Hunnies Body Boutique had traffic. Real traffic, from real customers, arriving at a Shopify store with good products and leaving without buying. The instinct was to buy more traffic. The fix was to stop the store losing the traffic it already had, and it lifted conversion 55 percent in under two months without a dollar spent on acquisition.
Almost every conversion problem I have worked on came down to one of the nine things below.
1. Clarity in five seconds
A visitor should be able to answer three questions almost immediately: what is this, is it for someone like me, and what do I do next. Most homepages answer none of them, because they open with a slogan.
"Elevating brands through innovative digital solutions" tells a visitor nothing. "Web design and SEO for businesses that need the phone to ring" tells them what it is and who it is for. The second one is less impressive to write and considerably more effective to read.
Test it properly: show your homepage to somebody outside your industry for five seconds, then ask them what the business does. If they hesitate, the headline is decoration.
2. One job per page
A page that asks a visitor to book a call, download a guide, sign up to a newsletter and follow on Instagram gets fewer of all four. Attention does not divide neatly.
Decide the single action that page exists to produce, make that action the visually dominant element, and demote everything else. A secondary option is fine when it is genuinely secondary in weight, size and position.
3. Proof where the doubt is
Proof belongs next to the claim it supports, not filed in a testimonials section nobody visits. If your service page says you rank businesses quickly, the evidence for that goes on that page, near that sentence.
What counts as proof, in descending order of usefulness: a named client with a specific measured outcome, a named client with a specific quote, a recognisable logo, a review score with a link to the source, a count of projects. What does not count: an anonymous five star quote, a stock photograph of a smiling person, or a number with no way to check it.
One caution I give every client. If your homepage claims one number and your about page claims a different one, you have not doubled your credibility, you have destroyed it. Prospects notice, and so do the models now summarising your site.
4. Mobile first, actually
Most sites are designed on a large monitor and checked on mobile at the end. It shows. The result is a site that works beautifully in the layout nobody uses and awkwardly in the one that carries most of the traffic.
The specific failures I find repeatedly: text below 16 pixels, tap targets too small or too close together, forms that reflow when the keyboard opens, sticky headers that eat a third of the screen, and horizontal scroll caused by one element that is a few pixels too wide.
Tablet is worse, because designers check phone and desktop and assume the middle takes care of itself. iPad portrait at 768 pixels is where a surprising number of otherwise competent sites fall apart, and it is a real share of traffic in several of the markets I work in.
5. Remove the friction you cannot see
Every field in a form costs you completions. Every step in a checkout costs you orders. Ask for what you need to respond, not for what would be nice to have in your CRM.
Name, email and the question is usually enough. Phone number, company size, how did you hear about us, and a mandatory dropdown of budget ranges are all things you can ask in the reply, once the person is already talking to you.
The exercise that fixes this fastest: complete your own enquiry form or checkout on your own phone, on mobile data, as if you had never seen it. Most owners have never done this. Almost everyone who does finds at least one thing they immediately fix.
6. Answer the objection on the page
Every buyer has three or four reasons not to proceed. Price, timeline, whether you have done this before, what happens if it goes wrong. If the page does not address them, the visitor does not email to ask. They leave.
Put the awkward answers on the page. Price ranges rather than "contact us for pricing". Realistic timelines rather than "fast turnaround". What happens if they are unhappy. Pages that do this convert better and attract better enquiries, because the people who contact you have already accepted the terms.
7. Speed is a conversion metric
Load time is treated as an SEO issue and it is a conversion issue first. People leave slow pages before they have seen anything to evaluate, and on mobile connections the effect is severe.
The usual culprits are large unoptimized images, too many third party scripts, fonts that block rendering, and on WordPress a stack of plugins each loading assets on every page whether needed or not. Fixing those is unglamorous and it is often the highest return work available on a site.
8. One obvious next step
The call to action should say what happens, not what the visitor must do. "Get a free site audit" outperforms "Submit". "See if we are a fit" outperforms "Contact us". The difference is that one describes a benefit and the other describes work.
Repeat it. A long page needs the action offered at the top, in the middle and at the end, because people decide at different points and should not have to scroll back to act.
9. Measure one change at a time
The reason most conversion work cannot prove its value is that six things changed at once and nobody can say which one mattered. Change the highest impact thing, measure it against a comparable window, then move to the next.
This is slower than a redesign and it is the only approach that produces knowledge rather than opinion. It is also how you end up with a site that keeps improving, because you know which of your assumptions were right.
Where to start
If you do only one thing, complete your own enquiry form on your own phone. If you do two, rewrite your headline so a stranger understands the business in five seconds. Those two cost nothing and between them they explain most of the gap I find between sites that produce enquiries and sites that only look like they should.
