How to Build an AI Visibility Strategy from Zero: The Playbook for Brands Starting Invisible

If your brand is invisible to ChatGPT, Perplexity, and Google AI Overviews right now, you are not behind. You are in broad company. Presenc AI's 2026 benchmark of 2,847 brands puts the cross-industry median AI visibility score at 49 out of 100, which Presenc reads as significant room for improvement for the average brand. A low score usually does not mean you executed badly. It usually means you have not started building the signal on purpose.

The playbook below starts from what the research says AI engines cite, then builds those signals deliberately rather than tweaking meta tags or buying links.

Why Starting from Zero Is Not the Disadvantage You Think It Is

The conventional wisdom says you need domain authority, years of content, and a massive backlink profile before AI engines will notice you. The data says otherwise.

In Ahrefs' analysis of 4 million AI Overview URLs, about 38% of cited pages also ranked in Google's top 10 for the same query, against about 76% in its July 2025 study (Ahrefs notes its methodology and the model behind AI Overviews changed, so the two are not like-for-like). Pages ranking 11 to 100 account for 31.2% of citations. Pages beyond rank 100 account for another 31%. That means more than 60% of the pages AI Overviews cite are not in the traditional top 10. The gatekeeping function of Google rankings is weakening, and that creates a real opening for brands starting from nothing.

Cyrus Shepard's May 2026 meta-analysis of 54 studies scored 23 features by how consistently they correlate with AI citations across ChatGPT, Gemini, and Perplexity, on a scale that runs from 2 to 9.5. Domain Authority scored 5.0. Schema scored 5.6. The highest-scoring features were URL accessibility (9.5), search rank (9.4), fan-out rank (9.3), preview control (9.2), and query-answer match (9.2). Shepard notes these are correlations across studies, not confirmed ranking factors. Three of those five are things a brand can usually address quickly. Taken at face value, the list says you do not need ten years of accumulated authority to be in the running: a crawlable site, clean preview controls, and content that directly answers the question line up with the strongest correlates.

Step 1: Audit Your AI Visibility Baseline Across Four Engines

You cannot build what you cannot see. Before spending a dollar on content or PR, you need to know exactly where you stand across ChatGPT, Perplexity, Gemini, and Claude.

Here is a manual audit you can run yourself. Pick 10 category queries that your buyers would type: not branded searches, but problem and comparison queries. "Best CRM for mid-market SaaS." "How to reduce churn in a subscription business." "What is the most reliable project management tool for remote teams." Run each query across all four AI engines. Record three things for each:

1. Presence: Did the engine mention your brand at all?
2. Position: Were you in the first cited source, in a list, or buried at the end?
3. Accuracy: Was the information about you correct?

Linksii's State of AI Search Visibility benchmark found that 104 of 200 tracked brands only appear when AI is asked about them directly. They are invisible on category, comparison, and problem queries. If your brand only shows up when someone types your exact name, you do not have visibility. You have a business card.

The Pondral AI Visibility Index breaks visibility into five scored factors: presence (71.7/100 average), prominence (57.6), context accuracy (66.2), citation link (25.7), and competitive presence (57.4). The weakest factor across 200 brands is Citation Link at 25.7 out of 100, which Pondral's current methodology defines as whether a graded answer cites a link to the brand's own domain (we did not check how its v2.0.3 run scored it). These are averages from one vendor's June 2026 single-run snapshot, which Pondral calls directional and says its methodology has since changed. In the same snapshot the Presence average was 71.7.

Track your baseline scores across all four engines and all 10 queries. This is your starting line. You will measure against it in 90 days.

Step 2: Fix the Technical Floor Before Anything Else

This is the step most brands skip. They jump straight to content strategy or PR campaigns while their site is blocking the very machines they want to cite them.

Shepard's analysis gives URL accessibility the highest score of any feature correlated with AI citations, 9.5 on his 2-to-9.5 scale. A page an AI system cannot reach cannot be cited from it. Check three things today:

1. robots.txt: Make sure you are not blocking ClaudeBot, GPTBot, PerplexityBot, or OAI-SearchBot. A single disallow line can keep an entire platform's crawler out.
2. Preview control: Shepard's list flags that preview controls such as a nosnippet directive can influence whether your content is used in AI answers. Audit every page template.
3. Crawl response: Your pages need to return clean 200 responses. Redirect chains, soft 404s, and interstitials break AI crawling just like they break Googlebot.

Semrush's US clickstream study found that 92 to 94% of Google AI Mode sessions ended without a visit to any external website. For those sessions the click is not the conversion event; the citation is. If the machine cannot read your page, it cannot cite you. This is the floor. Build on it or build on nothing.

Step 3: Earn Independent Brand Mentions (This Is the Lever)

Here is where the playbook diverges from everything you have been told about SEO.

Ahrefs studied 75,000 brands and found that branded web mentions correlate with AI Overview brand visibility at 0.664. Backlinks correlate at 0.218. That is a 3 to 1 ratio. In that data, the signal most strongly associated with brand visibility in AI Overviews is not your link profile. It is how often your brand is named across the web.

