“People also search for” is Google’s way of telling a searcher which other people, companies, or queries tend to travel with the thing they just looked up. It shows up in two places. On a knowledge panel, it is a row of related entities (other founders, rival brands, co-stars, spouses). Under an organic result, it is a cluster of follow-up queries that appears when someone clicks a link, comes back to the results page, and Google guesses they did not find what they wanted. In both cases, the people also search for module is a public statement about who Google thinks you belong with, and you did not write it.
That last part is why founders email me about it. Someone Googles their own name before a funding round and finds a competitor, a former cofounder they parted with badly, or a stranger with a similar name sitting in a neat row of headshots beside them. The question is always the same: why is that person there, and can I get them out?
The short answer is that you cannot delete the box, but you can change what feeds it. The longer answer is the rest of this post.
What the box is, in both of its forms
Google has been running the entity version for well over a decade. On October 19, 2012, the official Inside Search blog announced that “People also search for” in Knowledge Graph panels now came “with explanations.” TechCrunch covered the change the same day: hover over an item and Google would tell you why it was there. Their example was Gone With the Wind, where the panel pointed out that the film’s director had also directed other movies in the row. The feature covered actors, films, TV shows, and family ties between famous people, and Google said it would only show an explanation when an interesting connection existed.
That history tells you what the entity version is built on: the Knowledge Graph. The items in the row are entities (things Google has a stable identity for), and the row is a set of edges in a graph: same director, same industry, married to, competes with, often searched in the same session.
The second form arrived later and behaves differently. SISTRIX, the German search analytics firm, documented it as the “short click box.” According to their write-up, it appeared on mobile from October 2016 and Google rolled it out to desktop in all countries in February 2018. It shows up directly beneath a result after a user clicks that result and returns to the results page quickly, a pattern SEOs call a short click or return-to-SERP. SISTRIX found the box only on page one, not under every result, and noted that the suggested queries were specific to the result clicked, distinct from the generic “Searches related to” list at the bottom of the page. Search Engine Land later reported a redesigned version on desktop where the suggestions either load after a short delay or appear on that same click-and-return.
So there are two distinct systems wearing the same label. One is entity-to-entity. The other is query-to-query, triggered by behavior. When a client says “Google is showing my competitor next to me,” my first question is which version they mean, because the fixes differ.
Google also keeps reshuffling the furniture around these modules. In 2022 Google began testing a “Things to know” card that groups subtopics of a query, and Digital Information World reported it appearing on roughly 1% of tracked results that June while still in testing. AI Overviews now sit above much of the classic results page for many queries. None of that retired the related-entities row. It means the row is one of several places where Google publishes its opinion about your neighborhood, and the same inputs tend to feed all of them.
Why does Google put other people next to your name?
Google does not publish the recipe for either version, so be wary of anyone who claims to know the weights. What Google does say, on its own Knowledge Panel help page, is that knowledge panels are “created automatically,” that the facts in them come from public sources, licensed data, and content owners, and that Google itself cannot directly decide when a panel appears. Read that with the 2012 explanation feature in mind and a reasonable model falls out.
Two kinds of signal are in play. The first is structural: the web describes you in relation to other entities. Your Crunchbase profile lists cofounders. A Wikipedia article (if you have one) names your company and your predecessors. Trade coverage says “X, a rival to Y.” Every one of those sentences is a candidate edge in the graph. The second is behavioral: in aggregate, people who search for you also search for someone else, in the same session or close to it. If a podcast host interviews you and your biggest competitor in back-to-back episodes, a slice of their audience will Google both of you that week. If you share a name with a local news figure, people searching for them will land on you and vice versa.
The person in your row is there because one or both of those signals point at them. Often it is both. A former cofounder shows up because old articles name you together and because investors who search one of you search the other. A competitor shows up because analysts write comparison pieces and because buyers search both brands during evaluation.
This is not Google passing judgment. It is Google reflecting the company you keep in text and in search behavior, which moves the question from “how do I complain to Google” to “what is the web saying about who I am next to.”
The Three Neighbors Model
When I audit a people also search for problem, I sort every name or query in the box into one of three neighbor types. I call it the Three Neighbors Model, and it decides what you do next.
The first is the Peer Neighbor. This is someone who belongs in your row because you share a category, and their presence helps a searcher understand who you are. A fintech founder appearing next to two other fintech founders is a Peer Neighbor relationship. It is usually fine, and sometimes it is flattering. If you are a newer brand and Google puts you next to the category leader, you have been handed a seat you wanted.
