# Finding 086: No-hit head-signal repeats stayed answer-stable but differed by origin trace

## Date

2026-07-02

## Status

Published

## Summary

This comparison reviewed the completed Perplexity and Copilot/Bing p37/p38
plain head-signal repeats against the `/lab/reading/head-signal-isolation`
fixture.

| Client | Attempts | Answer-side result | Direct-origin result |
|---|---|---|---|
| Perplexity | `manual-client-perplexity-plain-head-signal-20260702-001-p37`; `manual-client-perplexity-plain-head-signal-repeat-20260702-001-p38` | Both returned `fetched:false`, `pages_opened:0`, and `Failed to fetch url content`. | Both had no exact target-page hit. p37 had ancillary `/robots.txt` and `/` PerplexityBot activity; p38 had no related origin activity in the bounded prompt window. |
| Copilot/Bing | `manual-client-copilot-bing-plain-head-signal-20260702-001-p37`; `manual-client-copilot-bing-plain-head-signal-repeat-20260702-001-p38` | Both returned `fetched:false`, `pages_opened:0`, and `Failed to get web content.` | Both had no exact target-page hit and no related root, robots, or sitemap activity in the bounded prompt windows. |

All four prompts used ordinary site-owner review wording and avoided asking for
source areas, page-head metadata, structured data, code-like tokens, marker
values, or hidden fields. The repeated answer-side no-fetch result was stable
for both no-hit clients. The origin-side trace was stable for Copilot/Bing and
variable for Perplexity.

## What does this mean?

For site owners, SEO/AEO teams, publishers, and researchers, repeated failed-fetch answers are not all the same measurement outcome. Copilot/Bing twice showed a clean no-hit: the client said it could not fetch the page and the server saw no related visit. Perplexity twice said it could not fetch the page, but one run still produced root and robots traffic while the repeat produced none. That means AI visibility audits need to track the assistant answer, the exact target-page hit, and any nearby origin traffic as separate facts.

## Method

This comparison uses the published p37/p38 findings and their response
artifacts:

- Finding 081: Perplexity p37 plain head-signal target no-hit with ancillary
  `/robots.txt` and `/` activity.
- Finding 082: Copilot/Bing p37 plain head-signal clean target no-hit.
- Finding 084: Perplexity p38 plain head-signal repeat clean target no-hit.
- Finding 085: Copilot/Bing p38 plain head-signal repeat clean target no-hit.

Each run was submitted in a fresh client chat or thread, copied into the
matching response artifact, logged with `npm run manual-client:log`, and
reviewed against the local direct-origin event log inside a bounded timestamp
window. The p38 prompts repeated the p37 plain site-owner review condition for
the two clients that had produced target no-hits in the p36/p37 comparison.

## Evidence

| Evidence type | Result |
|---|---|
| Perplexity answer stability | p37 and p38 both returned `fetched:false`, `pages_opened:0`, and `Failed to fetch url content`. |
| Perplexity target-page retrieval | No exact `/lab/reading/head-signal-isolation` target-page hit appeared for either p37 or p38. |
| Perplexity ancillary activity | p37 had `mr3cowad-56fbnx7p` for `/robots.txt` and `mr3cowo8-6b0r2qng` for `/`; p38 had no related root, robots, sitemap, or exact attempt-id activity in the bounded prompt window. |
| Copilot/Bing answer stability | p37 and p38 both returned `fetched:false`, `pages_opened:0`, and `Failed to get web content.` |
| Copilot/Bing origin state | p37 and p38 both had no exact target-page hit and no related root, robots, or sitemap activity in the bounded prompt windows. |

```json
{
  "comparedAttempts": [
    {
      "promptCode": "manual-client-perplexity-plain-head-signal-20260702-001-p37",
      "confirmedHitFromPrompt": false,
      "rawEventIds": [],
      "ancillaryOriginActivity": {
        "present": true,
        "paths": ["/robots.txt", "/"],
        "rawEventIds": ["mr3cowad-56fbnx7p", "mr3cowo8-6b0r2qng"]
      }
    },
    {
      "promptCode": "manual-client-perplexity-plain-head-signal-repeat-20260702-001-p38",
      "confirmedHitFromPrompt": false,
      "rawEventIds": [],
      "ancillaryOriginActivity": {
        "present": false,
        "paths": [],
        "rawEventIds": []
      }
    },
    {
      "promptCode": "manual-client-copilot-bing-plain-head-signal-20260702-001-p37",
      "confirmedHitFromPrompt": false,
      "rawEventIds": [],
      "ancillaryOriginActivity": {
        "present": false,
        "paths": [],
        "rawEventIds": []
      }
    },
    {
      "promptCode": "manual-client-copilot-bing-plain-head-signal-repeat-20260702-001-p38",
      "confirmedHitFromPrompt": false,
      "rawEventIds": [],
      "ancillaryOriginActivity": {
        "present": false,
        "paths": [],
        "rawEventIds": []
      }
    }
  ]
}
```

