TL;DR: Seven patterns we see on almost every service-business site, ranked by how often they cost leads: the homepage doesn't name the buyer (most common), CTAs all say Contact us with no context, no proof above the fold, services are listed not explained, no pricing or pricing is hidden, FAQs are filler not objection-handling, and the contact form has 9 fields when 3 would do. Fixes: name the buyer in the first 100 words, one CTA with a specific promise (Get a quote in 48 hours), proof at the top, real prices or ranges, real FAQs, a 3-field form. Most fixes ship faster than the rebuild teams are considering.
We audit a service-business website roughly once a week. Most of them have the same handful of problems, in the same order, and the fixes are usually faster than the rebuild they're considering.
Here are the seven patterns that show up consistently, what they cost, and what to do instead.
1. The homepage doesn't answer "who is this for"
The single most common failure. The homepage says what the business does (we build websites, we paint houses, we do dental implants) but never names the buyer. Result: every visitor wonders "is this for me," and most leave before they decide.
Fix: Name the buyer in the first 100 words. "We build websites for service businesses doing $250k–$5M in revenue." "We do dental implants for adults who lost a tooth and don't want a bridge." Specific. Filtering.
The homepage should make 80% of visitors say "this isn't for me" and the other 20% say "yes, exactly me." A homepage that tries to qualify everyone qualifies no one.
2. The CTA is "contact us" with no context
Three CTAs in the nav, all saying "Contact us." A footer CTA saying "Get in touch." A floating chat widget asking "How can we help?" None of them tell the visitor what happens next or why they should bother.
Fix: One primary CTA, with copy that promises something specific. "Get a quote in 48 hours." "Book a 15-minute fit call." "See if we'd be a fit." The CTA names the next step. The visitor knows what they're signing up for.
Three CTAs convert worse than one. Always.
3. Stock photography that looks like everyone else's
A header photo of two people in a meeting. A hero image of a city skyline. A team photo from Unsplash where one of the people is also on three other agency websites in your niche.
This pattern destroys trust on contact. Buyers can spot stock photography in 200 milliseconds. The moment they spot it, the site stops being credible.
Fix: Original photography of the actual team, the actual workspace, or the actual work. Even a phone-shot photo of your studio at 3pm on a Tuesday beats a stock image of a meeting that didn't happen. If you can't shoot original, lean into illustration, type-led design, or video. Anything that isn't borrowed.
4. No proof anywhere on the homepage
No client logos. No testimonials. No outcome numbers. No named past projects. Visitors land, read the headline, scroll, and have no evidence that anyone has bought from you and won.
Fix: Three proof elements minimum on the homepage. A logos strip, even if you only have four logos. One named testimonial with a real human's name and role. One outcome line per case ("Booking up 40%" or "Replaced their template site in two weeks").
Logos alone do meaningful work. We've seen homepage conversion lift from adding a six-logo strip below the hero. The logos don't have to be Fortune 500. They have to be real.
5. The contact form is a 9-field interrogation
Name, email, phone, company, role, company size, budget, timeline, what services interest you, message. Submit. By field 6 the visitor's interest has died.
Fix: Three fields maximum on the first ask. Name, email, and either a "what do you need help with" picker or a one-line message. The rest happens on the call. Friction before commitment is friction that kills the lead.
If you genuinely need budget and timeline before you reply, build a multi-step form that paginates the questions and shows progress. Don't put all 9 on one page.
6. The pricing page says "Contact for pricing"
Or worse, there's no pricing page at all. Result: every visitor with a budget under your floor wastes your time on a fit call. Every visitor over your floor self-disqualifies because they assume you're overpriced.
Fix: Show ranges, even if they're directional. "Projects from $5,000." "Starting at $10k for a custom build." Filter the bad fits before the call. The visitors who continue are the ones in your band, and the conversation moves twice as fast.
The fear is "we'll lose the chance to upsell." The reality is you'll lose less time on bad fits, which more than offsets.
7. The site loads slowly on a phone in a parking lot
A site that hits Lighthouse 60 on mobile costs you 30–50% of mobile leads. Most service-business websites we audit land in the 40–70 range because they're running WordPress with 22 plugins, or because they bolted Webflow with a stack of third-party scripts.
Fix: Test the site at webpagetest.org with a mobile profile and a 4G connection. If LCP is over 2.5 seconds, you have a real problem. The fix is usually image compression, removing unused JavaScript, and auditing third-party scripts.
A static site on Vercel with no plugins hits 95+ Lighthouse with no work. A WordPress site needs real attention to get there.
The bonus pattern
There's an eighth one we see almost as often. We left it off the seven because it's harder to fix.
The site has no consistent voice. Every page reads like a different person wrote it. The homepage is corporate-friendly. The about page is folksy. The blog is academic. The pricing page is curt. Visitors don't notice consciously, but they don't trust it either, and they leave without a CTA click.
Fix: One person owns the voice. Every page goes through them. Lock voice early in the project and audit it on every redesign.
How to use this list
Three things to do this week.
- Open your homepage and time-yourself reading the first 100 words. If you can't tell who the site is for at second 30, fix that first.
- Count CTAs above the fold. If it's more than one, decide which is primary and demote the rest.
- Run your site through webpagetest.org on a 4G mobile profile. If LCP is over 2.5s, that's where the next $500 of work goes.
If you want a 15-minute audit of your service business site against this list (no Loom, just an email with the three things to fix first), send us the URL. We send these back in 48 hours.