AI SEO
Get Mentioned and Cited When AI Answers Your Buyers
A share of your buyers now ask ChatGPT, Perplexity, or Google AI Overviews before they open a search box. Duomark works both sides: AEO, making your pages the easiest source to extract and cite, and GEO, growing the trusted off-site mentions those models draw on. Nobody controls an AI answer. What can be worked on is your odds of being the source it picks.
- AEO
- GEO
- AI Visibility Monitoring
kenapa brand saya tidak muncul di chatgpt
Why your brand is missing from AI answers
Before adding new pages, we check how answer engines read your brand today. These four patterns come up most often.
Ask about your service and the AI names competitors
Models answer from the sources that mention and cite a name most consistently, not from whoever has the better product. If your brand rarely appears in third-party articles, listings, reviews, and coverage, there is nothing for a model to pick you from. This is the GEO side of the work, and that presence can be built.
Your pages rank on Google, but the AI quotes someone else
Ranking and being cited are two different things. Answer engines lift passages that stand on their own: question-shaped headings, clear entity definitions, a complete answer in a single block, and clean structured data. Pages that read well for people are often not yet tidy for a machine to extract, and that is what AEO fixes.
The AI states something wrong or outdated about your business
Models assemble your entity from many sources at once: your site, Google Business Profile, directories, social profiles, even old articles. When the brand name, address, or service list disagree across them, the wrong version travels into the answer. The fix is boring but effective: one set of facts everywhere, with Organization schema as the anchor.
You have no idea whether you are mentioned, because nobody tracks it
AI answers vary by user and by session, and shift again when models update, so checking a prompt or two yourself tells you very little. We build a list of the prompts your buyers actually use, then track whether your brand is mentioned or cited across ChatGPT, Perplexity, Gemini, Claude, and AI Overviews, including which sources those answers pull from. Once it is measured, it becomes clear what to fix first.
beda ai seo dan seo tradisional
Traditional SEO or AI SEO?
Most businesses need both, not one instead of the other. Classic SEO is still the foundation: if your pages cannot be crawled, the content is unclear, and the brand has no authority, answer engines have nothing to work with.
| Criteria | Traditional SEOright when your priority is traffic and conversions from Google result pages | AI SEO (AEO + GEO)right when your buyers already ask ChatGPT, Perplexity, or AI Overviews first |
|---|---|---|
| Where you appear | On the search results page: blue links, the Map Pack, and featured snippets. | Inside the answer itself: ChatGPT, Perplexity, Gemini, Claude, and AI Overviews, as a brand that gets mentioned or a source that gets cited. |
| How the user finds you | Someone types a keyword, scans a list of results, then picks and clicks. | Someone writes a longer prompt, the model summarizes an answer, and only part of the readers click through to the source. |
| What wins the spot | Keyword relevance, content quality, backlinks, and technical health. | The most extractable content, plus facts that stay consistent and are corroborated by sources beyond your own site. |
| How it is measured | Keyword rankings, Search Console impressions and clicks, organic traffic, and conversions. | Share of voice per prompt: how often your brand is mentioned or cited, and which sources the answer pulls from. The numbers move per session and shift whenever a model updates. |
| What it is best for | Existing demand with clear volume, transactional pages, and local businesses chasing calls and visits. | Considered purchases, technical categories, and brands that want to be named when buyers ask for a recommendation. |
apa itu aeo dan geo
The Three Disciplines Behind AI SEO
Nobody controls an AI answer. Three things are actually in your hands: making pages easy to extract, strengthening your brand footprint off-site, and measuring what really shows up in answers.
AEO: Pages That Are Easy to Cite
Answer engines lift the clearest passage that can stand on its own. We set up entity definitions, question-shaped headings, and answer paragraphs that are complete in one block, then lock it in with structured data. All of it has to be crawlable and renderable without JavaScript getting in the way.
FAQPage, Article, Organization schema + question-shaped headings + entity definitions + render auditGEO: Off-Site Authority and Mentions
Generative models learn from many sources, not just your website. GEO works the outside: a consistent brand entity across the web, expert commentary from your team in the media, digital PR, third-party listings, and reviews. The more often the same facts appear on sources those models ingest, the better your odds of being named.
entity consistency + digital PR + expert commentary + third-party listings and reviewsAI Visibility Monitoring
We build the list of prompts your buyers actually use, then check them on a set schedule across ChatGPT, Perplexity, Gemini, Claude, and Google AI Overviews. We log whether your brand is mentioned, whether it is cited, and which sources the answer pulls from. Answers vary per user and change as models update, so what we read is the trend, not a single screenshot.
prompt set + scheduled checks + report on mentions, citations, and source domainsThe platforms we cover
Each engine picks its sources differently, so no single tactic works across all of them. These five are the platforms we track and optimize for, and results are reported per platform rather than merged into one number.
Google AI Overviews
The AI summary above Google's search results. Closest to classic SEO: pages that are crawlable, already indexed, and answer cleanly are the ones most likely to be cited.
ChatGPT
Widely used for early research before choosing a vendor. We track whether your brand gets named on category and comparison prompts, including when its web browsing is active.
Gemini
Tied closely to Google's ecosystem, so your entity consistency and Google Business Profile carry weight here.
Perplexity
An answer engine that always shows its sources, so it is the most transparent to measure. We log which of your pages get cited and on which prompts.
Claude
Often used for longer, analytical questions. What gets picked up here is a complete, self-contained explanation, not a promotional snippet.
cara kerja jasa ai seo
The monthly AI SEO cycle
These five stages run every month, not once. Each cycle adds pages that are easier to cite and new mentions on sources the models read, so the effect compounds.
