Most of your customers are on a phone. Is your site?

Most of your customers meet your website on a phone, at a bus stop, over cell data. Here's how to see what they see, why it breaks, and what to do about it tonight.

Someone hears about you from a neighbor, pulls out their phone at the bus stop, and types your name. What happens in the next three seconds decides whether they call. If the text is too small to read, the menu won't open, or the phone number is a picture instead of a button, they don't complain. They tap back and pick the next name on the map. This guide gives you a 30-second test, explains why so many sites fail it, tells you what to fix and in what order, and gives you the exact words to use with the person who built your site.

The 30-second check

Do this now, on your own phone, not a laptop. Turn off wifi first so you're on cell data, because that's what your customer at the bus stop has. Open your site and answer eight questions honestly.

Did it load in under three seconds? Count out loud. If you got to "four", that's a no.
Can you read the text without pinching or zooming?
Is your phone number a button? Tap it. It should offer to call. If nothing happens, or it opens a photo, that's a no.
Is there one obvious button at the top: Call, Book, or Directions?
Can you find your prices within two scrolls?
Does the menu open when you tap the three lines, and close again?
Did anything pop up and cover the screen? A cookie banner, a newsletter box, a chat bubble sitting on top of the Call button.
Do the photos look sharp, and did they appear at once, or did you watch them load one by one?

Every "no" is a leak. Three or more, and your site is quietly losing people every day. Write the answers down; you'll use them in the script below.

For a second opinion that isn't your own eyes, open pagespeed.web.dev, paste your address, and choose Mobile. Ignore the technical list. Look at two things: the big performance score (under 50 is a problem, over 80 is good) and the line "Largest Contentful Paint", which is roughly how long the main content takes to show up. Anything above two and a half seconds needs work.

Why old sites break on phones

Most sites that fail the test were built for a desktop monitor and then squeezed onto a phone as an afterthought. The site your web guy showed you on his big screen in 2016 is not the site your customer sees on a subway platform in 2026. Here are the specific ways they fail, in plain words.

Fixed widths. The page was designed 1,200 pixels wide. A phone is about 390. Either the page shrinks until the text is unreadable, or it hangs off the right edge and you scroll sideways to find the menu.
Hover menus. On a desktop, the menu opens when the mouse hovers over it. A finger can't hover. So the dropdown never opens, or it opens and won't close.
Huge photos. A photo straight off a camera can be 5 megabytes. Six of them on the homepage is 30 megabytes, downloaded over cell data before the page settles. That's your "four seconds".
Tiny tap targets. Links designed for a mouse pointer are too small for a thumb, so people hit the wrong one and give up.
Text trapped in images. The phone number, hours, or menu was uploaded as a picture. It can't be tapped, copied, or read by Google, and it blurs when it shrinks.
Popups and overlays. A box designed to sit politely in the corner of a monitor covers the entire phone screen, and the X to close it is a four-pixel dot.
Old plugins and builders. Sliders, animated headers, and page-builder add-ons that were never updated. They load slowly, break in new browsers, and one of them is usually why the menu doesn't work.

None of this is your fault, and none of it is rare. Sites age like any tool nobody maintains, which is the whole reason a monthly plan exists.

The site on your web guy's monitor in 2016 is not the site on a subway platform in 2026.

A worked example: what a slow site costs

Made-up figures, so you can swap in your own. Suppose your site gets 300 visits a month, and 200 of them are on phones. Say your site takes six seconds to load on cell data, and, being generous, one in four phone visitors gives up before it loads. That's 50 people a month who wanted to find you and didn't get to the page.

Not all of them would have become customers. Say one in five would have. That's 10 lost customers a month. For a veterinary clinic at $45 a visit, that's 10 × $45 = $450 a month, or $5,400 a year, gone to whoever's next on the map. For a mechanic at $180 an average ticket, it's $1,800 a month. Compare that with what it costs to fix: a rebuilt, phone-first site is inside the monthly plan, from $499 a month, with a $1,200 setup fee paid once. The veterinary clinic earns that back in the first month of not losing people.

The exact numbers will be different for you. The shape won't. A leak at the front door costs more than any repair.

What to fix first, in order

If you have someone who maintains the site, hand them this list. If you're on a builder platform and do it yourself, most of these are settings you can change tonight.

Make the phone number a real link. Everywhere it appears, it should be tappable and offer to call. On most builders, select the number, click the link icon, and type tel: followed by the digits with no spaces, like tel:3475597878. Do the same for the address with a link to your Google Maps listing.
Shrink the photos. Every image should be under about 300 kilobytes and no wider than 1,600 pixels. Most phones can resize before upload; most builders have a "compress images" setting. Six photos at 200 kilobytes is 1.2 megabytes instead of 30.
Put one button at the top. Call, Book, or Directions, in a color that stands out, visible without scrolling. Our three-button guide explains which one goes first for your kind of business.
Kill the popups. Turn off the newsletter box, the "welcome" overlay, the animated chat bubble that sits on the Call button. If you need a cookie notice, make it a thin bar at the bottom, not a wall.
Replace text-in-images with text. Retype the menu, the hours, the prices as actual words on the page. A bonus: Google can finally read them.
Fix the menu. If the three-lines menu won't open or close, it's usually a plugin. Update it, switch to the builder's default menu, or remove the add-on.
Set text to at least 16 pixels and buttons to at least the size of a fingertip, roughly 44 by 44 pixels. On most builders, this is in the theme or style settings.
Test again on cell data, on your phone, and on someone else's phone that isn't the same brand as yours.

