A lot of owners see decent traffic and a quiet phone and figure they need a whole new website. I’d hold off on that, cause you almost never do, and the work usually starts on a handful of pages already doing most of the lifting.
Where should I start when my conversion rate is low?
Go after the pages that get the most visits and the fewest calls or forms, cause that’s where a fix pays you back fastest. But first get tracking in place for calls, forms and booked consultations, so you’re counting real leads and not guessing.
Skip that and a month later nobody can say whether the new page worked, since half the leads came in by phone and the phone was never counted. So put call tracking on the number that shows on the site and mark calls, form submissions and booking confirmations as key events in Google Analytics 4.
Then sort your landing pages by sessions and pull the five with the most traffic and the least to show for it. For most service firms that’s the homepage, two or three main service pages, and maybe a blog post that got popular for a question nobody expected. Honestly, those few pages are the whole first project.
What should the top of the page say?
I’d put three things up there, who you help, what you do for them and what happens when they reach out, with the phone number and one clear button right next to it. Use your buyers’ words, not your brochure’s, cause that first screen is where most people decide whether to stay.
The Nielsen Norman Group found that 79% of its test users scanned any new page they came across, and only 16% read word by word. So a page that opens with “Welcome to our firm, founded on a commitment to excellence” just spent its best spot on a sentence nobody needed. Now picture a builder whose first line tells a homeowner they build custom homes on the owner’s own land in that county, on fixed-price contracts, and that a site visit can be booked this month.
Write it the way you’d answer on the phone. Asked if you handle someone’s kind of case, you wouldn’t read them the mission statement, you’d say yes or no and what happens next, and the page should do the same thing in its first two sentences. The rest of the page just proves it.
How do I make a stranger trust the page enough to call?
They have to be able to check it, so use real names and credentials, a physical address, numbers you can back up and links to your sources. Generic praise and stock photos work against you cause they look like every other firm out there.
The Stanford Web Credibility Project drew on research with more than 4,500 people, and its guidelines are pretty plain about it. Make your information easy to verify, show a real organization behind it, and make contact easy. For an injury firm that could be the attorney’s bar admissions and the kinds of cases they’ve handled, for a wealth manager the firm’s registration and how fees work, and for a builder the warranty terms in writing.
Reviews help when they’re real and you show them honestly. Northwestern’s Spiegel Research Center found purchase likelihood for a product with five reviews was 270% greater than for one with none, but that’s product research and not research on hiring a professional, so I’d read it as a few genuine reviews beating zero and not as a number to plan around. And don’t ever invent one. The Federal Trade Commission announced a final rule on August 14, 2024 banning fake reviews and testimonials, and that covers AI-generated ones too.
Copy that sounds generated for any firm in any city hurts trust too, and I get into fixing that in how to make content not sound like AI.
How many fields should a contact form have?
Only what you need to start the first conversation. For most service firms that’s a name, a phone number or email and a short box for what they need, and the rest can wait for the call, cause every extra field is one more reason for somebody to quit halfway.
The best-known example is old but clean. In a 2008 case study the Chicago web firm Imaginary Landscape swapped an 11-field contact form for a 4-field one, and the form’s conversion rate went from 5.4% to 11.9% across two-month periods with no drop in lead quality. They concluded the number of fields mattered, not the number of required ones. It’s one company and a small sample, 10 submissions against 26, so I’d read it as a direction and not a promise. Still, a service firm rarely needs a fax number on its contact form.
Make the phone number tappable on a phone. And put a “book a time” option by the form for people who won’t wait on a callback.
Does page speed still affect conversions?
It does, mostly on phones, and that’s where a lot of first visits happen. A slow page makes the firm behind it look careless, and the buyer has two other names to try.
A 2020 Deloitte study commissioned by Google found a 0.1 second improvement in mobile site speed went with an 8.4% increase in retail conversions and an 8.3% improvement in bounce rate on lead generation information pages. A tenth of a second won’t save a weak page, but heavy sliders, autoplay video and oversized images on a service page are cost with nothing coming back.
Check readability too. The W3C’s accessibility guidelines call for a contrast ratio of at least 4.5 to 1 for normal text, and plenty of polished sites miss it with light gray type, so open your pages on a phone outside and see what you can actually read.
What about visitors who come from ChatGPT or Google’s AI?
A lot of them land on your homepage with a short list of firms already in their head, so now the homepage has to work like a landing page. Tell them up top they’re in the right place and how to reach you, then back up what the AI said.
SE Ranking’s July 2026 analysis, as reported by Hartzer, found 60% of AI-referred traffic landed on homepages, against 17% from organic search. And a twelve-month study by MySEOConsult on a single B2B software company found AI visitors converted to a free trial at 8.79% against 3.54% for organic search. The author says it’s one site, mostly ChatGPT traffic, and partly explained by which pages each group landed on. So these visitors can be worth a lot, and they tend to arrive at the one page most firms treat as a brochure.
Getting named in those answers at all is a separate job, and I cover it in the guide to AI visibility and in how to get your business cited by ChatGPT. This page is about what happens after somebody clicks.
How do I know if the changes worked?
