5 Things Every Website Should Do
Five jobs your website has to do, how to test whether yours does them, and how to fix the ones it misses.
Most business websites are not bad. They are just quiet. They sit there looking professional while the phone does not ring, and the owner assumes the problem is the design, or Google, or the economy.
It usually is not any of those.
A website has one job: take a stranger who has never heard of you and turn them into someone who calls, books, buys, or walks in. Everything else is decoration. When a site fails at that job, it is almost always because it is skipping one of five things. Not fifty. Five.
Nothing here requires a developer. Most of it you can check today, in your kitchen, on your phone.
Read it once end to end, then keep it open and run the scorecard at the back.
One warning before you start. You are the worst possible judge of your own website. You know what you sell, you know where everything is, and you have looked at that homepage two hundred times. Your customer has looked at it for four seconds and knows nothing. Every test in this guide is designed to get you back into that stranger’s head. Take them literally.
Answer three questions before anyone scrolls
When someone lands on your site, their brain is running three questions at once, and it wants all three answered before it decides whether to stay: what is this, is it for me, and why you instead of the other tab I have open.
That decision happens in the first screen, before a single scroll. If your top section says “Welcome to Our Website” or “Quality You Can Trust Since 1998”, you have answered none of them, and the visitor is already gone. They will not tell you they left. They will just be a number in an analytics report you never open.
This is the most common failure we see, and the cheapest one to fix. It is words, not code.
Run this test
The five second test. Get someone who does not know your business. A neighbor, your kid, a friend from another industry. Pull up your homepage on a phone, let them look for five seconds, then take the phone away and ask:
- What does this company do?
- Who do they do it for?
- Where do they do it?
If they hesitate on any of the three, your top section is not working. Do not argue. Do not explain what it meant. Their confusion is the data.
The squint test. Look at your homepage and squint until the text blurs. What is still readable? That is what people actually process. If the biggest thing on the screen is your logo or a stock photo of a handshake, you have given your prime real estate to something that sells nothing.
The top of the page tells a stranger what you do, who you do it for, and where, in one sentence they can read without effort. Under it sits one clear button. Around it, three or four short proof points, not paragraphs.
Steal this formula
[The one specific thing that makes you different, in one sentence.]
[Button: the next step] [Secondary: a lower commitment option]
Notice what none of them say: welcome, passionate, solutions, world class, one-stop shop. Notice what they all do: they let a specific person recognize themselves instantly, and they let the wrong person leave, which is a feature and not a bug. A site that tries to speak to everyone gets remembered by no one.
- Write the sentence before you touch the design. If you cannot say it in one sentence, the problem is not the website. It is the positioning, and no amount of design papers over that.
- Put it at the very top, in the largest text on the page. Larger than your logo.
- Cut the throat clearing. Delete "Welcome to", "We are proud to", "Our mission is". Start with the thing your customer wants.
- Add location if you serve a place. "Serving Naperville and the western suburbs" does more for your phone than any adjective ever will.
- Say what you do not do. "We do not do new construction" saves you the estimate calls you never wanted, and it makes the customers you do want trust you more, because only someone confident turns work away.
Make the next step obvious, and make it the same step every time
Every page needs one clear thing you want the visitor to do next. Call. Book. Request a quote. Get on the list. One.
Sites lose people by offering too many exits, not too few. A header with nine menu items, a slider with three competing buttons, a chat bubble, a coupon popup, a newsletter box, and a phone number in gray 12 pixel text at the bottom is not a helpful website. It is a decision the visitor did not ask to make, and the easiest way to resolve a hard decision is to make none and close the tab.
Pick the one action that is worth the most to your business and design the entire site to funnel toward it. Then repeat that action down the page, because people decide at different moments. Someone might be sold by your headline. Someone else does not believe you until they read the third review.
Run this test
- The thumb test. Open your homepage on your phone. Without scrolling, can you tap the thing you most want a customer to do? If your primary action requires a scroll, or worse, requires opening a menu, you are asking a stranger to work for the privilege of giving you money.
- The count test. Count every clickable thing in your top section. If it is more than three, you have a priority problem, not a content problem.
- The dead end test. Visit five random pages on your site. On each one, ask: if this page convinced me, what do I do right now? If any page ends with nothing but a footer, that page is a dead end. Every page needs an exit ramp toward the sale.
One primary action, styled the same way everywhere, repeated three to five times down a long page. A secondary action for people not ready yet. A phone number that is tappable on mobile and visible in the header on every page.
Steal these button swaps
The pattern: describe what the person gets, not what the button does mechanically. "Submit" describes your database. "Get my free quote" describes their afternoon.
Forms: shorter than you think. Every field you add is another reason to quit. Ask for what you genuinely need to have the first conversation, and get the rest on the phone. Name, phone or email, and one open field covers most businesses. Do not ask how they heard about you before you have earned the right. Do not ask for a budget range on a first contact form unless you are prepared to lose the people who have no idea yet, and some of those people were going to spend the most.