## Interpretation

The p37/p38 repeat set supports two different kinds of stability:

1. Answer-side no-fetch stability: both clients repeated failed-fetch answers
   across the original and repeat prompt.
2. Target-page no-hit stability: all four bounded origin reviews found no
   exact target-page retrieval for the requested head-signal URL.

It does not support one stable origin-side trace for every no-hit. Perplexity
p37 contacted `/robots.txt` and `/` without fetching the target page, while
Perplexity p38 did not show related origin activity. Copilot/Bing was more
stable across the two runs: both prompt windows were clean no-hits.

Because none of the four runs fetched the exact target page, this comparison
does not make claims about visible body text, page-head categories, hidden
page-head marker availability, or structured-data visibility for the no-hit
clients. It only compares repeated failed-fetch answers and server-side
origin traces.

## Limitations

- This comparison covers two runs per client from one account, plan, client
  mode, fixture, prompt family, and day.
- The p38 repeats reused the same plain wording style but were not randomized
  against other client tasks.
- The prompts supplied exact target URLs, so this measures direct URL opening
  rather than independent discovery.
- The lack of target-page events is bounded to each prompt window and the
  retained local event log.
- A clean no-hit does not expose whether the client skipped retrieval,
  failed before network contact, used product-side routing, or contacted an
  unobserved intermediate service.

## Publication Thesis Verification

- Thesis: In the p37/p38 plain head-signal repeat set, Perplexity and
  Copilot/Bing repeated failed-fetch answers and exact target-page no-hits,
  while Perplexity's ancillary origin activity varied between the original
  and repeat runs.
- Source: Findings 081, 082, 084, and 085; their browser-task artifacts,
  response artifacts, logged answer packets, and bounded direct-origin event
  reviews.
- Method: Compare copied answer fields, prompt codes, bounded timestamp
  windows, confirmed target raw event ids, and ancillary origin activity
  fields across the four completed no-hit runs.
- Bias: Small same-day repeat set on a synthetic fixture, using one account
  and one product mode per client. The prompt is plain relative to
  source-area prompts but still names AI-assistant readability and exact lab
  URLs.
- Consensus: Internally consistent with each individual finding. The answer
  fields repeat for both clients, target-page hits remain absent in all four
  reviews, and the only origin-side difference is Perplexity p37 ancillary
  activity versus p38 clean no-hit.
- Invalidation: A corrected origin review finding a target-page hit in any
  p37 or p38 prompt window, a response artifact assigned to the wrong prompt,
  or missing related origin activity in the p38 reviews would weaken this
  comparison.
- Verdict: Supported for the completed p37/p38 no-hit repeat set. The
  strongest claim is repeated failed target retrieval, not general platform
  behavior or hidden-field handling.
- Additional tests suggested: prepare a later randomized repeat packet for
  Perplexity and Copilot/Bing that alternates client order and URL variants
  to test whether clean no-hits and ancillary-only contacts remain stable.

## Next steps

- Design a randomized no-hit-client repeat packet only if stronger evidence
  about client order, timing, or URL-shape effects is needed.
- Keep page-head marker exposure comparisons limited to confirmed target-hit
  clients until Perplexity or Copilot/Bing fetches the exact fixture.