Put the calls, forms and booked consultations on the pages you changed, for the months after the change, next to the same months a year earlier. Month-to-month numbers get fooled by seasons and most service firms have them, so year over year is the fair test.
Write down the baseline first, plus the date each page changed and what you did. Change ten things in one day and you won’t know which one worked, but that’s honestly fine, cause the goal is more calls and not a research paper. For more on why one page pulls more calls than another, the full guide to conversion rate optimization for service firms covers measurement, testing and how it ties into search.
Structure helps too, cause a strong page per service with supporting pages linked around it gets buyers to the right answer faster, which I explain in topic clusters and pillar pages.
What it costs to wait
Usually it’s a few calls a month going to a firm whose page made the decision easier, and you never hear about them, cause a buyer who leaves doesn’t say why. And those are visitors you already paid for, one way or another.
The rarer worst case is a competitor who fixes their pages and also gets named in the AI answers your buyers read, so they get the introduction and the call both. I wouldn’t panic over it. Most of what’s on this page takes days and not months, so pick your five pages this week and start with the first screen.
How I know
My clearest example is my own ecommerce brand, a store and not a law firm or builder, so treat it as direction. On that brand, visitors from organic search convert at 9.2% against 1.1% for paid social, and organic search brings in 40% of revenue from 19% of visits. I think most of that gap is who’s arriving and what they find. Somebody who searched a question and found the answer is in a different spot than somebody an ad interrupted. If you want to know what a full rebuild costs and what each package includes, it’s all in my flat-fee pricing and packages.
Which fixes are worth doing first?
| Fix | Effort | What it helps | When to skip it |
|---|---|---|---|
| Call and form tracking | Low | Tells you which pages bring leads, so every other change can be measured | Never skip it |
| Rewrite the first screen of your top five pages | Medium | Scanners see who you help and how to reach you right away | If those pages already open with a clear answer and a phone number |
| Cut the contact form to three or four fields | Low | Fewer people quit halfway through the form | If you truly need the extra detail to route the lead, and ask it on a second step |
| Add proof a stranger can check | Medium | Credentials, address, real reviews and sourced numbers build trust | Never, but only use what’s real |
| Speed and contrast on mobile | Low to medium | Fewer people leave before the page loads or because they can’t read it | If the page already loads quickly and passes contrast checks |
| Full rebuild of 100 or more pages | High | Every page answers a buyer question, proves it and points to the next step, and it helps search and AI visibility too | If you only have a handful of pages and someone in-house who writes well |
See what buyers find before they ever reach your site
Before we talk, I run your buyers’ questions through ChatGPT, Claude, Perplexity, Gemini and Google and show you who gets named, then we look at whether your pages would win the call.
See who AI names instead of youFlat fee · Written guarantee · You approve every page
Questions people ask
Do I need a new website to improve my conversion rate?
Almost never. A redesign with the same vague copy tends to look nicer and sell about the same, so I’d spend that money on the words, the form, the proof and the phone version first.
How long before I see more calls from these changes?
The small stuff, like a shorter form or a phone number people can tap, can show up within a few weeks. Copy and proof changes need a few months before you can judge them fairly. Line up the same months a year apart so a busy season doesn’t fool you.
What’s the single biggest fix for a service business website?
For most service firms it’s the first screen on the homepage and the main service pages. People scan, and that opening is the part most of them ever see, so say who you work with and what you do in plain words, with the number and one button right there.
Should I add a chatbot or pop-up to get more leads?
I’d hold off. A pop-up that jumps over the first screen covers the exact thing a scanner came for, and a chatbot only earns its spot if it answers faster than the page. Get the page right, then try extras on your busiest pages and count the calls.
Do visitors from AI tools convert better than other visitors?
Maybe, but nobody has much proof yet and the numbers jump around. MySEOConsult tracked one B2B software company for twelve months and saw AI visitors convert at 8.79% against 3.54% from organic search, and the author says plainly it’s a single site. Check your own AI referrals in Google Analytics.
The leads you already have are leaving on the same pages
Every week those pages stay the same, some of the people who found you call someone else, and most of the fix is in the copy and structure. In writing, I guarantee at least 30% more AI visibility and 30% more organic traffic at month six, or I keep working at no charge, and I refund half the fee if either is still short at month twelve. Read the guarantee terms.
See who AI names instead of youFlat fee · Written guarantee · You approve every page
Sources
- Nielsen Norman Group: how users read on the web
- Stanford Web Credibility Project: guidelines for web credibility
- Spiegel Research Center, Northwestern University: how online reviews influence sales
- Federal Trade Commission: final rule banning fake reviews and testimonials (August 2024)
- Imaginary Landscape: fewer fields in a contact form sharply increases conversions (2008 case study, PDF)
- Deloitte: Milliseconds Make Millions, mobile site speed and conversions
- W3C: Understanding WCAG success criterion 1.4.3, contrast minimum
- Hartzer: why AI search traffic behaves differently, citing SE Ranking’s July 2026 analysis
- MySEOConsult: AI traffic converts better than organic search, 12-month B2B SaaS study
What changed: Published November 2026.