Google is like a child: how to educate it to activate a knowledge panel
Celebrities, companies, books, TV series, songs, podcasts, cities, places, monuments and even fictional characters: when you search on Google for these entities it appears (side on desktop, featured on mobile) a knowledge panel that allows the user to read brief pieces of information on the subject, so as to know and immediately frame that entity and, possibly, be encouraged to explore the subject in more depth. For a person with any kind of public profile, understanding how to get and populate this fact sheet is essential to ensure that the information contained is correct and complete, and there is a method to “convince” Google will activate a personal panel for ourselves or customers.
How to activate a personal Google knowledge panel
Officially, Google explains that the fact sheets “are automatically generated” and contain “information that comes from various sources on the Web”; sometimes, the search engine may “collaborate with partners that provide authoritative data on specific topics such as films or music and combine such data with information from other sources open on the Web”.
However, as Jason Barnard reveals on Search Engine Journal, there is a possibility to (legally) force the activation of a Google Knowledge Panel for ourselves, a customer or any other person, following a three-step path that serves to present that entity to Google. As a disclaimer, the author says that “at the moment” this process almost always works, but also that “the optimization of entities is full of subtleties, warnings and loads of depends“, so full certainty is never attainable.
Google and people, ambiguity issues
Ambiguity is a big problem for Google (and, in general, for machines), and people are a rather complex front to untangle because “there could be hundreds or even thousands of people sharing the same name and surname“, how easy it is to verify by searching on Linkedin or Facebook a name and discovering all the homonyms.
This means that confusion and misunderstandings are normal and can also be difficult to solve, because it is not easy to immediately identify the search intent: “after all, which Jason Barnard they mean?” , writes the author.
Therefore, it is necessary first of all to invest time to be clear, consistent and careful so that the process works, to make Google figure out which entity and which of the various homonyms we refer to.
Educating Google just like a child
Barnard uses a valid analogy to define the work needed to achieve the goal of activating a personal knowledge panel: we must think of Google as a child and try to educate and instruct it.
Based on his experience, in fact, the algorithms of the Knowledge Graph that guide the boxes of knowledge on Google are very similar to a child: “both want to learn, are sponges, thirsty for knowledge, and both learn more or less in the same way”.
Just like a child, Google needs:
- Information from the most authoritative source (logically it should be the person himself).
- A clear explanation
- Confirmation from reliable sources.
- A consistent message from all sources.
- Communication in easily digestible and comprehensible language.
- Time.
In this path, we are the teachers and we can take the initiative, defining the facts and guiding the process; to educate effectively, we must be “resolute, consistent, persevering and patient”.
The 3-stage process to obtain a knowledge panel for a person
The first step to get an informational tab on Google is perhaps the most delicate one, since it focuses on building an entity home to be identified by the search engine.
According to Barnard, the entity home is “the authoritative source that Google will use as a reference point for the person”: as we said before, the most authoritative source of information about a person is (should be) that person himself, and Google actively searches for first-hand information.
The article also suggests some examples of possible Entity Home, in order of choice from best to worst:
- The about me page on the personal site of the person in question.
- The home page on the personal site of the person.
- The about us page on the person’s company website.
- A social profile (Linkedin or Twitter are the ones that have proven most effective so far).
Barnard reports that he has activated information panels for many people: the best way is to “choose a domain owned and a page 100% dedicated to the entity”, which has always minimized time, the effort and hassle to reach the goal, although a social profile may “be an option” (confirmed by John Mueller) and a corporate website “is a possibility”.
Then there is an important consideration to keep in mind: this decision is long-term, because when Google accepted an entity home, make it change its mind “it proved exceptionally difficult”. It is not impossible, but it certainly “takes time, is slow, expensive and boring”.
How to create an effective entity home
The expert also offers practical advice on how to create an effective home for the entity, with an indication of the basic information to be entered – which will be those that Google needs to know the person / entity, reported at the end in the knowledge panel.
What is basic is to aim for clarity, without trying to explain everything in a single page, but trying to describe “who you are in a language that understands”. In particular, information on:
- Who I am (or who we work for).
- What do I do
- What are my relationships with other entities that Google knows (employers, family, schools, awards, etc.).
- What I have done before.
This is, in fact, a kind of curriculum vitae in a nutshell.
It is also important to add the schema.org markup to the Entity Home, useful to present key information in a machine-readable format, that is the “native language” of Google. In addition, it could also serve to “unambiguously indicate where Google can find the validation/confirmation it desperately needs”, as we will see later.
The validation/confirmation stage
The second key step is the validation and confirmation phase: as a child looks for other items that confirm the information obtained (parents, grandparents, siblings, teachers, local people identified as wise), so Google needs to find a corroboration of the data we provide it by finding “this same information from multiple trusted sources”.
That means we need to work to make sure that relevant and authoritative sources confirm what we’re communicating to Google through Entity Home – and Barnard himself, in the past, it had revealed which sources are reliable to Google in relation to knowledge panels – and consequently verify that “every page of the profile and relevant article on you” reports right and up to date news, and eventually work to correct them so as to be certain “that they confirm some or all of the factual statements reported on the Entity Home”.
This review is easy to do on places where we have good direct control, such as personal or corporate social profiles, but it can be harder if third parties have written about the person independently and we therefore need their collaboration for correction.