01
Baseline the prompts and current AI visibility
We build the list of prompts your buyers actually use, then run them across ChatGPT, Perplexity, Gemini, Claude, and Google AI Overviews. We record where your brand is mentioned, where it is not, and which sources each engine pulls from.
at the end of this stage: a prioritized prompt list plus a baseline of mentions, citations, and source domains per engine.
02
Audit extractability and entity consistency
We check whether your pages can be crawled, rendered, and read as self-contained answers rather than paragraphs someone still has to interpret. We also check that your brand facts, name, address, services, and description, stay consistent on your site and on third party listings.
at the end of this stage: a per page extractability audit and a list of entity data mismatches to fix.
03
Restructure and mark up the pages (AEO)
We rewrite headings into the questions people actually ask, and make sure the paragraph under each one answers in full without needing outside context. Then we add the relevant structured data, FAQPage, Article, and Organization, so engines read the same facts your visitors do.
at the end of this stage: restructured priority pages, with schema deployed and validated.
04
Earn off-page mentions and citable assets (GEO)
Generative models draw on sources outside your site, so we pursue mentions on media, directories, and industry publications those models ingest. Alongside that we build assets worth citing, such as your own data, expert commentary from your team, or a technical explanation nobody else has published.
at the end of this stage: a log of the new mentions and placements that went live, plus the citable assets they point to.
05
Re-measure, report, then iterate
We rerun the same prompt set and compare it against the baseline, so the shift in mentions and citations since last month is visible. Nobody controls an AI answer, but the pattern of sources an engine leans on is readable, and that pattern sets the priorities for the next cycle.
at the end of this stage: a monthly report of visibility changes per engine and the work plan for the next cycle.
checklist aeo dan geo
What we actually do
Concrete work, not ranking promises. We cannot control an AI answer, what we can do is make you the easiest source to extract and the one most corroborated elsewhere.
- Rewrite headings into the questions people actually ask
- Write answer paragraphs that stand alone and can be quoted as is
- Add and validate FAQPage, Article, and Organization schema
- Make sure key content survives crawling and rendering
- Align your brand entity data across listings, profiles, and directories
- Pursue mentions and coverage on the sources models read
- Prepare expert commentary and quotes the press can use
- Build citable assets such as original data or technical guides
- Track your buyer prompt set across ChatGPT, Perplexity, Gemini, Claude, and AI Overviews
- Log which domains and pages each engine cites in its answers
- Compare your presence against competitors on the same prompts
- Report the month over month shift in mentions and citations, as it is
harga jasa ai seo
AI SEO packages
Every tier runs the same method: AEO on your site, GEO off your site, and AI visibility monitoring. What changes between tiers is only three things: how many prompts we track, how many pages we work on each cycle, and how far the GEO work goes.
A starting point to see where you stand in AI answers today, and to clean up your most important pages so answer engines can read and cite them.
- 15 buyer prompts tracked across ChatGPT, Perplexity, Gemini, and Google AI Overviews
- AEO pass on 5 priority pages per cycle
- Baseline structured data: Organization and FAQPage
- Entity definitions cleaned up and core business facts made consistent
- Monthly report: which prompts mention you and which sources the answers pull from
Our recommendation for most businesses: wider prompt coverage, AEO running on a regular cycle, and off-page GEO work starting.
- 40 prompts tracked, including category, comparison, and alternative prompts
- AEO pass on 12 pages per cycle
- Full structured data: Article, FAQPage, Organization, and Service or Product
- GEO begins: one citable asset per month plus outreach to relevant media and directories
- Competitor comparison: who gets cited when your brand does not appear
For brands that want broad prompt coverage and continuous AEO plus GEO work every month, with one strategist owning your account.
- 80 or more prompts tracked, covering your full category and its follow-up questions
- Continuous AEO: 25 pages per cycle
- Full GEO: regular citable assets, digital PR, expert commentary, and cleanup of third-party listings and reviews
- Dedicated strategist with a direct communication channel
- Quarterly deep review: content direction, entity health, and gaps in the sources AI cites
pertanyaan soal ai seo aeo dan geo
Common questions
can you guarantee my brand gets cited
No, and anyone promising that is worth avoiding. Nobody controls an AI answer: results differ per user, per session, and shift when models update. What we can work on is the odds, by making you the most extractable source and the one most corroborated elsewhere.
how is this measured
We fix a list of prompts your real buyers actually use, then run them on a regular schedule across ChatGPT, Perplexity, Gemini, Claude, and Google AI Overviews. We log whether your brand is mentioned, whether it is cited with a link, and which sources each answer pulled from. Because answers vary, what we read is the trend over time, not a single screenshot.
does this replace normal seo
No. AI SEO runs on top of your existing SEO, it does not replace it. Answer engines mostly pull from pages that are crawlable, indexed, and already trusted, so if your SEO foundation is weak, that gets fixed first.
how long until it shows
AEO changes can surface as soon as pages are recrawled, often within weeks. GEO is slower because mentions and citations accumulate gradually, usually over months. Keep in mind visibility can also drop back when models update, so this is ongoing work, not a one time fix.
does my business actually need this yet
Not necessarily. Check one simple thing first: whether your buyers actually ask AI before choosing, which tends to happen on considered, research-heavy purchases. If your site is still thin or technically unhealthy, your money does more good fixing basic SEO first.
what happens if the ai says something wrong about my business
Usually the model is just repeating what it found: an old page, an outdated listing, a stale article. We trace the sources, correct them at the source, then publish the right facts in a clear self-contained form and keep them consistent everywhere your brand appears. There is no edit button on the model side, so a correction takes time while the old sources get displaced.
Find out what AI says about your business
We can start with something simple: run a handful of the prompts your buyers actually use, then show you who gets mentioned today and which sources those answers cite. From there it becomes clear whether this work is worth starting.