Feed it to your favorite LLM

Guide

AI visibility for translation agencies: the measured index

Short answer

The one live answer in this sample named no agency and judged agencies on the people and process assigned to a project. It linked American Translators Association guidance and the ISO 17100 page, and it advised a small paid pilot. The method: ask buyer questions, store every raw answer, and count.

The AI Visibility Index: Translation agencies

On October 14, 2026, claude-personal and codex received three buyer questions about translation agencies. One asked for the best agencies in Rio de Janeiro, Brazil. One asked how to choose an agency. One asked what translating a training manual into Spanish and Portuguese costs, with technical review. Only the how-to question returned an answer, so every count below comes from 1 answer. That answer was general and did not mention Rio de Janeiro or Brazil.

Who the models name

The one live answer named no translation agency. This batch therefore holds no measured data on which Rio de Janeiro agencies assistants recommend.

Where the models send clients

The answer linked the American Translators Association (ATA) guide to buying translation services. It linked a second ATA page that explains certified translation. It linked the ISO page for ISO 17100, the standard for translation service processes. It sent buyers to the receiving institution to confirm requirements for official documents.

What the models judge on

The answer opened with one rule: "Choose an agency based on the people and process it will assign to your project." It judged eight points in a table. Those points were subject expertise, target-language writing, separate review, and disclosed AI use. The rest were confidentiality, clear scope and pricing, accountability, and continuity. Locale mattered, with the example of French for France versus French for Canada. Separate review mattered, because "proofreading for spelling alone won't catch mistranslations." Machine translation and generative AI had to be disclosed, with the human review spelled out. Confidentiality covered who sees files and whether files can be used for model training. Continuity covered keeping the same linguists and exporting the glossary and translation memory. The answer said ISO 17100 "doesn't guarantee that every delivered sentence is correct." It said machine translation followed by post-editing falls outside that standard. For the United States, it separated a "certified translation" from a translation by a "certified translator." Red flags included equal expertise in every language and industry, and vague "human reviewed" claims. Undisclosed subcontracting and the lack of a written correction policy were red flags too. The answer's method was to shortlist three agencies and give each the same brief. It then advised a small paid pilot, scored by an independent reviewer.

How to improve your AI visibility as a translation agency

Treat each step below as guidance.

AI access

Allow the major AI crawlers on your service, process, and language pages. Keep your process page as plain text, because the one live answer judged agencies on process first.

AI business profile

List each language pair you handle, with the locale, such as Brazilian Portuguese or Latin American Spanish. Name the subject areas you cover, such as technical manuals, legal, or medical. Describe who translates, who revises, and who proofreads, as three separate steps.

Identity tags

Add Organization and LocalBusiness data with your Rio de Janeiro address, phone, and hours. List each language pair as a service, with the source and target language named.

Answer bank

Answer the cost probe in public: how you price a training manual into Spanish and Portuguese, and what technical review adds. Explain how you count words, and what changes the price, such as layout, rush, and repeated text. State when you use machine translation, and what human review follows it. State whether client files ever train an AI model.

Google profile

Choose a translation service category and use your real office address. List your language pairs under services. Ask clients to name the language pair and document type in reviews.

Listing consistency

Match the agency name, address, and phone across your site, Google, and every association directory you join. If you hold ISO 17100, link the certificate and name the certifying body, so a buyer can verify it.

Freshness and measurement

Update the language and subject pages when you add a pair or a reviewer. Ask these three questions each quarter, save the raw answers, and keep the sample size beside every count.

Translation Agencies-specific plays

Explain sworn translation for Brazil.

Some clients need a document accepted by a Brazilian court, registry, or university. Check the current rules for sworn translators registered with the state commercial board. Then explain who signs your sworn work and how long it takes.

Separate the two kinds of "certified."

The live answer distinguished a certified translation from a certified translator, using United States terms. Publish a page that says which one each client situation needs, and which one you supply.

Offer the pilot.

The live answer advised a small paid pilot with difficult terminology. Publish your pilot terms, the sample size, and how you take feedback.

Show the reviewer.

The live answer asked who reviews what. Name the revision step on every quote, with the reviewer's subject background.

Publish your continuity terms.

If clients can export their glossary and translation memory when they leave, say so on the site. The live answer listed export as a test of a good agency.

How this index is made

The factory asked three buyer questions through command-line engines on October 14, 2026. The public dataset stores each raw answer and refreshes in waves, so this page is one snapshot of a time series. This sample is thin: only 1 of the 3 questions returned an answer, and it named no agency. The Index methodology page explains the full method.

About Vizaeo

Vizaeo is an AI visibility service for law firms. It measures whether ChatGPT, Claude, Gemini, Grok, and Perplexity name your firm when clients ask for a lawyer, then applies the fixes that make a firm readable and quotable by AI assistants, with proof of each change. The Vizaeo sprint is $499 for 14 days of done-for-you fixes, with re-measurement at 30, 60, and 90 days against your baseline. If we miss the fix list, you get a refund.