Generative engine optimization

Generative engine optimization, done for you.

A generative engine does not return a list; it writes an answer. Ask ChatGPT which accounting firm to use, ask Perplexity for the best running shoe for flat feet, ask Gemini who makes the most reliable CRM for a ten-person team, and you get a paragraph with a few names in it and a row of citations underneath. The businesses in that paragraph were chosen because the engine found them, in sources it trusted, stated in a way it could use.

Generative engine optimization (GEO) is the work of becoming one of those sources. SEO Stuff does it as a fixed-price service: a baseline of what the engines say about your category today, the pages they will quote tomorrow, the placements that corroborate them, and an after-report that shows what moved.

Get a free SEO + AI search audit →See the Done-For-You Package

What is generative engine optimization (GEO)?

Generative engine optimization is the practice of increasing how often, and how favourably, a business appears in answers written by generative AI engines such as ChatGPT, Perplexity, Gemini and Claude. The term comes from a 2023 research paper that tested which page changes raised visibility in generated answers; the changes that worked were adding statistics, quotations and citations, while stuffing keywords did little or nothing.

How a generative engine picks its sources

When you ask a question with web search on, the engine first rewrites it into a few searches of its own, often more specific than what you typed. It reads the top results for each, discards what it cannot use, and composes an answer from the rest, citing the pages it drew on. Three habits matter for a business trying to be cited. The engine favours pages that answer directly, because it is looking for sentences it can lift. It favours agreement, so a claim that appears on your site and on two independent sites beats a claim that appears only on yours. And it favours sources it has learned to trust for a topic: review sites, trade publications, community threads, and well-maintained company pages with plain facts.

The research that coined the term measured this. Adding statistics, quotations and citations to a page raised its visibility in generated answers by up to 40% on the authors' benchmark, and keyword stuffing, the reflex of old SEO, did not help. That matches what we see in client baselines: the pages that get cited are the ones with numbers in them, a clear position, and sources of their own.

What we optimize, and what we leave alone

Prompt research: the twenty questions buyers in your category ask the engines, found by asking the engines and reading what they cite today
Pages written to be lifted: a direct answer in the first paragraph, specific numbers, named trade-offs, and a date
A facts page about your business (what you do, who it is for, what it costs, where you operate) so the engines stop guessing
Third-party corroboration: 3 DR50+ placements on sites already cited in your category
Crawler access, audited: whether your robots rules let OAI-SearchBot, PerplexityBot and Google's crawlers read the site, and whether an llms.txt file and schema markup are in place, with the fixes in the strategy guide
Nothing manipulative: no fake reviews, no hidden text, no mass-produced pages. Engines are good at spotting it and it reverses

How the work is done

1
Baseline the engines
Twenty buyer questions asked of ChatGPT, Gemini, Perplexity and Claude with web search on. Every answer and every citation goes on your dashboard, with who was recommended instead of you.
2
Find the gaps
A manual audit of your site against the sources the engines cited, and a strategy guide naming the questions you can own and what it takes. Within 7 business days.
3
Publish the pages
Ten pages in the shapes the engines quote, each built around one question from the research, each carrying the statistics, quotations and citations the engines reward.
4
Earn the corroboration
Three DR50+ placements where the engines already look for your category, so that what your pages say is backed by somebody else.
5
Re-run the baseline
The same twenty questions, after delivery, side by side with the first run. For ongoing weekly checks across fifty questions, the same team runs Rightcited.

GEO by business type

Generative engines ask different things of different businesses. These pages cover what the prompts look like in each category and what we build for it.

SaaS companies →

Engines answer "best X for Y" with the same comparison pages and review sites every time. The job is to be in those sources.

Ecommerce stores →

Product recommendations draw on buying guides, reviews and structured product data. We write the guides and earn the reviews.

B2B service firms →

"Who should we hire" is answered from case studies, trade coverage and plain pricing, which most firms have never published.

Small businesses →

A niche or local business can own its category's prompts outright, because almost nobody else has written for them.

