One landing page argued from sales calls and support threads, built at 360 pixels first, instrumented before launch, and tested only where volume can settle it.
About this service
We do not write a landing page from a brief. Before a word is drafted we read or listen to first-party evidence — sales calls, support threads, churn notes, app store reviews, session replays — and on the last nine builds the headline that shipped came from a customer's own sentence in seven of them. A brief tells you what the company wants to be true. The recordings tell you what the buyer says out loud. The distance between those two sentences is most of this job.
What arrives:
One page, reasoned end to end: the claim, the order of proof, the objection handling and the form, built in your stack and instrumented before it goes live. With it, a rationale naming what each block is doing and why it sits where it does, so the page can be disagreed with on its merits rather than on taste.
How we decide what goes above the fold:
The hero states who it is for and what changes, in the buyer's own words. Proof is placed where the objection appears, not in a logo strip at the top that nobody reads. Long or short is not a philosophy, it is a function of how much the reader must be told before the ask: for a twelve-dollar app that is three screens, for a forty-thousand-dollar programme it is considerably more, and anyone who answers that question before seeing your offer is guessing.
The mobile case:
Built at 360 pixels first and reviewed on a mid-tier Android handset rather than a simulator. A Largest Contentful Paint budget of 2.5 seconds on a throttled connection, fonts self-hosted and subset, hero imagery sized rather than shrunk. Forms tested inside the in-app browsers your traffic actually arrives in, where autofill, file pickers and third-party cookies behave differently from desktop Chrome and where a good-looking page quietly fails.
Testing, and when we refuse:
Detecting a twenty percent relative lift on a three percent conversion rate takes roughly 14,000 sessions per variant before the result carries any weight. If the campaign will not deliver that, we do not sell you a split test. We ship one page, write down the argument it makes, and change it on evidence rather than on a chart that is still noise. Where the volume is there, we move one variable at a time and name the metric before launch.
Instrumentation:
Events for the form, for the scroll depth that means something, and for the primary action, with consent handled under PDPA at the point of collection. If the conversion is a lead rather than a purchase, we agree what qualified means before the build, because a page tuned for raw form fills hands your sales team a worse week and a better-looking report.
What is not included:
Ad creative, media buying, brand identity work, and iteration after the reading period unless you retain us. We write English and Mandarin. For Bahasa Indonesia, Thai or Vietnamese we write the English master, brief a native writer we have worked with, direct the draft against the same evidence, and bill their time at cost.
Who this is not for:
Anyone who needs the page signed off by a committee that has not agreed on the audience. We will not run that argument for you, and a page written to satisfy four opinions converts worse than a page written to persuade one buyer. Also not for teams who want a proven template refilled. That is a real service, it costs a fraction of this, and we do not sell it.