At Hostinger, we spent years mastering the normal search engine optimisation playbook: rank excessive in Google and get clicks. However now ChatGPT and AI Overviews have modified the foundations.
Instantly, customers weren’t simply clicking the primary hyperlink. They had been getting their solutions immediately from AI-generated summaries. This was a significant shift in search conduct. Even if you happen to ranked #1, the clicks and the visitors weren’t assured anymore.
However that was solely half the issue. We additionally confronted a extra basic problem: we not totally managed our personal story.
That is the core problem of managing your model’s fame. AI methods like ChatGPT, Perplexity, and Google’s AI Overviews resolve what to say about you, however how they make these choices isn’t apparent.
And this issues extra every single day. OpenAI reviews that ChatGPT has 700 million weekly lively customers, representing a 400% improve from the earlier 12 months. Individuals are utilizing it for a mean of 16 minutes per day, which is getting very near the 18.2 minutes per day seen on main search apps.
We realized conventional search engine optimisation alone was not sufficient. We wanted a brand new technique to make sure AI highlighted our strengths, not outdated information or competitor speaking factors, and switch this new AI-driven world into a bonus.
That is the story of how we turned this problem right into a predictable benefit, and we’re sharing our full playbook so you are able to do the identical.
Subscribe For extra instructional movies!
Hostinger Academy
The “aha” second that constructed our playbook
At first, we had been flying blind. We might see what AI was saying about us, however not why. We’d see a competitor cited for a function or outdated details about us repeated, and we had no concept how they earned that spot.
So, we explored over 10 AI monitoring instruments and spent numerous hours analyzing prompts and responses.
The breakthrough got here once we understood that AI doesn’t simply learn a web site. It learns from information articles, accomplice blogs, Reddit threads, rivals, and even video transcripts. It’s constructing an image of a model by evaluating authority throughout all these sources.
We realized it’s extra like a pupil doing analysis, and it has its favourite sources. It doesn’t belief everybody equally.
And we discovered it constantly trusts three distinct forms of sources:
- Earned authority: Belief from high-credibility media sources.
- Borrowed authority: Belief from established companions, associates, and business specialists.
- Neighborhood authority: Belief from genuine conversations on platforms like Reddit.
This was the turning level. It reworked an unsolvable drawback (How can we affect AI?) right into a sensible, solvable one (How can we get our message into the sources AI already trusts?).
It gave us clear motion steps. As a substitute of shouting into the void, we might focus our efforts precisely the place they’d rely.
Contained in the playbook: Our three core methods in motion
We instantly translated this perception right into a tactical plan to construct earned authority, borrow authority, and develop group authority.
Right here is the precise framework we used to affect every of those three authority varieties, damaged down into sensible steps you possibly can take.
Technique 1: Constructing earned authority
This technique focuses on getting our model talked about by high-credibility business publications and information websites. We realized that to affect AI, this will’t be a “spray and pray” PR effort – it requires precision.
Tactic 1: Deal with high quality over amount
We rapidly realized that one point out on a website that AI deeply trusts is price greater than 100 small ones.
So, what are these extremely trusted sources, and the way do you get featured on them? An awesome place to begin is with OpenAI’s content partners. This record consists of main journalistic powerhouses similar to Condé Nast, The Guardian, and The Monetary Instances.
To get their consideration, you possibly can’t simply ship an ordinary press launch or slide into their DMs. You have to provide them data that’s distinctive and genuinely informative – one thing they will’t get anyplace else.
They’re in search of unique insights, professional commentary, new business evaluation, or information research that present actual worth to their viewers.
The way to implement this:
Don’t simply repackage what’s already on the market. As a substitute, create content material that gives actual worth, similar to:
- Unique analysis: Survey 100 of your clients a few scorching subject and publish the outcomes.
- Information insights: Analyze your individual inner information to disclose a brand new pattern. Present this information for others to quote, ensuring to guard buyer names and private particulars.
- Knowledgeable commentary: Compile your high 5 predictions in your business, with clear reasoning, and provide them as an professional quote.
Then, use instruments like Assist a Reporter Out, Muck Rack, or Prowly to seek out journalists who cowl your business and provide them your distinctive information or professional commentary.
In the event you’re on a finances, you are able to do this free of charge on LinkedIn. Seek for:
“[Industry] journalist”
“[Topic] editor”
Discover somebody who not too long ago lined the same subject. Ship a connection request with a word like this:
“Hello [Name], I noticed your article on [topic]. We simply discovered some attention-grabbing information on [related topic] that your readers would possibly worth. No strings connected, comfortable to share.”
Tactic 2: Grow to be your individual fact-checker
AI learns from the whole web, so it can inevitably encounter buyer complaints, outdated information, and any inaccurate data floating round. If left uncorrected, this information can develop into the “reality” that AI repeats to hundreds of thousands, immediately harming your model’s fame.