This is not a subtle difference. Ahrefs' own reading is that the signals tied to AI Overview brand visibility are rooted in text and language: language models learn from web text, so a brand named in text is evidence they can use whether or not a link comes with it. A plausible mechanism follows. When a credible publication writes "Company X's platform tracks AI citations across ChatGPT, Perplexity, and Gemini," the engine treats that as independent verification. It is evidence that the brand exists, operates in that category, and does that specific thing. A backlink proves a relationship between two pages. A brand mention proves a relationship between a brand and a concept in the mind of an independent author.

Evertune's first-quarter 2026 sampling of more than 108,000 product and category prompts on ChatGPT and Google AI Mode found that earned media accounts for about 32% of the domains AI models cite — a meaningful share, not a dominant one overall. The share varies sharply by topic: medical prompts averaged 58% on ChatGPT, and certain topics reached 87%. Where a category's trusted sources are editorial, earned media can carry most of the citations. The measured unit is source composition in the cited sample. It shows what engines cited, not why, and does not guarantee a future citation.

For a brand starting from zero in a category where editorial sources dominate the citations, the first investment is earned media rather than more owned content. Get mentioned by name in publications that AI engines already trust. Get quoted in industry analyses. Get cited in research. Independent mentions are the signal most strongly associated with AI Overview brand visibility in the Ahrefs data above, which is why they come before owned content in this plan.

Step 4: Build Structured Owned Content That Answers Specific Questions

Once you have earned media building your entity signal, the next layer is owned content that gives AI engines something to cite directly.

The key word is "specific." AI engines do not cite topic pages. They cite answers. Shepard's factor analysis scored query-answer match 9.2 on his 2-to-9.5 scale among the features correlated with AI citations. If your content directly answers a specific question that a buyer would ask, the engine has a reason to cite it. If your content covers a broad topic without answering any particular question, the engine has 50 other pages that do the same thing.

Here is the content structure that works for brands building from scratch:

Answer-first architecture. Put the direct answer in the first 40 to 60 words. Shepard scores important content placed near the top of the page 8.8 among the features correlated with citation. If your answer is buried in paragraph seven after three paragraphs of context-setting, you are working against that pattern.

Fan-out coverage. In the studies Shepard reviewed, ranking for related fan-out queries correlated with AI citation (9.3), as did ranking for multiple queries across a site (8.9). A brand that ranks across many variations of a buyer question is better placed than one ranking only for a single head term. Build content clusters: a pillar page that owns the primary query, surrounded by specific pages that own the sub-questions.

Named entities and specific data. Every claim needs a number or a name. "$1.32 billion" not "over a billion." "Ahrefs analyzed 75,000 brands" not "a study found." Shepard's list scores specific, verifiable facts 8.3 and specific claims over vague statements 8.1 among the features correlated with citation. The more concrete your content, the better its odds of being selected.

Structured formatting. Tables, numbered lists, comparison matrices, FAQ sections. These are not cosmetic. They are extraction patterns. Cyrus Shepard's meta-analysis of AI-citation studies scores AI-ready structure, content formatted so AI can easily extract it, 8.6 on his 2-to-9.5 scale among the features correlated with AI citations. Shepard's own caveat applies: these are correlations across studies, not proven causes.

Step 5: Target the Queries Where You Can Actually Win

A common mistake for brands starting from zero is to target the head terms. "Best CRM software." "Top project management tools." Those queries are dominated by brands with years of content and thousands of mentions. You will not win there on day one.

The head terms are where established brands are strongest, so the better aim is the long-tail problem queries, where Linksii's benchmark found 104 of 200 tracked brands never appear unless the AI is asked about them by name.

Linksii also reports a 71% visibility drop on its highest-intent, problem-framed queries for the brands it tracked, which suggests brands strong on "best CRM" can disappear on "how to reduce SaaS churn below 5%." Problem queries may be the white space where a brand with little existing visibility can get a foothold in AI answers.

Build a list of 20 to 30 problem queries your buyers would type. Check which ones have weak or missing AI answers. Start there. Own 10 problem queries before you compete on a single category query.

Step 6: Measure Four Metrics Every 30 Days

Evertune's playbook argues that cited domains change from month to month across major AI platforms, making visibility less stable than organic rankings (its 40% to 60% figure comes from a vendor roundup we could not trace to a primary, so we do not rely on the number). AI visibility is not a ranking you set and forget. Plan on monthly tracking.

The four metrics that matter:

1. AI share of voice. What percentage of responses for your target queries mention your brand versus competitors? This is the AI equivalent of market share. Track it across ChatGPT, Perplexity, Gemini, and Claude separately because they disagree. Superlines' tracking of a single brand (34,234 AI responses across 10 platforms, Jan 14-Feb 13, 2026) illustrates it: that brand's citation rate was 27.01% on Grok, 0.59% on ChatGPT and 0% on Claude. One brand's numbers, not a market rate.
2. Citation rate. The share of your tracked runs in which an AI engine cites your domain as a source, counted per run across the same questions and engines each time. Do not compare it with Pondral's 25.7 out of 100: that is a different unit, an average 0-to-100 component score for whether a graded answer links to a brand's own domain, from a June 2026 single-run snapshot, and a citation rate has no place on that scale.
3. Prompt coverage. What percentage of your target queries return a mention of your brand? If you are visible on 3 out of 10 queries, your coverage is 30%. The goal is to expand this systematically.
4. Accuracy. Is the information AI engines say about you correct? Inaccurate mentions can be worse than no mentions. Track and correct.

