Translation opportunities

Translation opportunities reads your Google Search Console data and takes your top 20 pages by clicks. It finds the languages people already use to find them, and the countries that already send you impressions. Then it turns each page’s main searches into the phrases people type in those languages, not word-for-word translations, and gets their search volume, cost per click and difficulty. You see which page to translate into which language first, and about how many clicks it could bring.

Your first run each month is free: it checks your top 5 pages in 2 languages, the ones people already search your site in (or the biggest markets when there are fewer), and ends by saying what a full run would add. A full run costs 1 credit ($1): the top 20 pages in up to 5 languages. The cost shows before you run it, and a run that fails gives back its credit, or the month’s free run. Sign in, connect Search Console (read-only), pick a site and run it.

What you get

• Each page and language worth translating, best first: searches a month, CPC and difficulty for the country that matters most.
• The searches it already gets: impressions from that country and from searches in that language, from the last 28 days.
• A range of clicks a month a translated page could get. It assumes the next realistic step from where your page ranks now, no higher than position 4, at a typical click rate.
• The phrase to target, the address for the translated page (such as /es/your-page) and the hreflang lines for both pages.
• What not to translate, and why: too few searches, or results that are very hard to beat.
• CSV and JSON downloads, and Copy for agent.

How it works

• It checks up to 5 languages: first the ones people already search your pages in, then the languages of the countries that send you impressions, then big markets.
• Searches in another alphabet, such as Arabic or Japanese, are found from the letters. Searches in another language with the same alphabet, such as Spanish on an English site, are found by Jev, an AI model.
• The phrases come from an AI model that is told to give what people in that country type, and to keep brand and product names.
• Keyword data comes from DataForSEO. A keyword’s numbers are kept for 30 days and shared between runs, without which site asked for them.

After you translate a page

• Have a native speaker check it before it goes live. A literal machine translation reads badly and ranks worse.
• Mark the item done. Dismiss a page and language you do not want to translate: it stays out of later runs.
• Run it again in a few months to see what is new.