Section: Technology & AI Author: V Reading time: ~27 min Sources and further reading: 22 items Topics: Google, digital footprint, security, deindexing, GDPR, public registers, own data SEO / working title: What can Google find out about you from a single name?
What can Google find out about you from a single name? The question sounds simple only until we notice that we are not comparing one thing. A search result may contain a true record and still belong to a namesake. Deindexing can remove a link while the page remains. The cadastre can show a property and still not be a people-search engine. A European request can change the European version of Search without deleting the same URL from google.com. This text therefore does not ask what “Google knows”. It asks what you will see about yourself in the public layers — and what of that is not a case file.
This text is not a guide to tracing another person, an address, a phone number or a place of residence. The tests below are an inventory of your own footprint. The point is not to assemble someone else’s life from registers. The point is to read your own visibility without panic and without false certainty.
1. In a private window your name is typed. Beside it stand someone else’s managing director and an old snippet
Before an interview, a job applicant opens a private window and types their own name as they use it in e-mail and on documents. They find an old profile of their own, someone else’s LinkedIn, an article about a namesake, a company listing that has nothing to do with them, and a snippet from a page that has already changed its text. A managing director opens the public register and sees a function recorded by law. A person with a common name finds someone else’s photograph. A reputation office handles hundreds of URLs. A journalist reads an official disclosure. Google ranks links; it is not the publisher of every page.
At first glance this is a dispute about what Google can find out about a person. In fact each description is of a different layer. The applicant compares visibility, not the completeness of an identity. The managing director compares the publicity of legal dealings, not a biography. The bearer of a common name compares a string match, not identity. The office compares the volume of URLs in requests, not the average Czech. The journalist compares a statutory disclosure, not a private file.
That is why everyone can be partly right and the shared conclusion still wrong. A search result can be unpleasant and still incomplete. It can contain a true fact, but about another person. It can show a public register, but only your role in legal dealings. It can hide a page after deindexing on a name search, while the page still exists.
The first rule of a safe inventory is therefore not: find out everything about yourself. It is: sort what you see. Links you published yourself. Official disclosures. News coverage. Obvious namesakes. Outdated snippets. Platforms that have their own visibility settings. Without that sorting we read a mixture as a verdict.
A name in Search is not a case file. It is a ranked selection with namesakes and gaps.
— Jiný Kontext
2. One name has at least six layers
A name in search looks like a simple key. It is not. It is a string of characters that passes through an index, registers, platforms, cache, the right to deindexing and territorial rules. When those layers are mixed, two opposite errors appear. One says: Google knows everything about me. The other says: once I deleted something, nobody will see it any more.
Six layers of one own footprint:
- Index and ranking. Google fetches publicly available URLs, stores information and ranks the results. What is not crawlable need not appear in name results. What is crawlable can appear even against the will of the person shown.
- Public registers and official disclosures. The public register, the Collection of Deeds, ARES, the insolvency register and other portals have statutory purposes. They are not a CV.
- Platforms and self-publication. LinkedIn, social networks, company sites, academic databases and old profiles often say more about visibility settings than about what a search engine “knows”.
- Namesakes and panels. The same name string can merge different people. A Knowledge Panel can attach the wrong context. A match of the name is not identity.
- Rights against the search engine and the source controller. Deindexing a result on a name search, erasure at the publisher, and the right of access against a controller are different acts [1][5][6][7].
- Territory, cache and copies. The European version of the results, google.com outside the EU, the cache, the snippet, another search engine and an archive are not one thing [2][14].
This layering is not a detail for lawyers. It decides where to turn when something fails. The wrong person in a panel is not the same as a live page. An outdated snippet is not the same as current content. A request to Google is not the same as a request to the publisher. A public register is not the same as a social network. And the cadastre is not a telephone directory of names.
Interpretive scheme: One name leads into six layers. The index shows links, the register shows a statutory role, the platform shows a visibility setting, the law addresses a specific controller, territory changes the scope and cache changes the time.
