Most conversion advice is button colours. Change the CTA to orange, add urgency, put a countdown timer on it. That advice circulates because it is easy to test and easy to write about, not because it works.

On the marketing sites we build and look after, the changes that actually shift enquiry numbers are structural. They are decisions about what the page says first, what it asks for, and how much work the visitor has to do to believe you. None of them are button colours.

Conversion is a clarity problem, not a persuasion problem

The mental model most people bring is persuasion: the visitor is undecided, and the page has to talk them into it. That is rarely what is happening.

What is usually happening is that a visitor arrives with a specific question, spends eight to fifteen seconds looking for the answer, does not find it, and leaves. They were not unpersuaded. They were unserved.

That reframe changes what you work on. Instead of trying to be more convincing, you work on being faster to understand. Every change below is a version of that.

1. Say what you do, for whom, in the first screen

The single highest-leverage element on any marketing site is the first thing a visitor reads. It has one job: let someone confirm within a few seconds that they are in the right place.

The failure modes are consistent:

Aspirational abstraction. “Transforming tomorrow, together.” Says nothing, applies to a bank and a bakery.
Internal language. Naming your proprietary methodology before naming the problem it solves.
Category without qualifier. “We are a design studio.” True, and identical to two thousand others.

What works is unglamorous and specific: what you do, who it is for, and what changes as a result. A B2B logistics client of ours replaced a headline about “end-to-end supply chain excellence” with a plain statement of what they moved, for whom, and out of which port. Enquiry quality improved immediately, because the wrong-fit visitors self-selected out and the right ones recognised themselves.

Being specific costs you volume and gains you fit. That is almost always the trade you want.

2. Match the page to the intent that brought the visitor

A person arriving from a search for “brand refresh cost” wants a number. A person arriving from a LinkedIn post about your new work wants to see the work. Sending both to the same homepage wastes half of them.

This is where most sites leak. The traffic is fine, the page is fine, and the pairing is wrong.

Practically:

Know your top five entry pages and what people were looking for when they landed. Search Console tells you the queries; analytics tells you the pages.
Make sure each of those pages answers its query directly, near the top, before it does anything else.
Give each one a next step that fits how ready that visitor is. Somebody researching cost is not ready for “book a strategy call”. They might be ready for a price range and a way to ask a question.

We wrote a whole article that exists for this reason. People search for what a brand refresh costs, so there is a page that answers it with real ranges rather than a contact form. It converts better than a sales page would, because it does the thing the visitor came for.

3. Reduce the number of decisions per page

Every additional option on a page splits attention. A homepage with nine equally weighted paths is a homepage with no path.

The fix is hierarchy, not subtraction. You do not have to delete the secondary content – you have to make it visibly secondary. One primary action per page, stated more than once, with everything else clearly subordinate.

A related and very common problem: navigation with fourteen items. Long menus are a symptom of an unresolved argument about what the business does. Resolving that argument is brand strategy work, not web design work, which is why we cover it in what brand strategy actually is.

4. Lower the cost of the ask

Conversion rate is heavily determined by what you ask for and how expensive it feels.

“Book a 45-minute consultation” is expensive. It implies a calendar commitment, a sales conversation, and probably a pitch. “Send us a short note about your project” is cheap. It costs a paragraph and carries no obligation.

Look at your own forms with this lens:

Field count. Every field costs you completions. Ask for what you need to reply usefully and nothing else. Name, email, and a message box will do for most service businesses. Phone number is optional far more often than people think.
Framing. “Request a quote” implies a negotiation. “Tell us what you are trying to do” implies a conversation.
What happens next. Say it. “We reply within two working days” removes the fear of the void that stops people submitting.

We changed our own contact page from a quote request to an invitation to describe the problem, and the volume went up while the quality held. The people who were going to be serious were still serious. The people who were hesitating had one less reason to.

5. Put proof where the doubt is

Testimonials in a carousel at the bottom of the homepage are decoration. Proof works when it sits next to the specific claim a visitor is doubting.

If you claim you deliver on time, the evidence for that belongs next to the claim, not on a separate page. If your price is above market, the case for the difference belongs on the page where the price appears.

Useful forms of proof, roughly in order of strength:

A named client saying a specific thing. “They rebuilt our site in six weeks and enquiries went up” beats “great to work with” by a distance.
Numbers you can stand behind. Real outcomes, real timeframes. Never invented ones.
Work you can look at. For creative businesses, the work is the argument.
Recognisable logos. Weak on their own, useful as reassurance alongside the above.

Anonymous quotes attributed to “CEO, Financial Services” read as fabricated even when they are not. If a client will not be named, the quote is doing less than you think.

6. Make it fast, especially on mobile

This is the least interesting item on the list and one of the most consequential in the South African market. A large share of your visitors are on mobile, on variable connections, on mid-range devices.

The usual culprits are ordinary: uncompressed images, an uncompressed hero video, fonts loading late, and a stack of third-party scripts. Fixing those is a day of work and often worth more than a redesign.

Test on a real mid-range Android on mobile data, not on a laptop on fibre. What you see on your own machine is not what your market sees.

What does not move the needle

Things we have watched people spend real time on that changed nothing measurable:

Button colour and micro-copy tweaks in isolation. Detectable only at traffic volumes most marketing sites will never see.
Chat widgets on low-traffic sites. If nobody is staffing it, an unanswered chat is worse than no chat.
Fake urgency. Countdown timers on services with no deadline. Visitors recognise it, and it costs trust.
More pages. Adding thin pages to look substantial dilutes the ones that were working.
Redesigns without a diagnosis. A visual refresh that does not change structure or message changes conversion by roughly nothing. Expensive way to find that out.

How to work out what your site needs

Before changing anything, spend an hour on this:

List your top five landing pages and the query or campaign that brought people there.
For each one, write the question the visitor arrived with. In their words, not yours.
Open the page and time how long it takes to answer that question. If it is more than a few seconds, that is your fix.
Check what you are asking for on each of those pages, and whether the ask matches how ready that visitor is.
Load each page on a phone on mobile data and count the seconds.

That exercise finds more real conversion problems than any tool will. It also tells you whether your problem is actually the website at all – sometimes the honest answer is that the traffic is wrong, and no amount of page work will fix a mismatch upstream. Deciding what to measure after that is its own discipline, which we set out in the return-on-creative metric.

Conversion work is not a dark art. It is mostly answering the question people came to ask, sooner, and asking for less in return.

If your site is getting traffic but not enquiries and you want a straight read on where it is leaking, send a short note with the URL and what a good month looks like for you. We will point at the two or three things we would change first.