Run this audit on the same 10 queries you baselined in Step 1. Compare every 30 days. A practical 90-day cadence is: Days 1 through 30 for audit and initial fixes, Days 31 through 60 for early visibility signals, Days 61 through 90 for measurable share-of-voice shifts. Full strategic maturity takes three to six months.

The 90-Day Timeline: What to Do and When

Here is the compressed version:

Days 1 through 14: Technical cleanup and baseline audit. Remove robots.txt blocks on AI crawlers. Fix nosnippet directives. Run the 10-query baseline audit across four engines. Document your starting scores.

Days 15 through 30: Earned media sprint. Pitch three to five publications in your industry. Get quoted in industry roundups. Provide data or commentary for analyst reports. Independent mentions are the signal most associated with AI Overview brand visibility in the Ahrefs data.

Days 31 through 60: Content build. Publish five to eight answer-first articles targeting your highest-priority problem queries. Use the fan-out structure: one pillar, four to six sub-question pages. Every page should directly answer a specific query in its opening paragraph, cite primary sources, and link to your other relevant content.

Days 61 through 90: Measure, adjust, compound. Rerun the 10-query audit. Compare against your baseline. Double down on the queries where you gained visibility. For the queries where you are still invisible, diagnose whether the gap is earned media, content structure, or technical access.

Why This Works: The Compounding Effect of Entity Authority

Here is what most AI visibility guides miss. The individual tactics, whether they are earned media, content structure, or technical fixes, are not the strategy. The strategy is the compounding effect.

The working theory is a flywheel: earned media mentions build your entity signal, a stronger entity signal makes your owned content more likely to be cited, and each citation reinforces the signal for the next engine that processes a query about your category. The evidence above is correlational, so treat the loop as a theory to test against your own baseline, not a proven mechanism.

In G2's March 2026 survey of 1,076 B2B software buyers, 51% said they start research with an AI chatbot more often than with Google, up from 29% in G2's 2025 buyer report, and 69% said they chose a different vendor than initially planned because it was part of a chatbot's recommendation. Brands that are invisible to AI engines may be losing consideration they will never see, because the buyer can find the answer before visiting a website.

The case for starting now is that brands that build AI visibility early, even from zero, may reach positions that are harder for competitors to displace, while those that wait for AI visibility to "mature" risk learning late what they were missing.

Gartner predicted in February 2024 that by 2026 traditional search engine volume would drop 25%, with search marketing losing market share to AI chatbots and other virtual agents. Attention is moving. And the brands that move with it, starting today from wherever they are, are the ones that will be in the answer.

This is Machine Relations: the discipline of earning AI citations and recommendations for a brand by making that brand legible, retrievable, and credible inside AI-driven discovery. It works the same way whether you are starting from zero or reinforcing an existing position. The only variable is when you start.

FAQ

How long does it take to build AI visibility from zero?

Expect 30 days for your first measurable signals and 90 days for meaningful share-of-voice shifts. Full strategic maturity takes three to six months. This timeline is this page's own planning guidance, not a measured benchmark. The variance depends on how aggressively you invest in earned media, which is the fastest lever.

Is it possible to rank in AI answers without a strong domain authority?

Yes, on the evidence available: it is not a gate. Domain Authority scored 5.0 on a 2-to-9.5 scale in Cyrus Shepard's meta-analysis of features correlated with AI citations, and in the Ahrefs data above more than 60% of AI Overview citations come from pages outside the traditional top 10. Higher-scoring correlates in the same list include URL accessibility (9.5) and query-answer match (9.2), and Ahrefs finds independent brand mentions correlate with AI Overview brand visibility at 0.664.

What is the most important first step for an invisible brand?

Fix the technical floor. Check that AI crawlers (GPTBot, ClaudeBot, PerplexityBot) are not blocked by your robots.txt or nosnippet directives. This takes 30 minutes and costs nothing. If the machine cannot reach your pages, no amount of content or PR will make you visible.

How much should a company invest in earned media versus owned content?

For brands starting from zero, the ratio should skew heavily toward earned media in the first 60 days. Branded web mentions correlate with AI Overview brand visibility at 0.664 versus 0.218 for backlinks. Earned media is a plausible way to build the entity signal that supports your owned content; the correlation is the evidence, and whether it causes citations is untested.

Do I need specialized AI visibility tools to get started?

Not on day one. The manual 10-query audit across ChatGPT, Perplexity, Gemini, and Claude gives you a baseline that no tool can replace. As you scale past 30 queries and need monthly tracking, platforms like Evertune, Presenc, and Pondral offer automated monitoring across multiple engines.