This meant we would have liked to shift from passive monitoring to lively fame administration. Whereas reporting errors to AI platforms is an possibility, the simplest technique is to manage the narrative.
You obtain this by turning into the supply of factual data that’s straightforward for AI to crawl, making you the go-to reference for correct details about your model.
The way to implement this:
This works even if you happen to don’t have a giant PR group. Step one is to actively monitor what AI says about you. You are able to do this by frequently checking outcomes for prompts your clients would truly use, for instance, “How good is [Your brand] for [X purpose]?” or “What are the cons of [Your brand]?”
Once you discover inaccuracies – just like the AI claiming you don’t have a market while you do – it’s essential right the file publicly.
Essentially the most direct manner is to replace your web site or publish a brief weblog put up clarifying the reality. AI methods usually pull details immediately out of your website, so conserving that model messaging constant and correct actually issues.
However what if the criticism is truthful? This can be a large alternative. If AI highlights an actual problem, like gradual response instances or a web site bug, your first job is to repair it.
After you have, make that repair a part of your content material technique. Share the way you improved. You could possibly create a brief video, write a data-backed social put up, or create a easy buyer story.
This turns a public criticism into a robust, clear story about your responsiveness. You don’t want a PR machine for that – simply honesty, consistency, and visibility.
Technique 2: Use borrowed authority by means of companions
This technique includes utilizing the belief AI already has in your companions, associates, and business friends. By having trusted third-party websites endorse you, like assessment websites and professional bloggers, you “borrow” their authority, which is a robust sign to AI.
Tactic 1: Empower your present associates and companions
The best place to begin is with the companions who already speak about you. We found our associates’ content material, whereas authoritative, was constructed for human readers, not machines. It wasn’t structured in a manner that allowed an AI to parse and quote simply.
This tactic includes empowering your companions with clear pointers to make their content material AI-friendly, utilizing their present authority as a direct asset in your model.
The way to implement this:
We turned our pointers into an AI-ready playbook for our associates. You are able to do the identical by sharing this straightforward roadmap together with your companions:
Use question-based headings: Ask direct questions in your H2 and H3 tags. AI prefers this semantic format.
✅ Good: What’s [feature]?
✅ Good: How [feature] works for small companies
❌ Unhealthy: Understanding [feature]
❌ Unhealthy: [feature] defined
Reply first, clarify second: Begin the part with a direct, factual reply, then increase on it.
For instance, think about your heading is the direct query: “What’s vibe coding?”
Your very first sentence, proper beneath that heading, ought to be the concise reply: “Vibe coding is a software program improvement technique that makes use of AI-based instruments to hurry up prototyping and help creativity.”
Then, within the subsequent paragraph, you possibly can present the extra context and rationalization: “This technique differs from conventional improvement by specializing in… [expansion follows here].”
Add credible metadata: Guarantee each article has a transparent writer title and an “Up to date: [Date]” on the high to construct credibility.
Use AI-readable buildings: Use easy HTML tables for comparisons (not pictures of tables) and bulleted lists for options.
Hyperlink strategically: Use descriptive anchor textual content, like “use circumstances of [Product] in design”, as an alternative of imprecise anchors like “click on right here” or “learn extra”. You must also place hyperlinks the place they take advantage of sense. As a substitute of grouping them within the intro, put a hyperlink proper within the paragraph the place you’re discussing that subject. This helps AI perceive why the 2 pages are associated.
Tactic 2: Recruit the winners
We determined to seek out out which web sites AI had already cited most frequently for our matters, after which made them our new partnership targets.
That is highly effective as a result of AI has already recognized these websites as authorities, so any point out or partnership with them provides your model a big, pre-vetted credibility enhance.
We name this the two% rule: if a supply is cited in 2% or extra of AI solutions for a subject, they’re a high-value goal.
The way to implement this:
Essentially the most direct approach to get this information is through the use of a devoted AI monitoring software similar to Profound. It lets you analyze your goal key phrases to generate a visible quotation share report, the place you possibly can determine the domains AI trusts most.
You can even do that manually free of charge:
- Go to Google and sort in your essential key phrases. See which internet sites are linked within the AI Overviews.
- Go to ChatGPT or Perplexity, ask round 10 completely different questions on your subject, and log which web sites get cited most frequently.
That is your new goal record for partnerships. Deal with the highest 2%. In the event you analyze 100 responses and Website A is cited 5 instances (5%), they’re a first-rate goal. Website B, cited as soon as (1%), will not be. This separates one-hit wonders from true authorities.
After you have your record of “2% winners,” contact them strategically. These websites are already profitable, so don’t method them with a chilly pitch asking for a hyperlink.
As a substitute, provide to assist lock of their authority with content material you create collectively that’s fully distinctive.