If you must ask for more, tell them why. “We ask for your address so we can check if you are in our service area” turns an intrusion into a courtesy.
- Name your one action. Write it down. If your team would answer this question three different ways, that is your first problem.
- Put it in the header, on every page, as a button that looks like a button.
- Repeat it after each section that does persuasive work.
- Delete competing asks. The newsletter box, the second slider button, the chat widget that fires at three seconds. If it does not lead toward the one action, it is taking from it.
- Make the phone number tappable and put it where a thumb naturally rests.
- Test your own form. Fill it out from your phone, right now, as a customer. Time it. Then wait and see what actually lands in your inbox, and how long it takes. This one check finds broken forms on live business websites more often than anyone wants to admit.
Load fast, on a phone, on a bad connection
Speed is not a technical nicety. It is the first promise your business makes. A site that takes six seconds to appear has already told the visitor something about how you handle things, and they believe it.
Most of your visitors are on a phone. A good share of them are on a mediocre connection, in a parking lot, in a basement, on hotel wifi. Your site does not get judged on your office fiber connection and your desktop monitor. It gets judged in the worst conditions your customer happens to be in.
Speed also feeds everything else. A slow site drags down how well you rank, how many people reach your form, and how much your ads cost if you run them.
Run this test
Google publishes the three measurements that matter, and gives you the tool for free.
- PageSpeed Insights. Run your homepage and your top service page through pagespeed.web.dev. Look at the mobile tab, not desktop. Three numbers matter.
- The real world test. Turn wifi off on your phone. Load your site on cell data, ideally somewhere with two bars. That is your actual customer experience. Count the seconds out loud. It feels a lot longer when you are the one waiting.
CLS is the one that makes people tap the wrong button because an ad or image loaded late and shoved everything down.
What good looks like: the main headline and the primary button appear almost immediately. Nothing jumps. Images fill their space instead of popping in and pushing content around. Tapping feels instant.
In the order that usually moves the needle most:
- Images. This is the number one cause, and it is not close. A photo straight off a phone or a stock site can be several megabytes and is often being squeezed into a space a fraction of its size. Resize images to the size they actually display, compress them, and serve modern formats like WebP.
- The hero video or slider. Autoplaying background video and rotating sliders are the two heaviest decorations in common use, and neither one is persuading anybody. Sliders in particular tend to get one thing seen, the first slide, while carrying the weight of all of them.
- Plugin and script sprawl. Every tracking pixel, chat widget, popup tool, review carousel, and font library loads code. Audit what is installed. If you cannot name what a tool does for the business, it goes.
- Fonts. Loading six weights of three font families costs real time. Two families, two or three weights each, is plenty for a business site.
- Reserve space for things that load late. Images and embeds should have their dimensions set so the layout does not shift when they arrive. This is what fixes the jumping.
- Hosting. Cheap shared hosting is cheap for a reason. If the server is slow to respond, nothing else you do fully repairs it.
One caution. Do not install a speed plugin and call it done. Aggressive caching and minification tools can and do break layouts, forms, and checkout. Test every page after any speed change, especially the ones that take money.
Prove that real people trust you
You can say you are reliable. Everyone says that. Nobody believes it, because saying it is free.
Proof is different. Proof is a real customer, with a real name, saying a specific thing. Proof is a photo of your actual crew and your actual work instead of a stock model in a hard hat who has clearly never held that tool. Proof is the number of years, the number of jobs, the license number, the service area map.
This is the section most business owners skip because it feels like bragging. It is not bragging. Someone about to spend real money with a stranger is looking for a reason to feel safe. Give it to them, or watch them go find a competitor who will.
Run this test
- The stock photo audit. Walk through your site and count the photos that show your actual business: your people, your trucks, your shop, your finished work, your dining room. Then count the stock photos. If stock wins, your site is selling somebody else's business.
- The specificity test. Read your testimonials. Do any of them say anything a competitor could not have written?
- The reachability test. Can a visitor tell, in ten seconds, that a real company with real people is behind this site? Names, faces, a physical service area, a phone number that a human answers.
Wallpaper
"Great service, highly recommend"
Proof
"They came out the same day my water heater went, and the tech put down a mat before he walked through my living room"
Nobody invents a detail like the mat.
Reviews and testimonials placed where decisions get made, not quarantined on a “Testimonials” page nobody visits. Real photography. Named people. Specific detail. Recent dates. Numbers you can actually back up.
Put proof next to the ask. A review sitting directly beside the quote button does more work than twenty reviews on a separate page, because it answers the exact fear the visitor is having at the exact moment they are having it.
Steal these
"Hi Sarah, glad we got that sorted today. If you have a minute, a quick Google review helps a small business like ours more than anything else. Here is the direct link: [link]. Thank you either way."
the team, the vehicles, the shop or office, three finished jobs, one in-progress shot, one candid of someone actually working, the owner.
