OnlyDetect's 30 September 2026 snapshot contained at least one stored social link for 365 of 461 eligible profiles. That leaves 96 without a recorded link. Link presence provides a possible verification route; it does not independently establish that the destination belongs to the same person.
We counted directory records and network labels without tracing private relationships, identifying subscribers, or publishing a map of individuals' accounts.
What the directory recorded
The public aggregate dataset describes the selected sample and exact observation times. The median number of stored links per eligible profile was two, and the maximum was five.
Among the recorded network labels, Instagram appeared on 267 profiles, Twitter on 251, and TikTok on 134. These categories overlap: the same record can have links to several networks. Their counts must not be added together to claim a number of unique people.
Scroll across the table on smaller screens.
| Selected network label | Profiles with a link | Stored link rows |
|---|---|---|
| 267 | 268 | |
| 251 | 251 | |
| TikTok | 134 | 134 |
The table uses the database's labels. The full download includes the other recorded network categories. It contains aggregate counts rather than individual account URLs.
Why profiles and links have different counts
The Instagram column contains one more link row than profile count. This means the directory has multiple rows for at least one profile under that network label. It does not, by itself, establish a duplicate creator, a fake account, or an error: a person can publish more than one professional account.
Choose the counting unit before interpreting the result. Counting links answers how many rows are stored. Counting distinct profile IDs answers how many directory records have such a link. Neither is automatically a count of unique real people.
What makes a link useful evidence
A link on a third-party directory is a lead. Stronger evidence connects the relevant public profiles through a creator-controlled page or another primary source with an explicit account reference. Check the handle and the destination, and keep the date of the observation.
The official-link guide gives the practical verification sequence. This study did not independently open and authenticate every stored social destination, so it does not publish an ownership-accuracy percentage.
Missing links are not negative proof
The 96 records without stored social links may reflect source coverage, extraction limits, or a creator choosing not to cross-link accounts. Our count cannot distinguish those explanations. It does not establish that those creators lack public professional accounts elsewhere.
Do not fill the gap with private identifying details or accounts inferred from a photograph. A public verification task can remain unresolved without escalating into personal identification.
How to cite the findings
State that this is a dated OnlyDetect sample of 461 eligible records. Use “stored links” and specify whether a number counts profiles or link rows. The SQL shows the distinct-count calculation and can be rerun against an authorized matching dataset.