What to say to whoever built your site

Most web people mean well and are busy. A vague "the site looks weird on my phone" goes to the bottom of the pile. A specific list gets done. Send this by email so there's a record.

"Hi [name]. I tested our site on my phone over cell data and found some problems I'd like fixed. 1. It takes about [X] seconds to load; I'd like it under three. 2. The phone number isn't tappable; please make it a tel: link everywhere it appears. 3. [The menu doesn't open / the popup covers the screen / the text is too small to read / the prices are inside an image.] Can you tell me what it'll cost, how long it'll take, and whether any of it is included in what I already pay you? Thanks."

Then have answers ready for the usual replies.

If they say "it looks fine on my end": "Please test it on a phone over cell data, not on wifi or a laptop. Here's a screenshot of what I see."
If they say "that's how the template works": "Then I need a different template. What would it cost to move to one that works on phones?"
If they quote a price larger than the original site: "I'll think about it. What do I own, the domain and the files, if I decide to rebuild elsewhere?" Our ownership guide explains why that question matters.
If they don't reply within a week, that's your answer, and it's the same answer you'll get in month seven when a form breaks.

If you'd rather not have the conversation, the first call with us is free and fifteen minutes. We'll run the test with you on the phone and tell you honestly whether it's a fix or a rebuild.

Phone first. Everything else is decoration.

What good looks like

A site that passes doesn't look fancy. It looks obvious. Here's the target, so you know what to ask for.

It appears in under three seconds on cell data, and the main content in under two and a half.
The first screen shows your name, what you do, where you are, and one big button to call or book. No slideshow.
Text is readable without zooming, on a bright sidewalk.
The phone number, address, and hours are text, tappable, and the same on every page.
Prices or a menu are two scrolls away at most.
Photos are real, sharp, and small enough to load instantly. Real photos, not stock.
Nothing pops up. Nothing slides in. Nothing covers the button.
If your customers speak Spanish, Russian, Georgian, or Ukrainian, a language switch at the top, and the whole page in that language, not just the homepage.

This is the whole spec for the sites we build. Phone first, one button, real photos, real words, fourteen languages. The AI does the slow part, building and testing across screen sizes, which is why the site sits inside the monthly plan rather than carrying a $2,500 price of its own. We do the part that needs a person: we come to your business, take the photos, write every word, and check the finished site on a real phone before it goes live.

Common mistakes and the fix for each

Testing on wifi in the back office. Fast wifi hides a slow site. Fix: turn wifi off and test on cell data, ideally somewhere with two bars.
Judging by the homepage only. The homepage got fixed once; the services page still has the 2016 layout. Fix: tap through every page a customer would visit, especially the one with prices.
Adding a chat widget to "help". It sits on top of the Call button and loads a megabyte of code. Fix: if you want chat, use one that stays out of the way and loads after the page. Our Website Bot is built that way.
Uploading photos straight from the camera. Fix: resize to 1,600 pixels wide and under 300 kilobytes before upload, every time.
Fixing it once and never checking again. Browsers update, plugins break, the builder changes its template. Fix: run the 30-second check on the first of every month, or have a monthly plan that does it for you.

Your checklist

Tested on my own phone, wifi off.
Loads in under three seconds.
Text readable without zoom.
Phone number tappable and rings.
One clear button at the top: Call, Book, or Directions.
Prices or menu within two scrolls.
Menu opens and closes.
No popups covering the screen.
Photos sharp and under 300 kilobytes each.
Checked every page, not just the homepage.
Email sent to my web person with the specific list, or a call booked to talk it through.

Questions owners ask

My site looks fine on my phone. Am I done? Test it over cell data, on a phone that isn't the same model as yours, and on every page. If it passes all eight questions on both, yes, you're done for this month. Set a reminder to check again next month.

Can't I just fix the old site instead of rebuilding? Sometimes. If it's two or three items from the fix list, a good web person can do it in an afternoon. If it's fixed widths, hover menus, and an old page builder together, you're paying to patch a foundation that'll crack again. Ask for both prices and compare the first-year totals, which our cost guide shows how to do.

What's a good load time, really? Under three seconds on cell data for the page to be usable, and the main content visible in about two and a half. Faster is better; there's no such thing as too fast.

Do I need an app instead? No. Nobody downloads an app for a dental office or a diner. A fast site that works on a phone does everything an app would, with no install.

Will a phone-first site look bad on a computer? No. Built properly, the same site rearranges itself for a big screen. Phone first just means the phone version is designed first, because that's where most customers meet you.