According to the article, “Google may have already figured out what the official profiles of the person are, but it is often not safe as there will be hundreds or perhaps thousands of profiles with the same name”, so it is “absolutely vital that there is consistency between these”, which must consistently aim at the entity’s home. On the other hand, however, Google is also able to understand that this “is simply a repetition from the source itself without independent arbitration“, so we need a further effort.
Third-party confirmation is in fact essential and is much stronger than autoconferma: correcting the contents on which we control is the basis, “but convincing the independent parties to add credibility is the key to connect the dots so that they do not break“.
Last step: an infinite and self-confirming cycle
Thinking like a human being, repetition can become annoying, but for a machine trying to understand the world “it is the only thing it craves for”.
The creation of the entity home (step 1) is useless and does not mean anything if Google has not understood and accepted this choice, and it is up to us to choose: we should instinctively understand which page is the Entity Home, and then bet on this best option and work to convince Google.
The search engine is looking for the most authoritative page in the web about the person, and generally prefers a page that is on a site owned by 100% of the person, while the other options (sites that we do not own 100%) require a greater effort.
How to make Google understand which one is the entity home
Barnard then describes the process of infinite cycle and auto-confirmations that we must set to force Google to approve our choice on the home of the entity: from this home, you have to link to corroborating sources (both previous and third party ones), and from these reinforcing sources link back to the Entity Home.
This is what the machine sees:
- CV on the home page of the entity concerned (first-hand news).
- Link to a dedicated page on another site.
- Confirmation of this information (more or less authoritative, more or less duplicated, but still confirms).
- Link back to the entity’s home page.
- (Re) confirmation of this information.
- Link to a dedicated page on another site.
- Confirmation of this information (more or less authoritative, more or less duplicated, but still confirmations).
- Link back to the entity’s home page.
- Starting over and over again.
This is a brute force educational method – not what we would give to a child, the author laughs off, “but Google is special” and, being a machine, it needs such a special process. If “you’re not famous, this is what works” to get a personal knowledge panel.
Avoiding Google choosing
According to Barnard, Google eventually attributes an Entity Home to the person, regardless of whether we instruct him or not: however, if we let this happen without intervening, we risk that “the choice is that of a confused child, based on fragmented and confusing information“.
For this, it is better to try to educate Google to understand what is the entity home we prefer, rather than leave it alone to try and bet on what is the “best hypothesis” (it would be, as in the famous game, pinning the tail to the donkey).
Why does this method work?
The meaning of this three-step process to activate an information sheet on Google lies in its simplicity: in practice, it gives the search engine “what it is actively looking for”, because what we do concretely is “educating the child” by providing information, explanations, indications for confirmation and centralisation.
The key is precisely this idea of centralizing information in a single place – the entity home – a concept that recently the aforementioned John Mueller had used in reference to the EAT signals that Google needs to define the quality behind the pages and their authors.
In this context, if Google “doesn’t understand who you are, it can’t apply EAT signals; to understand who you are (and therefore be sure when it applies those signals), it needs a reference point“, which can be precisely the entity home in our case.
The information are already online, scattered on the Web, probably spread on many pages; generally, they are fragmented, and therefore a pivot/hub/reference point is needed to connect the dots, that is to realize what Mueller calls reconciliation.
We can help the machine by showing exactly how to connect those dots, because the Entity Home simply confirms what it thought it knew, without being certain. In a situation in which the machine is confused and fails to clarify the facts in its own “mind”, the home will be the reference – the “crutch” – which provides you with the findings on the facts that must confirm through verification and corroboration.
How long it needs to activate a knowledge panel
If we effectively accomplish this process, we can activate an information sheet on the person in a few weeks or even days: only that the machine understands the key facts about an entity and that the entity’s owner can confirm the information through reliable sources.
However, in some cases this work may take longer, such as when there is:
- Immense ambiguity (especially for common names “like Simon Cox”, the author says).
- A lot of confused noise (fake news, more famous homonymous).
- Historical baggage (Google has already incorrectly connected the dots and needs to be re-educated).
However, according to Barnard there is no doubt about the “if” we can get a place in the Knowledge Graph, and therefore in the information sheet: it is rather a matter of “how much thought, time and commitment it will take to educate Google”.
Who can gain a knowledge panel on Google
Google’s Knowledge Graph has no guidelines for notability: it just wants to understand everything. Therefore, educating Google and getting a place in the Knowledge Graph is possible for everyone regardless of notability.
If executed correctly, the three-step process – Entity Home construction, confirmation and infinite loop – can push any person into the Knowledge Graph.
As the author says, “Google does not judge you, it simply wants to understand; and if you educate it correctly, it will understand”.
What the machine needs is that “a centralized and authoritative place on the Web” so that to reconcile the (honest) CV that we provide; and wait for us to present you an entity home that you recognize and trust, “a place where he obtains information directly from the source”.
In short, entering the Knowledge Graph is accessible for each entity, and once in the Knowledge Graph, an information sheet “exists” in Google: but the fact that Google actually shows the knowledge panel for a given user search query is another matter.
Therefore, we will see our knowledge panel based on certain circumstances, and in particular:
- Google’s level of understanding.
- Google’s trust in this understanding.
- Google’s assessment of the probability that the user’s intent is our name (often very geo-sensitive element).