3. The best inventory is the one whose error you see before an outside reader does
For your own digital footprint it is not enough to count results. What matters more is knowing which error is likely and how to recognise it. Sometimes the problem is too much visibility. Sometimes it is a mix-up with a namesake. Sometimes an outdated snippet that survived a correction of the page. Sometimes an official fact that someone reads as a private biography.
| Type of failure | What it looks like | How to test it on yourself | What limits the harm |
|---|---|---|---|
| Namesake | Someone else’s profile, article or listing under your name | In a private window, separate your own links, registers, news and obvious namesakes | A string match is not identity |
| Outdated cache | The snippet lives on after the page has changed | Compare your own URL, the live page and the snippet | Ask for a refresh of outdated content [14] |
| Deindexing is not erasure | The link disappears on a name search; the source remains | After a request, distinguish the result from the live source | Two acts: Google and the publisher |
| Territory | The European version differs from the result outside the EU | Read C-507/17 as a rule of scope, not as a promise of global erasure | European measures make circumvention harder; they do not delete the site [2] |
| Register as biography | A company function is read as a whole life | Check only your own role and the statutory purpose of the disclosure | Publicity of legal dealings is not doxxing |
| Public profile | The search engine shows more than you want | Check the public visibility of your own profile | Changes appear with a delay [20] |
The best inventory is therefore the one that expects an error in advance. It does not ask only how many links sit on the first page. It asks what each link is, whether it belongs to you, whether it is current, who published it, what rights you have against it, and whether its purpose is to inform the public about a role, or about a private life.
4. A search result is a ranked selection. It is not a case file
A case file has an originator, rules, a procedural purpose and responsibility. A search result has an index, ranking, snippets, ads, panels and links to other people’s pages. These two things can meet in the same browser, but they are not the same. When a search engine shows a court decision, an article or a register entry, that does not mean it has assembled a complete dossier on a person.
CJEU judgment C-131/12 of 13 May 2014 matters precisely because it recognised the special role of a search engine in a search by name. In that processing the search engine operator is a controller of personal data, and under certain conditions the data subject may ask for a link to be removed from the results [1]. The court did not say that Google is the author of every page. It said that ranking and making links available by name has an effect of its own.
It does not follow that a search result has the completeness of a police file. It follows that the index can significantly shape what an ordinary reader sees first about a person. That is exactly why an inventory of your own footprint is worth doing. Not because Google knows everything. Because a ranked selection sometimes decides the first impression.
5. A name is not a unique identifier
A name is a social tool, not a primary key in a database. Two people can share a name, a similar profession, the same city, an old profile and someone else’s article. A search engine does not see a person the way a family or an office does. It sees strings, links, context and signals of relevance.
As of September 2026 the dossier does not contain current official frequencies of given names and surnames from the Ministry of the Interior for the 2026 population. There is therefore no figure here for how many namesakes the average person has. The mechanism is more precise: the more common the name, the greater the risk of mix-up; the more public the role and the rarer the combination of name plus field, the easier the assignment; neither of those turns the result into a case file.
A Knowledge Panel can look like an authoritative card for a person. That is convenient, but not decisive. The panel can join data that belong to a publicly known person, an organisation or another bearer of the name. If someone else’s card appears next to your name, the first reaction should be neither panic nor the assumption that someone has assembled your identity. It is an assignment error, to be handled through the tools of that service.
A safe inventory therefore works with categories, not with guesswork. “This is my profile.” “This is an obvious namesake.” “This is an official role.” “This is an old snippet.” “I do not know.” The last sentence is better than a false link.
6. A public register records you as a function. Not as a biography
Czech public registers have a statutory purpose. Act No. 304/2013 Coll. regulates the public registers of legal and natural persons and the disclosure of data related to registered persons and facts [15]. The Ministry of Justice portal or.justice.cz makes the public register and the Collection of Deeds available [16]. ARES of the Ministry of Finance aggregates data on economic entities from selected source registers [17].
If you are a managing director, a member of a body, a co-founder of an association or a natural person in business, you can appear in these systems. That is the publicity of legal dealings, not a narrative of private life. The entry says that a given person has or had a given role in a given entity. It does not say what their character is, where they spend their time or what they think.
How to test your own register entry
The test belongs only to your own role. If you know you are registered with a given entity, open your entity in or.justice.cz or ARES and note what is disclosed by law. This is not about assembling other people’s registers. It is about understanding what can be visible about you as a person in a role, and why.
It does not follow that public registers are a security flaw. It follows that their result must not be read as a biography. Official publicity can be legitimate and still reputationally sensitive if a reader pulls it out of its purpose.
7. The cadastre asks about a parcel. Not about a telephone directory
The real-estate cadastre is about properties and rights in them. Act No. 256/2013 Coll. regulates the cadastre and its publicity, including limits on inspection in § 52 [18]. ČÚZK describes the cadastre as a system for recording real estate, not as a service for finding people by name [19].
