Here’s something that surprises a lot of clients: when we start a website project, nobody opens a design tool for weeks. No homepage mockups, no colour palettes, no “here are three concepts to choose from.”
That’s deliberate, and it’s the single biggest difference between a website that performs and one that just exists. Most underperforming websites weren’t built badly — they were built too early, before anyone worked out what the site actually needed to achieve, for whom, and how. The design was an answer to a question nobody had asked yet.
So this post is a look behind the curtain at how we actually plan a website before designing a single page — partly because clients find it useful to know what they’re paying for, and partly because even if you never hire us, thinking through these steps yourself will make whatever you build better.
Why “start with design” fails
When a project starts with visuals, everyone anchors on the wrong things. The conversation becomes about taste — do we like this blue, should the logo be bigger — instead of outcomes. Taste debates have no right answer, so projects drag; and worse, a site can win every taste debate and still fail completely at its actual job, because its job was never defined.
The approach we use instead borrows from design thinking: understand deeply before you create anything, define the real problem, and only then design — testing as you go. It sounds slower. In practice it’s faster, because you build once, in the right direction, instead of redesigning your way toward a goal nobody wrote down.
Step 1: Understand the business, not the brief
Every engagement starts with questions that have nothing to do with websites. What does the business actually sell, and what’s the most profitable part of it? Where do customers come from today? Which enquiries are gold and which waste the team’s time? What’s the growth plan for the next two years?
We do this because the brief a client arrives with — “we need a refresh,” “our site looks dated” — is almost never the real problem. Dig for an hour and it turns out the actual issue is that the site attracts the wrong kind of enquiry, or that the sales team can’t point prospects to anything that explains the premium offering, or that a competitor is winning the searches that matter. You can’t design for a problem you haven’t found.
Step 2: Understand the customer — with evidence, not guesses
Next, we get specific about who the site must serve. Not “everyone” — the two or three types of visitor who actually drive the business, what each one is trying to accomplish, what they need to believe before they’ll act, and what would make them hesitate.
Where we can, we use evidence over opinion: analytics from the existing site (what people actually do, where they leave), the questions customers genuinely ask the sales team, search data showing what people in Singapore actually type, and — increasingly — what AI assistants say when asked about the business and its competitors. Every guess we replace with evidence removes a future argument and a future redesign.
Step 3: Define what the site is for — in one sentence per goal
This is the step that changes everything downstream. Before any structure or design, we force clarity on the site’s jobs, ranked. For most SMEs it’s some version of: generate qualified enquiries for X service, make the business credible enough to shortlist, and support the sales conversation after first contact.
Each goal gets a measurable definition — not “more leads” but “enquiries for our premium service, through this form, from this kind of visitor.” Once that exists, every later decision has a referee. Should the homepage lead with the company story or the service? Check the goals. Do we need that animation? Check the goals. Taste debates mostly evaporate, because “do we like it” gets replaced by “does it serve the job.”
Step 4: Architecture before aesthetics
Only now do we sketch the site — and still no visuals. This is structure: what pages exist, what each page’s single job is, how a visitor moves from landing to acting, and what content each page needs to do its job.
Two things get decided here that most rushed projects skip. First, the conversion path — the deliberate route from “first visit” to “enquiry,” with every step earning the next. Second, findability — how the structure serves search and, now, AI answers: what topics the site needs to own, and how pages should be organised so both Google and AI engines can understand and cite them. Getting this right at the planning stage costs nothing; retrofitting it after launch is expensive.
We also plan for the unglamorous realities at this stage — how the site will handle personal data properly under the PDPA, what needs to integrate with the client’s existing systems, and what the team will realistically maintain after launch. A plan the client can’t sustain isn’t a plan.
Step 5: Content before design
Here’s a rule we hold to that saves enormous pain: real content comes before visual design. Design built around lorem ipsum is fiction — the layout that looks elegant with placeholder text falls apart when the actual message, the real service names, and the genuine proof points arrive. So we draft the core messaging first: what each key page says, in what order, to move its visitor toward its job.
This is also where the earlier steps pay off. Because we know the customer’s real questions and the site’s ranked goals, the content practically writes its own outline — and it’s built from day one to be the kind of clear, well-structured, evidence-backed material that ranks in search and gets cited by AI.
Only then: design
By the time visual design starts, it isn’t guesswork — it’s the final layer over a structure that already works on paper. The designer isn’t asked “make something nice”; they’re asked to give a clear, credible, distinctive visual form to a defined message, for a defined audience, in service of defined goals. That’s why the design round goes quickly and the feedback stays productive: everyone is judging the work against the plan, not against personal taste.
And because the plan defined success measurably back in step 3, launch isn’t the finish line. We check the site against its jobs — are the right enquiries coming in, from the right people — and adjust based on what real visitors do. The plan makes the site improvable, because there’s something concrete to improve against.
The honest takeaway
None of this is secret, and you could run a version of it yourself: understand the business problem, know your real customers, define measurable jobs for the site, structure before styling, write before designing. The reason projects skip it isn’t that it’s hard — it’s that opening a design tool feels like progress and asking hard questions doesn’t. But the weeks spent planning are precisely why the site works — and why it keeps working after the novelty of the new design wears off.
Where Oasis Web Asia comes in
This process is, in short, what you’re actually buying when you work with us — the thinking that makes the design worth paying for. It’s the reason our tagline is that we understand your business first, then build, and it’s the kind of web development Singapore SMEs come to when they’ve had enough of sites that look fine and do nothing. If your current website was built straight from a design concept and it shows — wrong enquiries, no enquiries, or a site the team quietly avoids sending people to — the fix usually starts with the questions above, not another coat of visual paint.
If you’d like us to walk through those questions with you for your own business, that’s exactly the conversation we like to have.
Start a conversation → — get a free consultation with our Singapore-based team.