You have seen the artificial intelligence (AI) box at the top of Google, noticed your company is not in it, and maybe watched a competitor sit there instead. The question is whether that is fixable, and the answer starts with something most of the advice skips: Google's AI Overviews and Perplexity are built differently, reward different things, and should not be blurred into one generic "AI search" strategy. This is the platform-specific piece of a larger subject I cover in AI Search Visibility for Manufacturers, and the full picture is in How Modern Manufacturers Get Found.
The short version: Google's AI Overviews draw on Google's existing index, so conventional ranking still matters, plus an answer stated plainly enough to lift. Perplexity searches the live web and cites what it retrieves, which gives specific, quotable content a chance even against bigger competitors. The question of how to show up in Google AI Overviews, and in Perplexity, comes down to the same underlying work aimed at two different doors.
How does Google's AI Overview work?
Google's AI Overview draws from Google's existing search index, so conventional ranking still matters. By Google's own documentation, there are no additional requirements and no special optimizations for appearing in AI Overviews or AI Mode, and its generative AI guidance says these features are built on the same core Search ranking and quality systems as ordinary Search.
That is the known part, from the platform itself, and it answers the question "How does Google's AI Overview work" at the level anyone outside Google honestly can. The practical consequence follows from what an Overview is. It is a summary, so a page whose answer sits plainly near the top is easier to lift accurately than a page that ranks well but buries its answer under six paragraphs of company history. Strong search engine optimization (SEO) is necessary here, and it is not sufficient: ranking gets you into the pool the summary draws from, and answer clarity affects what is worth summarizing.
So if you have searched some version of how to optimize manufacturing website for AI search, the honest Google-side answer is unglamorous: do the ranking work you already should be doing, and structure your pages so the answer to the buyer's question is stated cleanly instead of implied.
How do you appear in Perplexity search?
Perplexity works differently by design. Its own documentation says it searches the internet in real time when asked a question, gathers information from sources it judges authoritative, summarizes, and cites each answer with numbered footnotes linking to the originals. It is retrieving live, not only drawing on an index position you built up over years.
What that design appears to reward is different from what Google rewards: specificity, recency, and precision quotable enough to survive being lifted into an answer. This is the part of the article that should interest a smaller manufacturer most. Genuinely specific technical content, the documented operating range, the named material, the answer to the exact question asked, can get cited in situations where the same company would never outrank a larger competitor in conventional results. Being outgunned on domain authority is not the end of the story on this platform.
Honesty about the limits: nobody outside these companies can specify how sources are chosen, the industry's term Perplexity optimization describes informed practice rather than settled knowledge, and Cobalt does not measure Perplexity citation. What is documented is the mechanism; what is observed is that specific, citable pages appear; the rest is inference, and I will label it that way.
How do you rank in AI search results?
The on-page work that serves both systems is the most actionable part of this subject: a direct answer near the top of the page, clean and descriptive headings, claims that can be quoted without distortion, and specific figures with named sources rather than vague assertions.
The underlying principle ties the list together. Both systems are assembling an answer from other people's pages, so both need material they can lift without misrepresenting it. Writing that is easy to quote accurately is the shared requirement. A sentence like "handles continuous duty at 400 degrees Fahrenheit" survives being quoted. A sentence like "engineered for the most demanding applications" gives a system nothing to work with, and leaves the opening to a competitor's more specific sentence. When owners ask some version of "How to rank in AI search results" or "How to improve answer engine optimization?", that list and that principle are the answer for any platform.
Worth as much as the list: what Google says you can skip. Since Google states no special optimization is needed for its AI features, anyone selling you a separate Google-AI-features package is selling work the platform itself says you do not need. The work above is ordinary good publishing, aimed carefully.
How does AI find businesses in the first place?
Beneath the page-level tactics sits entity-level groundwork: consistent company information everywhere it appears, clear authorship, corroboration across independent sources, and structured data that accurately identifies what a page describes.
The plain answer to "How does AI find businesses?" is that a system has to resolve scattered mentions into one confident identity before it can name you. Inconsistency works directly against that. A company whose name, location, and description vary across its website, directories, and industry listings is harder to resolve into a single entity, and a system that cannot resolve you confidently has a reason to name someone it can. This is also the cheapest problem in this article to fix, because you can audit it yourself in an afternoon: search your company name and check whether your site, your Google Business Profile, your industry directories, and your association listings agree on who you are, where you are, and what you make.
How do you get AI to recognize and cite your business?
Bring the halves together: the entity must be resolvable, and the content must be worth citing. Neither alone is enough. A perfectly consistent company with nothing substantive published gives the systems an identity with nothing to say. Deep content from a company whose basic information contradicts itself gives them something to say with no confident identity to attach it to.
Corroboration is where most manufacturers fall down. Independent sources saying what your site says is one of the factors these systems draw on, and it is the factor manufacturers most consistently lack. Authority signals are one of the four areas we score in the audits we run, and none of the manufacturers we have audited has scored strong on them. The expertise is usually real. The independent footprint confirming it usually is not there.
On expectations, I will tell you what governs the outcome rather than quote a timeline: how quickly your material gets indexed, how much depth you publish, and the pace at which corroboration accumulates from sources you do not control. Anyone quoting a fixed schedule is guessing on your behalf. On measurement, this platform is where Cobalt's actually lives: we measure citation within Google's AI Overviews across the full network of terms a manufacturer should own. No reliable measurement exists for Perplexity, so we do not claim results there. The same published, corroborated material serves both systems, and that is what we build to.
The Buyer Visibility Audit checks what comes back when buyers search the applications you serve — in conventional results and in Google's AI answers, alongside the competitors who appear instead of you. Free, one to two business days, scoring Buyer Visibility, Authority Signals, AI Readiness and Conversion Ready.
Frequently Asked Questions
How to get a company to show up in AI searches?
Work both halves at once. Make the company resolvable: consistent name, location, and description everywhere you appear, with structured data that matches the visible page. Then give the systems something worth using: substantive answers to the questions your buyers ask, stated plainly enough to quote. Companies that show up reliably tend to have both; companies missing either tend to be absent.
How to get AI to recognize your business?
Recognition is the entity half. Audit whether your basic information agrees across your website, your Google Business Profile, directories, and industry listings, then fix the disagreements. Add clear authorship so expertise attaches to named people, and keep your structured data honest. None of this requires a vendor; it requires an afternoon and some discipline.
How to get cited in AI search results?
Citation is the content half. Publish specific, verifiable material: figures with named sources, stated operating boundaries, direct answers positioned near the top of the page. Then build the independent footprint, because corroboration from sources you do not own is one of the factors these systems draw on, and it is the one most manufacturers lack. Precision is liftable; a slogan gives the system nothing to lift.
Editorial Note
I source from primary research where I can reach it, credible reporting of it where I can't, and leave out what I can't verify at all. Cobalt's own client figures include the measurement method and sample size.
The recommendations reflect my experience, not a settled answer. Search and buyer behavior keep changing, your situation isn't identical to my clients', and I still get things wrong. Verify anything that matters.
Find out where you stand
If you want to know what buyers and AI systems currently see when they look for a company like yours, the Buyer Visibility Audit will tell you. It is free, delivered in one to two business days, and scores four areas: Buyer Visibility, Authority Signals, AI Readiness, and Conversion Ready, including which competitors come up when buyers search the applications you serve. No obligation and no pitch attached to it. If what it shows is worth acting on, we can talk about what that would involve.
Request Your Free Buyer Visibility Audit →Free. No obligation. No pitch.