The second is the Shadow Neighbor. This is someone who is there because of a specific story: a lawsuit, a split, a controversy, a name collision. Shadow Neighbors are attached to you by a narrow set of documents and a burst of search interest. They feel permanent, but they are often the most movable, because they depend on a limited pool of pages and a spike in curiosity that fades unless something feeds it.
The third is the Missing Neighbor. This is the entity that should be in your row but is not. Your company. Your book. The well-known investor who backed you. The award you won. When the people you want to be associated with do not appear, it usually means the web has not stated the connection clearly enough for Google to treat it as an edge, or searchers are not making the connection in their own queries.
Most people notice Shadow Neighbors first, because those sting. The work that changes the box is mostly about Missing Neighbors. You seldom push someone out of the row. You crowd them out by making better connections louder.
Audit your own neighbors this week
Start in a clean browser window, signed out, and search your name exactly as a stranger would. Then search it with your company name, and again with your title. Note every entity in the knowledge panel row (if you have a panel) and every query in any short-click box you can trigger by clicking a result and coming back. Do the same on mobile, since layouts differ. Screenshot everything with the date, because you will want a baseline.
Now label each entry with its neighbor type. For every Shadow Neighbor, find the documents that connect you. Search both names together and read the first two pages of results. You are looking for the specific articles, profiles, and pages where the two of you appear in the same sentence. There are fewer than you fear. For every Missing Neighbor, search the pairing you want (“your name” plus “your company,” or plus “the investor”) and see how much clear, crawlable text states the relationship. Often the answer is a LinkedIn line and little else.
Last, check the structured sources Google draws on. Look at your Wikidata entry if one exists, your Crunchbase and company pages, your speaker bios, and the About page on your own site. Write down every place where the description of you is thin, stale, or wrong. An executive whose only detailed bio online is from three jobs ago will keep being associated with that old company and its people.
Shape the box without trying to game it
Do not try to manipulate the behavioral signal. Paying people to search pairs of names, or running click schemes, can blow up in public, and it fights a system built on aggregate behavior across many searchers. The durable approach works on the structural side and lets behavior follow.
Begin with your owned properties. Your site’s About or leadership page should state, in plain sentences, who you are, what company you lead, what you founded before, and who you work with. Add Person and Organization schema markup with sameAs links to your verified profiles so that every page points to one consistent identity. If your knowledge panel is eligible, claim it through Google’s verification process. Google says verified representatives can suggest changes to panel facts. That will not let you edit the related row, but cleaner facts give the graph a sharper entity to connect.
Then work on earned coverage, which is where a Missing Neighbor becomes a Peer Neighbor. A profile in a trade publication that names you alongside the company you built and the category you lead creates the kind of co-occurrence the graph is made of. A podcast appearance next to a respected peer creates a new pairing in both text (show notes, transcripts) and behavior (listeners searching both names). A contributed article under your byline on a credible site adds another page where your name sits beside your expertise instead of beside an old dispute. This is the core of what we do with AEO work, because the same associations that shape the people also search for row also shape how AI answer engines describe you.
For Shadow Neighbors, starve the story. Do not issue a press release about the old conflict. Do not post about it. Every new mention renews the pairing. If an article is inaccurate, ask the publisher for a correction, since a corrected page is better than an untouched one. If the association is a pure name collision, add disambiguating detail everywhere you control (middle initial, city, company) so that pages about you are clearly about you.
What can’t you control?
You cannot control which entities Google chooses to show, how many it shows, or when the layout changes. You cannot remove a legitimate association that the public record supports. If you and a competitor are the two names in your category, you will appear together, and trying to fight that is a waste of budget that should go toward being the better-described of the two.
You also cannot control timing. The box reflects crawled content and search behavior, and both move at their own pace. A big news moment can reshuffle it within days. Steady repositioning through coverage takes longer. Google’s help documentation does offer a Feedback link at the bottom of every knowledge panel, and you should use it for anything wrong or offensive, but think of feedback as a report, not a request that will be granted.
What you do control is the volume and clarity of the sentences that connect you to the right neighbors. That is a PR problem as much as an SEO one, and it rewards people who treat their name as an entity that needs a consistent story across every page that mentions it.
Here is the next step. Open a signed-out browser today, run the three searches above, and screenshot the row. Label each entry Peer, Shadow, or Missing. Then pick the single Missing Neighbor that matters most to your next twelve months (the company you are raising for, the category you want to own) and get one piece of earned coverage that names you together in the first paragraph. Rerun the screenshot in 30 days and compare.