SEO for Latino-owned businesses in the United States

My own Search Console has been collecting a demand signal I never wrote for. Over sixteen months, queries like "agencia de seo estados unidos que ayudan a negocios locales latinos" logged 154 impressions at an average position of 28.5. "posicionamiento en buscadores en washington" logged 225 at position 39.9. Those positions are terrible, and that is the point: there is measurable demand and almost nothing decent answering it. A market where the average answer sits on page four is not a competitive market. It is an empty one.

What is SEO for Latino-owned businesses in the United States?

By researching the Spanish queries separately rather than translating the English ones, because the words differ by country of origin. A new business picks one language; a mid-authority one adds the second with its own research; a consolidated one runs both, paired properly so neither cannibalises the other.

Quick answer

• The core question is not which language to publish in. It is which language your customers type in, which is often not the one they speak at home.
• Translating your site is not a strategy. A correct translation that nobody searches for ranks for nothing.
• Separate pages per language and intent. One page trying to serve both ends up serving neither.
• The demand is real and badly served. Queries in this space sit at positions between 17 and 67 in my own data.

What kind of problem this is

The dominant assumption is that serving a bilingual market is a translation problem: publish in English, run the site through a translator, done. That assumption is comfortable, cheap, and produces sites that rank for nothing in either language.

The reason is that search behavior does not follow spoken language cleanly. A second-generation business owner may speak Spanish at home and search in English out of habit. A recent arrival may search in Spanish for personal services and in English for anything involving paperwork. The same person switches by topic, not by identity.

That makes this a research problem before it is a writing problem. And research on this market is exactly what almost nobody does, which is why the positions in that table are what they are.

The mistake that costs the most, with numbers

I audited a 312-page recipe site whose content had been machine-translated into English and never reviewed. One page ranked in position 1 for its main query and received zero clicks. Another sat at position 2.86 with 316 impressions and one click.

The diagnosis fit in one sentence: it was not a rankings problem, it was a clicks problem. The page for a chicken stew appeared for "sudado de pollo recipe" with a title that read "chicken stew". Google considered it relevant. The person searching for sudado de pollo did not recognize their dish and kept scrolling.

That case is the whole argument of this article in miniature. The translation was correct and useless. Nobody had checked which words the audience actually types, and a technically accurate title lost to the term people search for. The full diagnostic method is in the Spanish guide on rankings without clicks, and the query research behind it in keyword research.

How the research actually works

Start with your own data, not with a tool

Search Console already lists the queries your site appears for, in every language, whether or not you wrote for them. That list is the only data source that is genuinely yours, and in a bilingual market it answers the language question directly: it shows what your existing visitors type.

Filter by country and compare. The same business often shows two distinct query sets that never overlap.

Harvest autocomplete separately per market

Suggestions differ between the United States, Mexico, Colombia and Spain, and those differences matter here. A term used in Bogotá may be unknown in Los Angeles even among Spanish speakers.

In one harvest across three Spanish-speaking markets, 11,220 seed queries produced 14,715 unique suggestions, of which 8,875 survived cleaning. The method costs nothing and reveals the exact phrasing people use, which is the part a translator cannot give you.

Check what Google already shows

Search your target query and look at the results. If the top ten are in English for a query written in Spanish, Google has decided that this audience accepts English answers. That single observation saves months of arguing about language strategy.

Structure: separate pages, and say so

One page cannot serve two languages. It splinters the signal, confuses the crawler, and reads badly to both audiences.

The structure that holds is a page per language and per intent, connected with hreflang so search engines know they are equivalents. And there is a rule about that connection worth stating plainly: a half-declared pair is worse than no declaration at all. If page A points to page B and B does not point back, Google discards the whole signal. Both directions or neither.

For local businesses, the same logic extends to the Google Business Profile, which in this market often outweighs the website itself. The strategy behind that is covered in the Spanish guide on local SEO.

When NOT to do this

When the business serves one language only and knows it. A restaurant whose customers all order in English does not need a Spanish version; it needs a better English one.

When the site is not indexed properly yet. Two language versions of an invisible site are two invisible sites.

And when the budget only covers one thing well. A single excellent page beats two mediocre ones, and this market is not competitive enough to require both from day one.

What to expect

No timeline is promised here, and anyone who promises one is selling something they cannot deliver. What the data supports is the size of the opening, not its speed.

