Why Churches Lose Bilingual Visitors Before They Ever Walk In

A woman is standing in her kitchen on a Thursday evening with her phone in her hand, looking at your church's website. She has been thinking about going back to church for a few months now. Her mother went. She stopped sometime in her twenties, and now she has a seven-year-old who asks questions she does not feel equipped to answer.

She wants to know one thing: whether there is anything for a seven-year-old on Sunday morning, and whether she has to sign him up in advance.

Your website is in English. She reads English fine, mostly. But "children's discipleship pathway" is not a phrase she has encountered, and the page does not say what time anything starts, and after two minutes of scrolling she calls her son in from the other room and asks him to read it.

That moment, the handing of the phone, is the moment most churches never see and never count. Sometimes the son reads it and explains it and they go. Often he is fourteen and busy and says "I don't know, Mom, it doesn't say," which is true, and she closes the tab.

This post is about that gap. It is not primarily about translation software, and it is not an argument that your church should start a Spanish-language service. The data on that last point is more surprising than most church leaders expect.

The size of the thing

Start with the ground truth, because most churches are working from an impression rather than a number.

In the Census Bureau's detailed analysis of American Community Survey data, 67.8 million people in the United States, close to one in five, speak a language other than English at home. About 62% of them speak Spanish. Among Spanish speakers at home, 39% report speaking English less than "very well."¹ More recent ACS releases put the share at roughly 22%, so if anything the number has drifted up since.

That last number is the one to sit with. It does not mean they speak no English. It means English is effortful. It means reading a dense page of church-specific vocabulary in a second language, at the end of a long day, while deciding whether to walk into a room full of strangers, is one more cost on top of a decision that already feels expensive.

And this is not only a border-state phenomenon. It is Nashville and Des Moines and suburban Atlanta and small towns in Nebraska with a meatpacking plant. The map churches carry in their heads is usually ten or fifteen years out of date.

The counterintuitive part

Here is where most church strategy on this goes wrong.

The instinct, when a church notices it has Spanish speakers in its neighborhood, is to ask "should we start a Spanish service?" It is a serious question with serious costs: a pastor who preaches in Spanish, a worship team, a second set of volunteers, a room, a time slot, and a multi-year commitment you cannot walk back gracefully.

Pew Research surveyed U.S. Latinos on exactly this. Among Hispanic churchgoers, 82% say their church offers Spanish-language services. Only 40% say they would prefer a church that offers them. About half (51%) say it does not matter to them either way.²

Read that again, because it is easy to skim past. Twice as many people attend a church offering Spanish services as say they would go looking for one. That is not the same as saying they dislike it. Most of them are simply indifferent, which is arguably the more useful finding: the language of the service is not what is deciding whether they show up.

The generational data explains why. Among Hispanic immigrants who attend religious services, 79% say most or all the services they attended in the past year were in Spanish. Among their U.S.-born counterparts, that drops to 35%.³ And the household language picture shifts fast: Spanish is spoken exclusively at home by 52% of foreign-born Hispanic adults, 11% of the second generation, and 6% of the third.⁴ Roughly half of second-generation Latino adults are bilingual rather than Spanish-dominant.

So the family in your neighborhood is very often not a Spanish-speaking family. It is a mixed-language family. Grandma prefers Spanish. Mom is comfortable in both and switches depending on the topic. The kids are functionally English-first and roll their eyes at their own accent. They are going to attend an English service without much complaint.

What they will not do is fight through an English-only front door with no help.

The loss is not happening in the sanctuary. It is happening earlier, in the part of the process that costs almost nothing to fix.

Where the loss actually happens

There are three moments between "I've been thinking about church" and "we walked in," and language friction shows up in all three.

The search. People search in the language they think in, and for local searches that often means Spanish or a mix. Iglesia cerca de mí. Iglesia bautista en español. Google is reasonably good at surfacing English pages for Spanish local queries, so this is not a wall so much as a handicap: you are competing for that result against churches whose pages actually use the words being searched. The cheapest fix is not a translated site, it is a Google Business Profile with a description and service list that a Spanish speaker would recognize. (If your profile is not claimed in the first place, start there before anything else in this post.)

The website. Not the theology, the logistics. What time. Where to park. What to do with the kids. Whether you have to sign up. What people wear. These are the questions every visitor asks, and they are exactly the questions that are hardest to answer from a page written in ministry vocabulary in a second language. "Nursery," "connect card," "life groups," and "next steps" are all terms that do not survive translation intact, and half of them barely survive in English.

The question. This is the one nobody staffs. A visitor gets 80% of the way through your site and has one specific question left. On a Thursday night, or a Saturday at 11 p.m., there is no one to ask. In English that visitor might email the office. In Spanish, facing an English contact form, most people simply do not. The cost of asking is higher, so fewer questions get asked, and a question that does not get asked becomes a visit that does not happen.