Short, no guilt, one tap. Send it the same day.
- Get more reviews, systematically. Not once a year when you remember. Ask at the moment of relief, which is right when the job is done and they are happy. A text with a direct link beats a business card with a QR code every time.
- Take real photos. You do not need a photographer to start. A modern phone in good light is enough. Real beats polished.
- Turn one job into a story. Pick a project you are proud of. Write four short paragraphs: what the customer needed, what made it tricky, what you did, how it turned out. One page. That is a case study, and it will out-sell your services page.
- Show your credentials plainly. License numbers, insurance, certifications, years in business, association memberships, service area. Boring, and it closes deals.
- Keep it honest. Never post a review you wrote, never invent a number, and never claim a certification you do not hold. One discovered fake costs more trust than a hundred real reviews build.
Be findable, and be easy to reach
The best website in your town is worth nothing if nobody lands on it, and nearly nothing if the people who land on it cannot get hold of you.
Findable means someone searching for what you sell, in the place you sell it, can actually find you. Reachable means that once they decide, contacting you takes one tap and the message reliably arrives.
That second half is where good businesses quietly bleed. A form that posts into a void, a phone number with a typo, a contact page that only offers a form when half your customers want to call, an email that lands in spam. Nobody complains. They just call the next company.
Run this test
- The search test. Open a private or incognito window and search the way a customer would, not the way you would. Not your business name. Search "what you do + your town". Where do you show up? Are you in the map results? Is your Google Business Profile claimed, filled in, and current?
- The one page per service test. List your services. Now check whether each one has its own page with real content. If eight services share one thin "Services" page, you are asking Google to rank one page for eight different searches, and it will do a mediocre job of all of them.
- The delivery test. Submit your own contact form. Then check: did the email arrive, how fast, did it land in spam, does it include everything the customer typed, and does it go to more than one person so a vacation does not swallow a lead. Do this every quarter. Forms break silently after plugin updates, host migrations, and email changes.
- The tap test. On a phone, tap your phone number. Does it dial? Tap your address. Does it open maps? Tap your email. Does it open mail?
A claimed and complete Google Business Profile with real photos, correct hours, and reviews you respond to. Your business name, address, and phone number written identically everywhere they appear. One page per service, each with enough substance to be useful. Page titles that describe the page in plain words. Contact options in the header, in the footer, and at the end of every page. Forms that deliver to at least two inboxes.
- Claim and finish your Google Business Profile. For most local businesses this outperforms everything else you could do this month. Categories, hours, service area, services list, real photos, and reply to every review, including the bad ones, calmly.
- Keep name, address, and phone identical everywhere. Suite numbers, abbreviations, and old phone numbers scattered across directories create doubt, for search engines and for people.
- Build one page per service. Explain the problem, your process, what it includes, what it costs or what drives the cost, and what happens next.
- Write page titles for humans. "Emergency Plumbing in Aurora, IL" beats "Home | Services | Company Name".
- Offer more than one way to reach you. Phone, form, email, and text if you answer texts. Let people choose their own comfort level.
- Make the site usable for everyone. Real text instead of words baked into images, meaningful alt text on photos, enough contrast between text and background, tap targets big enough to hit accurately, and a page that works with a keyboard. The published standard asks for a contrast ratio of at least 4.5 to 1 on normal text, and for tap targets of at least 24 by 24 pixels, with Apple and Google both recommending closer to 44 to 48. This matters for the customer with reading glasses in a parking lot, and it is increasingly a legal question too, not only a courtesy.
- Answer fast. None of this survives a three day response time. The fastest honest reply usually wins the job.
The 20 minute self-audit
Sit down with a phone and a timer and go in order. This page remembers what you tick, so you can come back and run it again next quarter.
0 of 26 done
How to read your score. Anything still unticked is a task, not a verdict. Take the three that are cheapest to fix and do those this week. Words are cheaper than design. Design is cheaper than a rebuild. Most sites need words.
Five ways people get this wrong while trying to fix it
They redesign instead of repair
A new look does not fix an unclear message. Plenty of sites need better sentences, not a new theme, and the rebuild simply re-skins the same confusion at a higher price.
They add more pages instead of better ones
Thirty thin pages perform worse than eight substantial ones, and they take five times as long to keep current.
They design for their peers
The award-winning site with the mysterious navigation and the slow scroll animation impresses other designers. Your customer just wants the phone number.
They hide the price entirely
You may have real reasons not to publish exact figures. Say something anyway. Ranges, starting points, what drives the cost. A page that refuses to discuss money reads as expensive and evasive, and it filters out the wrong people, meaning the ones who could afford you.
They set it and forget it
Hours change, staff change, services change, plugins update, forms break. A website is a tool that needs maintenance, not a brochure you print once.
