Most landing pages take six weeks to produce, and roughly five and a half of those weeks are spent waiting — for feedback, for copy, for a decision about a headline that will be changed again anyway. The work itself is not slow. The process around it is.
This post walks through the sprint I use to take a landing page from a conversation to a live URL, and more usefully, the structural decisions that determine whether a page converts. If you build your own pages, the second half is the part worth stealing.
Why Landing Pages Fail
Before the process, the diagnosis. Underperforming landing pages usually share one of four faults:
- They try to do two jobs. A page that asks visitors to book a call, download a guide and follow you on Instagram converts worse on all three than a page that asks for one.
- They describe the product instead of the outcome. Visitors do not want an accounting system. They want to stop dreading month-end.
- They ask for too much, too early. Every extra form field costs conversions. A nine-field form on a first-touch page is a rejection letter.
- They ignore the traffic source. Someone arriving from a price-comparison search needs a different page to someone clicking a retargeting ad for the fourth time. One page for all traffic means one page optimised for nobody.
None of these are design problems, which is why a redesign rarely fixes a page that is not converting.
Stage Zero: The Twenty-Minute Brief
Before the sprint starts, one short conversation covering five questions. Twenty minutes, and it prevents most of the revision cycles that stretch these projects into months.
- What is the one action? Book a call, buy, submit a form, register. One. If two are genuinely necessary, they need two pages.
- Where is the traffic coming from? Google Ads, Meta, an email list, a QR code on a flyer. This determines how much context the page has to establish before it can ask for anything.
- What does the visitor already believe? Cold audiences need the problem framed. Warm audiences find that patronising and scroll past.
- What is the honest objection? Price, switching cost, scepticism, timing. Whatever it is, the page has to answer it — pretending it does not exist does not make it go away.
- What proof do you have? Testimonials, numbers, client names, guarantees, case studies. Gather them before day one, because chasing a testimonial mid-build is the single most common cause of a sprint stalling.
Stage One: Structure and Copy
No design tools at this stage. The page gets written as plain text first, because a page that does not work as text will not be rescued by good typography.
The structure I start from, and adapt rather than replace:
The headline
States the outcome, for a specific person. "Bookkeeping for Johannesburg tradespeople who hate paperwork" beats "Professional Accounting Solutions" every time, because the second one could describe four hundred companies and speaks to nobody in particular.
The subheadline
One sentence on how it works or who it is for. This is where you qualify — deliberately repelling the wrong visitors is a feature, not a loss.
The primary call to action, above the fold
Visible without scrolling, on mobile, without exception. Button text describes what happens next: "Book a 15-minute call" rather than "Submit". Vague buttons make people hesitate, and hesitation is measurable.
Proof, immediately after
Logos, a number, a short testimonial. Whatever is true and specific. This sits high because the first thing a visitor decides is whether to keep reading, and proof is what buys the next ten seconds.
The problem, in their words
Three or four sentences showing you understand the situation they are in. Written using the phrases they use — which is why the brief asks what visitors already believe.
The offer, as outcomes
What they get, framed as results rather than features. "A live site in three weeks" rather than "responsive HTML/CSS build".
Objection handling
The honest objection from the brief, answered directly. Often an FAQ block, sometimes a guarantee, sometimes a single paragraph. Addressing the awkward question openly builds more trust than avoiding it.
The call to action, repeated
Same action, same wording. Never introduce a second, different ask at the bottom of the page.
This stage ends with the page written out and sent over. Feedback on structure and message happens here, in plain text, where changes cost minutes — not after the page is built, where they cost far more.
Stage Two: Design and Build
With the copy settled, design becomes a question of hierarchy rather than decoration. The page gets built mobile-first, because for most South African campaign traffic the desktop version is the edge case.
Things that get decided here, all of which affect conversion more than aesthetics:
- Form length. Every field removed lifts completion. Name and email is usually enough — you can ask the rest on the call.
- Load weight. A landing page carrying four megabytes of hero video loses visitors on mobile data before it ever renders. Images get compressed properly and nothing loads that is not needed.
- Distraction removal. Landing pages generally do not need full site navigation. Every link out is an exit.
- Contrast on the call to action. The button should be the most visually obvious element on the screen, in both colour and placement.
- Real device testing. On an actual phone, on mobile data, not just in a browser's responsive preview.
Stage Three: Wiring, Tracking and Launch
The stage that gets skipped when projects run late, and the one that determines whether you learn anything from the campaign.
- Test the form end to end. Submit it. Confirm the email arrives, in the inbox rather than spam, at the right address, with the right fields. Form notifications silently failing is more common than anyone admits.
- Confirm the thank-you page. A dedicated URL, not an inline message — it gives you a conversion event to track and a place to set expectations about what happens next.
- Install conversion tracking. Analytics plus whatever the ad platform needs. Without this you are buying traffic blind.
- Check page speed on the live URL. Staging environments lie.
- Set up the follow-up. An automatic acknowledgement email at minimum. Leads go cold quickly, and the gap between form submission and first human contact is where most of them are lost. If that follow-up is more than a single email, it is worth building the sequence properly.
What Happens After Launch
A landing page is a hypothesis. Some honest expectations:
Do not judge it on the first fifty visitors — that is noise, not data. Give it a few hundred before drawing conclusions. When you do evaluate, look at where people leave rather than the headline conversion rate alone. A high bounce at the top usually means a mismatch between the ad and the headline. Drop-off at the form usually means it asks too much.
And change one thing at a time. Rewriting the headline, shortening the form and swapping the hero image simultaneously tells you the page got better without telling you why.
Can You Really Do This in Days Rather Than Months?
Yes, with two conditions.
The first is a narrow scope. One page, one action, one audience. A sprint that grows a pricing table, a blog section and a second language is no longer a sprint.
The second is availability. The brief and the copy review both need a prompt response from you. The build is not the bottleneck — waiting is. Almost every landing page that takes six weeks spent five of them in someone's inbox.
The Short Version
Decide the single action before anything else. Write the page as text before designing it. Cut the form down. Test the whole path before sending traffic. Most of the work that makes a landing page convert happens before anyone opens a design tool — which is exactly why it can be done in days rather than months.
If you have a campaign coming up and need the page live before it starts, I run landing page sprints as tightly scoped, fixed-price jobs — designed, written, built, tracked and live, with the price and the launch date agreed up front.