Agencies →

Baselines and prompt research for every client on one credit pack, with done-for-you delivery you can put your name on.

GEO vs SEO, and GEO vs AEO

GEO vs SEO: SEO earns a position in a ranked list; GEO earns a place in a written answer. They share most of their foundations (authority, crawlability, content that answers the question), but GEO puts more weight on corroboration across sources and on the shape of the writing, and less on exact-match keywords. You cannot do GEO well on a site with weak SEO, and strong SEO alone does not get you cited.

GEO vs AEO: answer engine optimization is the older term and it started with Google's featured snippets and voice assistants. Today AEO usually means Google's AI Overviews and AI Mode, where the answer is extracted from a page, while GEO means the chat engines, where the answer is composed from several. The techniques overlap almost entirely, which is why SEO Stuff delivers them as one package rather than two.

Which package fits

The Done-For-You Package is the GEO service. The bundle adds volume once your authority is in place.

Done-For-You Package

Audit and strategy guide within 7 business days, 10 pages of content written to be cited, 3 DR50+ placements, 3,000 research credits and the dashboard. $999 once, delivered in 21 business days.

See the package

Premium Content Bundle

60 articles optimized for Google and AI search, with keyword and intent mapping, delivered in weekly batches over 4 weeks. $899 once. Best once your authority is in place.

See the bundle

Not sure yet?

The free SEO + AI search audit shows where you stand today in Google and in the assistants, and what to fix first. Manual, with a report within 48 hours.

Get the free audit

What the work has done in the generative engines

Before-and-after citation counts from clients using the same method.

24×

ChatGPT citations

A precious-metals client from 38 to 920 ChatGPT citations, with Copilot up from 27 to 332 and Gemini from 16 to 153.

457×

Perplexity citations

A telehealth client from 7 to 3,200 Perplexity citations, with Google AI Overview citations from 0 to 823.

271K

AI-search visitors in 6 months

A SaaS client cited across ChatGPT, Claude, Perplexity and Gemini for 40+ commercial queries.

Anonymized outcomes from ongoing client engagements using the method the package is built on. Client names withheld; your results depend on your market and what you publish.

Questions people ask

What is generative engine optimization in one sentence?

Making your business one of the sources a generative AI engine cites when it writes an answer to a buyer's question.

Is GEO just SEO with a new name?

Partly. The foundations are shared, and a site with no authority will not be cited by anyone. But generative engines weigh corroboration across independent sources and the shape of the writing (direct answers, statistics, quotations, citations) more than Google's ranking does, and they ignore keyword density entirely. Treating GEO as a rename leads to pages that rank and never get quoted.

Which generative engines does the service cover?

ChatGPT, Gemini, Perplexity and Claude are in the baseline and after-report. Microsoft Copilot and Google's AI Overviews appear in the case-study figures. The work is not engine-specific: the same pages and placements are read by all of them.

Do you track prompt volumes?

We research which prompts matter by asking the engines and reading what they cite, not by buying a volume estimate. Prompt-volume figures from tracking tools are modelled, and for most categories the twenty questions a real buyer asks are obvious once you look at the answers.

How much does GEO cost?

$999 one-time for the Done-For-You Package, which is the GEO service: baseline, audit and strategy guide, 10 pages, 3 DR50+ placements, 3,000 research credits, dashboard and after-report. Delivered in 21 business days. Ongoing weekly monitoring, if you want it afterwards, is Rightcited at $499 per brand per month.

How long does it take to see results?

Delivery is 21 business days. The engines pick up new pages and placements over the following weeks; most clients see meaningful movement in the 60 to 90 days after delivery, and the figures above came from engagements that continued beyond that.

Can you guarantee citations?

No. We guarantee the deliverables and the date, show you the baseline before we start, and show you the after-report when we finish. Anyone guaranteeing a specific answer from ChatGPT is guessing.

Three ways to start

See where you stand

A free, manual SEO + AI search audit of your site, with a report within 48 hours. No credit card, no sales call.