hundreds of published URLs still produced single digit monthly impressions for most months. Google's [AI features documentation](https://developers.google.com/search/docs/appearance/ai-features) says there are no additional requirements to appear in AI Overviews or AI Mode, and no other special optimizations are necessary. The same page frames eligibility as being indexed and eligible to show with a snippet. Our data fits that: we have little to optimize for in AI answers until the ordinary search foundations produce impressions.

**Choose queries by how stable the winner is.** The audit log tells you which queries are winnable. A query where one brand holds 13 of 13 slots is a long project. A query with a long tail of one time names is a place to start.

**Measure absence honestly.** An audit that records only wins hides most of what you learn. Our log recorded 260 misses, and the misses are the dataset: they name who holds the results, which source types dominate (listicles, directories) and where the field is open.

For the baseline on crawling, indexing and snippets, Google's [SEO Starter Guide](https://developers.google.com/search/docs/fundamentals/seo-starter-guide) remains the reference we measure against.

Methodology

All aggregates below were produced by a Python script (`/tmp/report_stats.py`) run on October 29, 2026. Raw output from the final run:

• Audit log: 260 audits, 20 queries, 13 runs, April 18 to October 3, 2026, span 169 days. Google visible true: 0. Perplexity true: 0, false: 180, null: 80. OnyxRank context non null: 0.
• Competitors: 420 distinct names, 1,411 total mentions, 5.43 average per audit, top 5 share 13.2%, top 10 share 22.0%, 238 names seen once.
• Search Console: 210 daily rows, April 1 to October 27, 2026, 199 impressions, 0 clicks. Query dimension: 18 rows, 41 impressions, weighted average position 31.3, median 36.4.
• IndexNow: 32 batch runs, 10,873 URL submissions, 9 single URL records, May 4 to October 26, 2026. Bing submit log: 11 records, 10 HTTP 400, 1 HTTP 200.

Definitions and limits:

• **Visible** means the audit found onyxrank.com in the first page of results for Google, or in a Perplexity answer. It is a binary judgment recorded by the audit, not a rank position.
• Perplexity could not be queried directly on some dates. Those 80 records are null, and where noted a web search proxy was used. Perplexity figures therefore rest on 180 checks.
• The 20 queries are a chosen set. They skew toward SEO agency and GEO topics and do not represent search in general.
• Competitor names come from audit notes and were not normalized beyond exact matching. "Thrive Agency" and "Thrive Internet Marketing" appear as separate names, so some concentration figures are understated.
• Search Console hides anonymized queries, so query level sums are lower than totals. The report uses the daily total for headline figures.
• This is one young site. These results do not generalize to established domains, and they are not an experiment: nothing was held constant.

Data

All figures come from OnyxRank's own tracking: the visibility audit log, the Search Console API for our domain property and our IndexNow and Bing submission logs. The underlying data can be requested as CSV by emailing [email protected]. You are welcome to cite any figure here with a link back to this page.

If you want to run the same checks on your own site, the free tools at [/tools](/tools) are a place to start.