GEO — GENERATIVE ENGINE OPTIMIZATION
Review sampled AI answers, check how they represent your business and improve the public information you can maintain. Monitoring may include ChatGPT, Gemini, Perplexity, Claude and Microsoft Copilot where access, market and available features allow. These are visibility targets, not promises of mentions or citations; we do not control model training or answers.
Choose visibility targets for the agreed market, language and query sample. Responses vary by product, search access and context; inclusion here is not a promise of mentions or citations.
A partir de USD 600/mês
A partir de USD 1.200/mês
A partir de USD 1.800/mês
A partir de USD 2.500/mês
A partir de USDÂ 750/projeto
A partir de USD 750/mês
Prompts, languages, platforms and context are documented so comparisons have a clear boundary.
Work focuses on verifiable facts and public sources you can edit or request corrections to, not promises to manipulate models.
Reports distinguish work completed, observed answers and limitations. AI mentions are not treated as visits or leads.
Honorários em USD pelo escopo indicado. Impostos, mÃdia, software/API e custos de terceiros são separados salvo inclusão explÃcita. Configuração separada. Escopo e condições são acordados antes do trabalho.
O escopo detalhado abaixo está em inglês; o idioma de execução é acordado na proposta individual.
One brand, market and delivery language unless stated otherwise. Caps are work limits, not outcome promises. Client supplies lawful data, access and material rights. Public AI actions and content require agreed review. No rankings, leads, revenue, hiring or third-party approval guarantees.
Contract, payment schedule, cancellation, hour rollover and support SLA require an explicit agreement. Legal, tax, investment and regulated work require the appropriate jurisdiction and an authorised professional role; this fee is not evidence of licensure.
One monthly documented sample: up to 25 prompts across 3 accessible engines, one brand/language/market, answer/source log and report. API/model access separately costed; sample is not market share.
Up to 4 existing pages of 1,000 words each reviewed/edited with facts, entities and appropriate visible-content-matching markup; baseline monitoring sample included. New research/development separately scoped.
Owned-source consistency program: up to 4 eligible profiles plus 2 page revisions/month and the monitoring sample. No promised Wikipedia acceptance, training-data control or citation.
One up-to-50-page technical SEO scope, up to 5 on-page tasks/2 briefs, documented 25-prompt/3-engine sample and up to 2 GEO page revisions monthly. Shared audits/content planning counted once.
One snapshot of 25 prompts across 3 engines, brand and up to 3 competitors scored in the same recorded answers; report/priorities, not indefinite monitoring.
One real business/location, up to 3 eligible profile reviews, 2 location/service-page revisions, 15 prompts across 2 accessible engines. No invented office, reviews or guaranteed visibility.
Share your website, approved business facts, target markets and languages, example prompts and any inaccurate answers you have seen. We will discuss a bounded review scope.
A scoped review of brand visibility in sampled AI answers, alongside improvements to factual, accessible public content. It is not a way to control models, buy citations or make a brand appear in every answer.
SEO reviews search accessibility, relevance and organic search performance. GEO adds observation of AI answers and their displayed sources where available. The work can share content priorities, but search traffic and sampled AI mentions are separate measurements.
No. Providers control retrieval, training and answer generation. Answers vary with prompt, language, location, account, mode and time. Publishing a correction does not guarantee that a platform will retrieve it, cite it or update an answer.
We agree a prompt set and sampling schedule, then record answers, mentions, sources and factual discrepancies with available context. Comparisons describe that sample only, not all users or queries. Website visits and enquiries require separate analytics and cannot be inferred from mentions.
ChatGPT, Gemini, Perplexity, Claude and Microsoft Copilot are possible visibility targets where access and relevant features are available. Platform coverage is agreed for your languages and markets; not every platform or mode provides web sources or citations.
Choose the monitoring, content, strategy, combined SEO, competitor or local information offer to discuss. The proposal defines prompts, platforms, sampling, pages, implementation and approvals, and identifies third-party costs. Fees require approval; no free audit is implied.