There is commercial research here worth borrowing, with a caveat. CSA Research's long-running "Can't Read, Won't Buy" work found that a large majority of consumers spend most of their time on sites in their own language and are substantially less likely to buy from a site that is not.⁵ The caveat matters: that study surveyed consumers in non-English-speaking countries, not bilingual Americans, so do not treat the percentages as directly transferable. Treat the direction as obviously true and the mechanism as familiar. Language friction reduces action. It does that for software purchases, and it does that for walking into a building where you do not know anyone.

The thing we ask children to do

There is a name in the research literature for what happened in that kitchen. It is called language brokering: children in immigrant families acting as informal translators and cultural interpreters for their parents, at the doctor's office, the bank, the school, and the parent-teacher conference.⁶

The research on it is genuinely mixed rather than uniformly negative. Brokering can build competence, closeness, and pride. It can also produce real strain, particularly when it is constant, and particularly when it inverts the parent-child relationship in settings where the parent is supposed to be the authority.

Set aside the psychology for a moment and consider what it means for a church specifically. If a mother's only path to understanding your ministry to her children runs through her fourteen-year-old, then the person deciding how your church gets described to her is a teenager who is skimming. He is not going to convey warmth. He is going to convey logistics, badly, and only the logistics he found in the first thirty seconds.

Churches spend enormous energy on how they describe themselves. It is worth noticing the cases where that description never arrives.

What to actually do about it

None of this requires a new service, a new hire, or a capital campaign. It requires knowing your neighborhood and lowering the cost of a question. Here is the order to do it in.

1. Find out what is actually spoken near you (30 minutes)

Stop guessing. Two sources, both free:

• Census data. Go to data.census.gov and look up table S1601, "Language Spoken at Home," filtered to your city, county, or ZIP code. It will tell you what share of your neighbors speak another language at home and how many report limited English proficiency.
• Your school district. Districts run a home-language survey at registration and most publish the results. This is often more current and more granular than Census data, and it tells you about young families specifically, which is the demographic most churches say they want to reach.

You may find Spanish. Depending on where you are, you may also find Vietnamese, Haitian Creole, Tagalog, Arabic, Portuguese, Korean, Amharic, or Marshallese. The useful thing about pulling the actual data is that it either confirms what you assumed or tells you something about your own neighborhood that nobody on staff had noticed, and both outcomes are worth thirty minutes.

2. Fix the "Plan a Visit" page before anything else

If you translate exactly one page, translate this one. Not your doctrinal statement, not your history, not the staff bios. The visit page is where the logistics live, and logistics are what language friction destroys.

Keep it concrete and short:

• Service times, written as times, not "join us Sunday mornings"
• The physical address plus one sentence on where to park and which door to use
• What happens with children, including whether you have to register and what the check-in process looks like
• What people wear, in plain words
• One human name, with an email or phone number, who will actually respond

3. Write for a second-language reader, which helps everyone

Most church websites fail bilingual visitors and monolingual visitors in exactly the same way, by using insider vocabulary. Every fix here is a fix for both groups.

Replace "nursery" with "childcare for babies and toddlers." Replace "connect card" with "a short form to tell us you came." Replace "next steps" with whatever the next step actually is. Write times as "9:00 a.m." rather than "first service." Assume no prior knowledge of your building, your programs, or church in general, because a good number of your English-speaking visitors do not have it either.

4. Make the question askable

This is the highest-leverage item, and it is the one most churches skip because it seems hardest.

Somebody has a question at 9:40 on a Tuesday night. The question is small. "Do I need to bring anything for my son?" If the only channel is an English contact form that promises a reply "within 2-3 business days," the question dies, and there is a reasonable chance the visit dies with it.

If they can ask in Spanish and get an answer in Spanish right then, at the moment they were actually wondering, the whole calculus changes. This is the specific problem we built Greetyr to solve, and it is worth explaining how it works, because the mechanism is less obvious than it looks.

Your church picks which languages it offers. That is a deliberate choice on our part rather than an automatic one: the greeter replies in a language your church has actually selected, not in whatever language a visitor happens to type. Visitors switch languages themselves with a picker in the chat bar, and the greeter stays in that language for the whole conversation.

The part that surprises people is that you do not have to translate your website first. The greeter reads the content you already have, in English, and answers in the visitor's language. Your service times, your kids' check-in process, your parking situation, and your staff contacts are already written down somewhere. A Spanish-speaking visitor asking "¿tienen algo para niños de siete años?" gets an answer built from the English page you wrote three years ago and forgot about. No translation project, no duplicate site to maintain, no second version to keep in sync every time your service times change.

Unlimited languages are included on every Greetyr plan, including the entry tier. We do not gate this one, because charging a church extra to speak to its own neighborhood in its own neighborhood's language seemed like the wrong business to be in.