That distinction matters for the security of your own footprint. If you know your own parcel or the address of a building, you can check what data attach to your property. Free inspection is tied to the property: a territory, a parcel, a building or a title sheet. An ownership overview for the whole territory of the Czech Republic and the collection of deeds have a narrower regime under § 52(2) [18].
This article deliberately does not describe how to join other people’s names to addresses. The safe question is: what is visible for my own property, if I know its identification? It is not: how do I find where another person lives. The difference is not cosmetic. It is the line between a personal inventory and a guide to interfering with privacy.
The cadastre can therefore be part of your own inventory, but not as a people-search engine. It is a layer of real-estate records. When someone reads it as a telephone directory, they are reading a different object.
8. Deindexing removes a link on a name search. It does not delete the source
Judgment C-131/12 brought a European right, under certain conditions, to ask for a link to be removed from name-search results [1]. GDPR Article 17 regulates the right to erasure, and Article 12(3) sets the period in which the controller informs the data subject about measures: one month from receipt of the request, with a possible extension of a further two months in more complex cases [5].
These are strong rights, but not magic. Deindexing at a search engine means that, if the conditions are met, the link will not appear for a query that contains the name. It does not mean that the publisher deleted the page. It does not mean that the same content is not elsewhere. It does not mean that an official document has lost its statutory purpose.
The CJEU in C-136/17 addressed sensitive data and deindexing; the EDPB in its Guidelines 5/2019 describes the balancing with the public interest, for example in journalism, official documents or a public function [4][6]. ÚOOÚ also reminds us that the right to erasure is not absolute [9]. That matters for expectations. Not every unpleasant result disappears merely because it concerns a person.
Before a deindexing request
First separate three things: a request to the search engine, a request to the publisher or source controller, and a possible complaint to ÚOOÚ. Google’s form for the European right to privacy protection concerns URLs that relate to you [12]. If you want the content deleted from the source page, you turn to the publisher or the controller. If the controller does not reply, a supervisory authority comes into consideration. The GDPR period is a period for information about measures, not a promise of erasure from the whole internet [5][8].
Deindexing removes a link. It does not delete the page.
— Jiný Kontext
9. European deindexing need not be global
Judgment C-507/17 of 24 September 2019 asked whether deindexing must be carried out globally. The CJEU held that Union law does not impose a duty of global deindexing; the search engine operator must apply measures that effectively make it harder for users in the member state to circumvent the deindexing [2]. The CJEU press release recalls the facts: the French CNIL imposed on Google, on 10 March 2016, a fine of 100 000 EUR for refusing global deindexing [3].
The figure of 100 000 EUR measures one sanction by a French authority against Google Inc. in that dispute. It does not measure the success of erasure, it does not say how many results return globally, and it cannot be read as what every user will see. Its role is different: it shows the conflict between national supervision and the global reach of a search engine.
For the reader the point is that the result on the European version and the result outside the EU need not be the same. That does not mean the right is worthless. It means that deindexing has a territorial and a technical scope. If we expect “deletion from the whole internet”, we will be disappointed by a right that was not built that way.
10. Cache and snippet can outlive the live page
Another common error appears after a page has been corrected or deleted. The live site may already be different, but search results still show an old snippet. That does not mean the publisher is necessarily lying. It means that the search engine may not yet have the current version, or that the preview still holds old text.
Google has a procedure for refreshing outdated content in search results [14]. This tool is not a universal erasure. It aims at the situation in which the content on the page has changed or disappeared and the search results still show old information. Again this is about a specific URL and a specific difference between the live page and the result.
The practical test on your own footprint is simple. For your own page or profile, compare the live content with the snippet in the result. If they differ, first check whether the change on the site is in fact public. Then deal with refreshing the outdated content. Do not read the cache as the live page, and do not read the live page as proof that Google has already corrected everything.
11. A LinkedIn public profile is a different page from the profile for signed-in users
LinkedIn is a good example of the difference between a profile for the platform and a public page for the web. A person can have a rich profile for signed-in users and still limit what an unsigned visitor sees. A search engine usually works with the publicly available version. Looking at your own profile while signed in is therefore not enough.
LinkedIn Help describes public-profile settings and states that changes can take weeks to months to appear in search engines [20]. That period measures the expected indexing delay according to the platform, not a guarantee that every search engine will react in the same way. It is a reservation against a quick conclusion: I changed it in the profile, so Google must show it at once.
