Marketing a business to a bilingual customer base

I spent sixteen years running multi-unit restaurants, and a good stretch of that was inside Korean-American businesses where the dining room ran in two languages at once. The same family would order in Korean at the table and pull up Google in English to check the parking. Nobody switched because of who they are. They switched because of what the task was.

That is the whole insight, and most marketing built for bilingual customers misses it. Agencies build a Spanish page or a Korean page, treat it as a box checked, and then wonder why it gets forty visits a year.

Why does one customer search in two different languages?

Because search language tracks the task, not the person.

Someone looking for a specific dish they grew up with will often type it the way they learned it, because that is the only name the dish has in their head. The same person checking your hours or your parking will type that in English, because that is the vocabulary of logistics they use every day.

Language choice follows the word, not the speaker. A customer who is completely fluent in English will still search a food name, a medical term, or a family word in the language they learned it in.

The population underneath this is not a niche. In the Census Bureau's December 2023 release of the 2018 to 2022 American Community Survey five-year estimates, 78.3% of the US population age five and older spoke only English at home, leaving 21.7% who spoke another language, and California carried the highest foreign-born share of any state at 26.5%. Among people who spoke a language other than English at home, 61.0% of Spanish speakers, 48.2% of Chinese speakers and 69.8% of Tagalog speakers reported speaking English "very well." Fewer than half of Chinese speakers did.

Read those proficiency numbers the right way. They are not saying that half your customers cannot function in English. They are saying that a large share of them are operating in a second language during a transaction, which is precisely when people revert to the word they learned first.

This is why "translate the site" is the wrong first move. You do not need your parking policy in two languages. You might badly need your dish names, service names, and category words in two languages.

Should I pick one language and do everything well in it?

Usually yes, and this is the answer most agencies will not give you because a second language is billable.

A half-built second language does more damage than no second language. A Spanish page with an English phone menu, English forms, and an English speaking front desk creates a customer who arrives expecting service they are not going to get. You spent money to manufacture a bad first impression.

The rule I use: do not market in a language you cannot serve in. If nobody on shift Tuesday afternoon speaks it, do not run ads that promise it.

A second language earns its place where the business genuinely operates in it. Real staff, real answers, real capacity. The market almost always exists. The capacity often does not.

The gap between a translated page and a page that works for a real customer is bigger than it looks, and the difference between translation and localization decides whether the page converts anyone.

What is wrong with a bilingual page that shows both languages at once?

It is the most common build, and Google's own documentation advises against it.

The instinct is reasonable. One page, two blocks of text, everybody served. Google's guide to multilingual sites says otherwise: help Google determine the language correctly "by using a single language for content and navigation on each page, and by avoiding side-by-side translations." It adds that translating only the boilerplate while keeping the bulk of the content in one language "can create a bad user experience if the same content appears multiple times in search results with various boilerplate languages."

The mechanism explains why the tag-based fixes people sell do not help here. Google states plainly that it "uses the visible content of your page to determine its language" and does "not use any code-level language information such as lang attributes, or the URL." A page that is half Spanish and half English is, to that system, an ambiguous page. No markup resolves it.

And here is the related claim to be precise about, because it circulates in a sloppy form. People will tell you that machine translation gets you penalized. What Google's spam policies actually name, under scaled content abuse, is "scraping feeds, search results, or other content to generate many pages (including through automated transformations like synonymizing, translating, or other obfuscation techniques), where little value is provided to users." Read the qualifiers. That describes generating volume from content that was not yours, at scale, adding nothing. It does not describe a restaurant running its five real pages through a translation tool and having a Spanish-speaking employee correct them before publishing.

The honest version is narrower and more useful: unreviewed machine translation is a quality problem and a trust problem before it is ever a policy problem. It reads wrong to the exact customers you are courting, which is a worse outcome than any ranking effect.

What actually needs the second language?

Far less than you think, and specific pieces of it more than you think.

In rough order of return:

1. The words people search for your product with. Dish names, service names, category names. These are the terms with no clean English equivalent and the ones a customer will type verbatim. 2. Your Google Business Profile description and services. This is often more valuable than a website page, because the profile is where the search actually lands. 3. The phone. If a caller cannot get an answer, nothing upstream matters. 4. Reviews. You do not write these, but you can ask in the language the customer speaks, and the review that comes back tells the next customer they will be understood. 5. The menu or service list itself. The physical or PDF one, in the room.

Notice what is missing. The About page, the blog, the mission statement. Those get translated first because they are easy, and nobody reads them in any language.

The ordering is not arbitrary. The best hard number on where local business contact originates comes from dentistry. Patient Prism's Dental Patient Access Report, published 2 July 2026, covers 8,280 dental locations and 11,552,668 patient calls in 2025, with referring source drawn from 5,022,887 attributed calls. It found 90% of attributable patient calls came from the Google ecosystem, with the Business Profile alone at 54%. It is dental rather than restaurants, and Patient Prism sells call intelligence to a customer base that already measures. It is still the largest attributed-call dataset available, and it points at the profile rather than the website.

How do I find out which language my customers are already using?

Where does community actually enter this?

Through people, not through pages.

Bilingual customer bases in the Bay Area tend to be dense and referral-driven. I will flag that as operator observation rather than research, because no dataset I would cite measures it. What I have watched is that a recommendation inside a group chat, a church directory, or a business association outweighs any ad you can buy against it.

The practical version is unglamorous. Join the association that already exists in your category. Ethnic business associations run on real reciprocity rather than networking theater, and the entry cost is showing up consistently rather than buying a sponsorship.

What to do this week

Open your Google reviews and count the languages across the last 150. Write down the split and the date. That number is your actual starting point, not your assumption about your neighborhood.

Take your top five products or services and write down what a customer calls each one in the other language. Not the dictionary word. The word your customers say. Ask two customers if you are not sure.

Put those words into your Google Business Profile services and products and your menu page. That is the one-hour job and it is the highest-return language work most businesses can do.

If you already run a bilingual page with both languages stacked, split it into two pages in one language each, and link them to each other.

Then ask your next ten satisfied customers for a review in whatever language they are comfortable in. Asking without sounding desperate is a matter of timing and phrasing, and it works the same way in every language.

Questions about your bilingual customer base?

Email me at eric@seod.com with your business name and I will count the language split in your Google reviews and tell you what it suggests about where your second-language effort should go first. I read Korean and Spanish reviews myself, so it is a real read, not a translation tool.

I answer these personally, and if the honest answer is that you have a monolingual customer base and should spend the money elsewhere, that is what you will get back.

Or keep reading more on multilingual and community marketing, including what actually works for Spanish-speaking customers in the Bay Area.