5. Do not promise what Sunday cannot keep

This is the part where a technology post should be honest about its own limits.

If your greeter answers in Spanish and a family shows up on Sunday and not one person in the building can say "welcome, we're glad you're here" in Spanish, you have made a promise you did not keep. That is worse than never having made it. The visitor's read will not be "the website had a translation feature." It will be "they made it look like there was a place for us here."

So before you turn on a second language, do the small unglamorous work:

• Find the people in your congregation who already speak it. They exist and are usually not on staff. Ask them, specifically, rather than making a general announcement.
• Put one or two of them at the door on Sunday, and tell your greeter team which faces to point people toward.
• Make sure your follow-up, the email or the text or the call that goes out in the first 48 hours, goes out in the language the person used to talk to you. Answering in Spanish on Thursday and following up in English on Monday is a strange thing to do to somebody.
• If you have no one, be straightforward about it. "Our services are in English, and Marta at the welcome desk speaks Spanish and can help you get oriented" is an honest and genuinely useful sentence. It beats both silence and overpromising.

The goal was never for software to have the relationship. The goal is for the software to remove the friction that was preventing the relationship, and then get out of the way. Everything Greetyr does is pointed at an in-person connection, and a language a visitor can use is the shortest path there.

Why this is worth doing now

Racially and ethnically diverse congregations in the United States have grown substantially over the past two decades. In research from Baylor University, evangelical congregations that qualify as multiracial (no single racial or ethnic group above 80% of attenders) rose from about 7% in 1998 to roughly 23% in 2019. Mainline Protestant congregations went from about 1% to 11% over a similar period.⁷

Those are real numbers and they are moving in a good direction. They are also nowhere near the diversity of the country itself, and the gap is not mostly a theological gap. Very few churches in America have decided they do not want their neighbors. What has happened is more ordinary. Churches built their front door in one language during a period when that was sufficient, the neighborhood changed, and the front door did not.

The remarkable thing about this particular problem is how cheap the fix has become. Ten years ago, reaching a Spanish-speaking family in your zip code meant hiring a bilingual staff member or maintaining a second website. Now the first, hardest step is mostly a matter of deciding to take it: knowing what is spoken near you, writing your logistics in plain language, and making sure a question asked in Spanish on a Tuesday night gets an answer in Spanish on Tuesday night.

The mother in the kitchen is not asking your church for much. She is asking whether there is something for her seven-year-old on Sunday, and whether she needs to sign him up.

Somebody should be there to tell her.

---

Greetyr is a digital greeter for church websites. It answers visitors' real questions from your existing content, in the languages you choose to offer, and points them toward the people in your building who can take it from there. Every plan includes unlimited languages.

Footnotes

1. U.S. Census Bureau, Language Use in the United States: 2019 (American Community Survey, published December 2022). 67.8 million people age 5 and older spoke a language other than English at home; approximately 62% of them spoke Spanish; 39% of Spanish speakers at home reported speaking English less than "very well." See: census.gov. ↩

2. Pew Research Center, The Shifting Religious Identity of Latinos in the United States, Chapter 5: "The 'Ethnic Church'" (2013 Survey of U.S. Latinos, published May 2014). 82% of Hispanic churchgoers say their place of worship offers Spanish-language services, compared with 40% who say they would prefer a church that offers them; 51% say it does not matter to them. See: pewresearch.org. ↩

3. Pew Research Center, same study. Hispanic immigrants who attend religious services are far more likely than U.S.-born Hispanics to say most or all services they attended in the past year were in Spanish (79% vs. 35%). ↩

4. Pew Research Center analysis of language use among U.S. Latinos across generations. Spanish is spoken exclusively at home by 52% of foreign-born Hispanic adults, dropping to 11% in the second generation and 6% in the third and higher generations. See: pewresearch.org. ↩

5. CSA Research (formerly Common Sense Advisory), Can't Read, Won't Buy, Ben Sargent. The original study surveyed 2,430 consumers across eight non-Anglophone countries, so the figures describe non-English-speaking consumers abroad rather than bilingual Americans. Cited here for the direction of the effect, not the precise percentages. ↩

6. For an overview of the language brokering literature, including both the developmental benefits and the documented strain, see Morales & Hanson, "Language Brokering: An Integrative Review of the Literature," Hispanic Journal of Behavioral Sciences (2005), and subsequent work by Weisskirch on family dynamics in child language brokering. ↩

7. Baylor University, "Racially Diverse Congregations in the U.S. Have Nearly Tripled in the Past 20 Years" (2020), reporting research on congregational diversity from 1998 to 2019. Multiracial congregations are defined as those in which no single racial or ethnic group exceeds 80% of attenders. See: baylor.edu. ↩