Listed below are some content material concepts:
- A co-branded information research, similar to a developments report or a “State of [the industry]” report.
- An unique efficiency report primarily based in your inner testing.
- Completely structured comparability tables they will simply drop into their high assessment pages
- A set of customized statistics derived out of your inner gross sales or help information.
The purpose is to counsel matters the place you present distinctive worth and may create a supply of knowledge that no person else can create.
This frames the collaboration as a win-win: you assist them reinforce their authority, and also you safe a high-value point out in your model.
This technique includes collaborating in conversations the place sincere, unfiltered discussions about your model are already going down.
AI platforms pull a large quantity of knowledge from on-line boards like Reddit as a result of they’re a supply of genuine human opinions and experiences.
That is the place AI learns what actual individuals assume, what their issues are, and the way they examine merchandise, making it a essential place to handle your model’s narrative.
Tactic 1: Have interaction immediately in AI-cited threads
Discover the precise threads AI is already citing (or new, related threads) and add genuinely useful, non-promotional feedback.
The aim is to not promote, however to be a useful professional. Since AI learns from these conversations, your genuine participation can immediately add context, make clear a function, or right misinformation on the supply, influencing the narrative AI will use sooner or later.
The way to implement this:
Use superior Google search operators to seek out related threads, similar to:
website:reddit.com “your model title”
website:reddit.com “[competitor name]”
website:reddit.com “[your topic]” “assist”
In your responses, be genuine. Reddit customers famously dislike inauthentic model promotion. Your aim is to be a useful member of the group, not a strolling commercial.
Construct an actual person profile by collaborating in discussions and sharing distinctive insights. This builds belief with each the group and the AI fashions that learn these threads.
Unhealthy Remark: I completely agree, that’s why our group at [Brand name] constructed a software to unravel this precise drawback. You’ll be able to test it out on our web site: https://www.hostinger.com/tutorials/ai-seo-playbook-brand
Good Remark: As somebody from the [Brand] group, I can make clear how function X works. It was designed to unravel [the problem user mentioned]. Right here’s the way it works: [more details]
Tactic 2: Shut the loop on suggestions
That is our strongest tactic for turning a public criticism right into a story about responsiveness. Once we discover a person criticism, we observe it, repair it, after which we return to the unique thread to answer.
That is extremely impactful for AI. As a substitute of simply studying “Model X has an issue,” the AI now learns the whole, constructive story: “Model X listens to clients, engages with them, and fixes its points.”
It transforms a unfavourable information level right into a public file of your trustworthiness, multi function public thread.
The way to implement this:
The method begins by treating complaints as actionable suggestions. The hot button is to create a easy monitoring system, like a spreadsheet. Once you discover a criticism, your first motion ought to be to log the hyperlink to the unique thread.
Share this suggestions internally and use it to make actual enhancements. That is essentially the most essential half: you’re turning a public criticism right into a constructive case research in your model’s responsiveness.
When you resolve the problem, the loop is closed by going again to that unique thread – even when it’s months later. Publish an replace to point out you adopted by means of. A easy reply just like the one under works completely:
“Hello [Username], circling again on this. Our group took your suggestions about [the issue] significantly, and we’ve simply carried out a repair. You’ll be able to see the replace right here: [Link]. Thanks once more for serving to us enhance!”
Constructing a resilient model that thrives with AI
So, what’s the massive takeaway? Getting AI to say the precise issues about you isn’t a few single trick. It’s about constructing a stable, reliable fame in all of the locations AI is already in search of solutions.
This isn’t simply an search engine optimisation job. It implies that all of your advertising efforts must be aligned, working collectively to make sure AI acknowledges and trusts your model as the very best supply of knowledge.
By focusing in your earned authority by means of media mentions, borrowed authority with companions, and group authority on boards like Reddit, you cease reacting to what AI says and begin informing its subsequent reply.
In our following playbook, we’ll present you the right way to take the lead and have your content material cited immediately as the reply. We’ll dive into our blueprint for content engineering for AI SEO.
The entire tutorial content material on this web site is topic to
Hostinger’s rigorous editorial standards and values.
Simon is a dynamic Content material Author who loves serving to individuals remodel their inventive concepts into thriving companies. With intensive advertising expertise, he always strives to attach the precise message with the precise viewers. In his spare time, Simon enjoys lengthy runs, nurturing his chilli vegetation, and climbing by means of forests. Comply with him on LinkedIn.
Deyimar Albornoz, Natural Advertising and marketing Strategist at Hostinger, has 10 years of expertise within the search engine optimisation area. She makes a speciality of creating and scaling natural progress methods that combine strategic planning, cross-team collaboration, and data-driven execution. All through her profession, Deyimar has centered on bettering visibility, visitors, and buyer acquisition by means of high-impact, non-paid channels.