The safe test is again your own. Check public profile visibility on your profile and open the public version without signing in. What an unsigned person sees is closer to what a search engine can index. If you do not want a given item shown publicly, you change it in the platform. If an old snippet holds on in search, you deal with delay and, if needed, outdated content.
12. Results about you looks for your data. It is not a search for other people’s
Google describes the Results about you tool as a way for a signed-in user to find and manage results that contain their personal contact details — for example a phone number, an address or an e-mail — if they themselves enter them [13]. The point of the tool is protection of your own data. It is not a service for entering other people’s contacts.
What needs to be tested
The test has a narrow brief: your own account, your own contact, your own result. If the tool finds a URL with your datum, you handle a specific request under Google’s rules. If it finds nothing, that does not mean there is no footprint on the internet. It means only that the tool, in that mode, did not find the match it is designed to look for.
This chapter is a safety brake. An inventory of your own footprint is not OSINT on other people. There is no place in it for procedures that combine another person’s name, address, phone, family and property. If we catch ourselves turning protection of our own footprint into a guide to someone else’s, we have changed the purpose.
The boundary of the test is simple: you enter data that belong to you, and you ask for action only on URLs that concern you.
13. The volume of erasure requests does not measure what Google knows about you
Figures on the right to be forgotten invite a shortcut. The study by Bertram et al. from 2019 analysed five years of requests at Google, from May 2014 to May 2019. It reports 3.2 million URLs in requests, about 502 000 requesters, and deindexing of 44.5 percent of requested URLs after assessment [11].
These figures measure the volume and outcome of requests at one operator in a given period. They do not measure what Google knows about the average person. They do not measure the number of pages on the internet. They do not measure the number of deleted sources. Nor do they say that 44.5 percent is the probability of success for your request; a specific request depends on the content, the public interest, the role of the person, recency and other criteria [6].
The same study states that 84 percent of requested URLs fell into the category of private individuals, which is the default at Google; at the same time 1 000 requesters, that is 0.2 percent, asked for more than 526 000 URLs, that is 16 percent of all requested URLs [11]. Some of those requesters are offices and reputation services. Again this is not the average Czech.
The Google Transparency Report stated, for data from January 2016, as of January 2018, that 1 percent of requesters accounted for about 20 percent of requested URLs, that is 1.4 million URLs [10]. That is a different time window from the Bertram et al. study; the figures do not add up. Surfshark, in a secondary analysis of the dashboard with collection as of 20 February 2025, reports for 2024 about 180 000 requests and more than 800 000 URLs, and the authors note that the interval may not cover the whole calendar year [22]. That too is not a count of deleted pages.
A good sentence is therefore not: Google deletes almost half the internet. It is: in a given period and for a given type of request it deindexed a share of URLs after assessment, while the volume of requests is unevenly distributed and does not describe the completeness of an individual footprint.
14. Visibility on networks is not the same as the index in Search
The Czech Statistical Office, in the VŠIT 2025 sample survey, for the 2nd quarter and persons 16+ in private households, states that 63.2 percent of persons used social networks, that is 5,492.6 thousand people [21]. That measures the use of social networks. It does not measure what Google indexes.
Among network users, according to the same survey, 15.7 percent had a fully public profile and posts, that is 863.9 thousand persons; 79.6 percent said they at least sometimes set privacy, and 4.7 percent do not know the settings [21]. This is a self-report in a sample survey, not a technical audit of indexability. A public post may be available to a search engine, but it need not be. A non-public profile may be less visible, but a screenshot or a quotation elsewhere changes the path.
CZSO also states that 21.2 percent of persons 16+ have asked for the deletion of data previously provided on the internet, that is 1,838.7 thousand persons [21]. This figure includes situations other than the right to be forgotten at Google, for example unsubscribing from a newsletter. It is not a statistic of deindexing.
Visibility on networks therefore cannot be turned into the simple sentence “Google sees your social networks”. The right question is: which part of my profile is public, what the platform allows search engines, what has already been taken elsewhere, and how long a change takes to appear.
15. The right of access aims at a specific controller. Not at the whole internet
GDPR Article 15 gives the data subject the right to obtain from the controller confirmation of whether it processes their personal data, and if so, access to those data and further information [5]. The EDPB in Guidelines 01/2022 sets out how the right of access is to work [7]. ÚOOÚ explains practical questions of processing data on the internet and the reply period [8].