What the policy says about translated pages, and what people think it says

Anyone researching this will read that unedited machine translation triggers a penalty. It is in more than one of the first results, stated as fact.

Google's spam policies do not say that. What they describe is "scraping feeds, search results, or other content to generate many pages", including automated transformations such as translating. The target is scraped material transformed at volume, not a business translating its own service pages.

The distinction changes the budget. Under the penalty reading, the second language is all-or- nothing: pay for full human translation or skip it. Under what the policy says, machine translation is a starting draft that needs someone who speaks the language to turn it into an answer to the question the reader actually asked, which costs far less than starting over and produces something the first approach never does.

And one mechanic that is not optional: hreflang needs return tags in both directions, and one error in a cluster makes the whole cluster get ignored. Half a signal is worth nothing, so a half-declared pair is worse than an undeclared one.

The pieces, and where each one is developed

Keyword research done twice, because Spanish queries in the US are not translations of the English ones and vary by country of origin: free keyword research.

Pages written rather than translated, with what the spam policies actually say about automated translation in translating a website for SEO, and the service side in Spanish SEO services.

The local layer, which for most of these businesses decides more than the website: local SEO for small business, Google Business Profile and how to get Google reviews.

And the trades where this concentrates: contractors, restaurants, dentists and law firms, where immigration work is the exception that reverses the plan.

What any of it costs is in SEO cost for a small business.

The three situations, and why one plan does not cover them

Businesses serving Spanish-speaking customers in the US arrive at this problem from three very different places, and the generic bilingual advice fits none of them well.

A business that operates in Spanish and has an English site is the most common case and the one losing the most. Its customers search in Spanish, its site answers in English, and the gap does not show up as a ranking problem because the Spanish queries were never targeted at all. The fix is research, not translation.

A business serving both languages equally has the opposite issue: two sets of pages competing for attention with one team maintaining them. Here the risk is that the second language goes stale, which is worse than not having it, because a page that contradicts the current price or service does more damage than no page.

And a business whose customers are bilingual but search in English does not need the second language at all for search, whatever the marketing instinct says. What it needs is for the people who arrive to recognise themselves once they land, which is a conversion problem and not a search one.

Telling which of the three you are is a fifteen-minute job with the query data you already have, and it decides everything that follows. The check is the same in all three cases: open the query report, sort by impressions, and see how many of the top fifty are in Spanish. If the answer is none and your customers speak Spanish, that is not a measurement of demand, it is a measurement of what you published. Search Console can only report queries for pages that exist, so an absence there is never evidence that nobody is asking.

Mistakes that repeat

• Translating the site and calling it a strategy. Correct words nobody searches for rank for nothing.
• Assuming spoken language predicts search language. The same person switches by topic.
• One page serving both languages. It splinters the signal and reads badly to everyone.
• Declaring hreflang in one direction only. Google discards a half-declared pair entirely.
• Skipping the Google Business Profile. For local businesses it often outweighs the site.
• Researching Spanish queries with Spanish-from-Spain data. The markets do not share terms.

Data and transparency

That Google's spam policies describe scaled content abuse as scraping and transforming content at volume, including translation, and do not name unreviewed machine translation on its own as a violation, was checked against the policy page on 12 August 2026. That hreflang requires reciprocal tags and that one error voids the cluster is documented behaviour.

The query table comes from Search Console for marketingnativo.com, a site I own, over a sixteen-month window; it is named because it is mine, and the figures are exact. The 312-page recipe site is client work: it is not named, and its figures are rounded. The autocomplete harvest of 11,220 queries and 14,715 unique suggestions across Colombia, Spain and Mexico is my own research, conducted in August 2026. No population, purchasing-power or market-size figures appear in this article because I could not verify any to a primary source. Across the sites I operate and audit, Search Console records more than 300 million impressions a year. Verified as of August 2026.

Primary sources, opened on 13 August 2026: Google spam policies.

What this changes

The usual framing for this market is cultural: understand the community, speak the language, build trust. All of that is true and none of it is a search strategy.

What the data says is narrower and more useful. There is a set of queries, they have volume, and the pages currently answering them sit at position 28, 39, 51 and 66. Nobody is fighting for that ground. The reason is not that it is hard: it is that serving it requires doing keyword research in two languages, which is twice the work and half as interesting as arguing about which language the audience prefers. The businesses that do the boring version will own it before anyone notices it was available.

Frequently asked questions