This right is strong, but addressed. We do not write to “the internet”. We write to a specific controller of whom we know or reasonably assume that it processes our data: an employer, a shop, a platform, a club, a website controller. Under Article 12(3) the controller must inform about measures within one month of receiving the request; where needed the period can be extended by a further two months [5].
It does not follow that every footprint disappears from the web in three months. It follows that a specific controller must respond under a specific legal regime. The right of access is there so that you know what a controller processes about you. The right to erasure has conditions and exceptions. Deindexing at a search engine is another layer. Mixing them into one sentence manufactures the wrong expectations.
16. An inventory of your own name says more than the myth of a case file
A practical inventory should be short, repeatable and your own. It is not a search for other people. It is not an assembly of addresses, property and family. It is about what you will see about yourself in ordinary public layers and which steps make sense.
Three tests of your own footprint
The first test is a search for your own name in a private window without signing in. Record the first results in four categories: my own publications, official disclosure or register, news coverage, obvious namesake. Do not add guesses. If you do not know whether a result belongs to you, mark it as uncertain.
The second test is your own contact. In a signed-in account use Results about you only with your own e-mail or phone number, which you yourself enter [13]. The aim is to see whether your contact details appear in results and what request can be filed. It is not a tool for other people’s data.
The third test is your own public page and your own statutory role. On LinkedIn check public profile visibility and the public version of the profile without signing in [20]. If you are registered in a company, an association or as a person in business, check your entity in the public register or ARES [16][17]. If you have your own property and know its identification, check the data for that property in the cadastre regime [18][19]. There is deliberately no procedure here for finding a person’s place of residence.
A useful result is not a long list of links. It is a short record in which each line answers four questions. Does the result belong to me, or to a namesake? Is the page live, or is it only a snippet in search? Who is the controller of the source? Which step makes sense: a change to your own profile, a request to refresh outdated content, a deindexing request, a request to the controller, or no action?
That record has one advantage. It stops the emotion from the first result becoming a bad procedure. An old public profile is handled in the platform settings. An outdated snippet is handled as outdated content. An official role is not handled with the same request as a blog post. Someone else’s namesake is not deleted from your life by reading them as Google’s mistake.
The result should not be a dossier. The result should be a map of visibility: what I published myself, what the state publishes for legal certainty, what a platform displays, what is a namesake and what is an old snippet. Such a map is less exciting than the myth of a case file. It is, however, usable.
17. The hidden cost is a mixture that we read as a verdict
The greatest risk of a name search is not only that we find something. It is that we read a mixture as a continuous story. One link is yours, the second belongs to a namesake, the third is an official role, the fourth an old snippet, the fifth an article that has a public interest. The mind easily assembles a person from that. The search engine did not provide a case file. It provided a ranked selection.
That is why it makes sense to repeat your own inventory after a change of role, after a publication, after deleting a profile, or before a situation in which the first impression matters. Not so that we live under the dictate of search results. So that we do not confuse three different jobs: repairing our own presentation, asking a controller to intervene, and accepting that some statutory disclosure has a purpose other than our comfort.
In security it is tempting to look for a secret procedure. Here ordinary evidence is better. The date of the check. The query you entered. The result that belongs to you. The result that does not. The URL for which you request a change. The controller to whom you write. The reply and the period. Such a record is neither a public profile nor a tracing tool. It is your own working note, which helps when an old snippet shows up again in a month.
The question is therefore not what Google will find out about you from a single name. It is: which layer are you reading right now — and whom do you see in it, when a string match is not identity and deindexing is not deletion of a site?
A rare name joined to a role that in practice creates an almost unambiguous identity would weaken this thesis. Even there, however, the result would not be a case file. A rule of mandatory global deindexing would also weaken it; C-507/17 does not impose that from Union law [2]. So would a flat free search for a person by name in the cadastre; § 52 of the Cadastre Act sets a narrower regime for the ownership overview [18].
The practical answer is more modest. Search for yourself, sort the layers, repair your own profiles, request deindexing where you have a reason, and do not read other people’s namesakes as part of your story. A name is not a case file. It is the start of an inventory.
Related texts in this series
- What a digital footprint is… — a name in Search is one layer of the footprint, not a whole CV.
- What does your mobile phone know about you? — an account and a device are a different vault from Google’s index.