Skip to content
OpenNow accepting new client engagements
All posts
Playbook5 min read

Our process: five phases, two tracks, websites to web apps

Every project we take on runs through the same five phases. What happens inside them depends on whether you need a website or a web app. Here's each step, where it sits, and why only web apps get a proof of concept.

mrkd teamJuly 24, 2026

Five phases, every project

Every project we run moves through the same five phases:

  1. Brief. What is this for, who uses it, and what changes when it ships?
  2. Inputs. What we need from you before we design anything.
  3. Design. The shape of the thing, settled before we build it.
  4. Build & launch. Production code, tested, live, and handed over.
  5. Grow. What happens after launch, on a monthly cadence.

The phases are the contract. You always know which one we're in and what has to be true before we leave it.

What happens inside each phase depends on what we're building. A landing page and a dashboard carry different risks. A marketing site lives or dies on the inputs you already have: your brand, your copy, your stack, the pages that already rank. A web app lives or dies on whether the workflow fits the people using it. So we run two tracks through the same five phases. Each has seven steps.

Track one: websites, landing pages, and stores

A landing page goes from brief to live in about three weeks. A full marketing site takes six to ten, depending on page count and how much content we're moving.

Discovery (Brief and Inputs)

On the website track, discovery spans two phases. First a short call: the goal, the audience, and the one action each page should drive. Book a demo, start a trial, buy the product.

Then your inputs, which is where website projects usually go wrong. So we front-load them:

  • Sites you admire, and why. "I like how Linear explains the product in one scroll" beats a moodboard.
  • Brand assets and copy. Logo files, type, colors, photography, and whatever copy exists. Gaps get flagged now, not in week four.
  • Your current stack. CMS, analytics, forms, CRM, hosting.

You get scope and a price within one business day.

UI/UX design (Design)

Sitemap and wireframes where the structure needs them. Then polished desktop and mobile pages in Figma, built as sections your CMS can reuse, not one-off layouts. You approve the designs before we write code.

Development (Build & launch)

Components mapped to CMS blocks on Builder.io, Storyblok, WordPress, or Shopify. Every section becomes a block your editors can drop onto any page. Work ships to a staging URL you review every week.

Content migration (Build & launch)

Copy, images, products, and posts moved in by script, not by hand. Every old URL gets a redirect and its metadata comes across, so the rankings you've earned hold.

Testing (Build & launch)

Real devices and browsers. Core Web Vitals, accessibility, forms, checkout, and tracking events. Nothing ships until it passes.

Launch (Build & launch)

Deployment, DNS, and analytics checks on go-live day. Your editors get a walkthrough of the CMS before handover.

Monitoring (Grow)

We watch speed, rankings, and errors after launch. Then new landing pages and tests on a monthly cadence, or your team runs it solo from the blocks we built. Both are good outcomes.

Track two: web apps and custom software

Dashboards, internal tools, and custom software carry a different risk: building the wrong thing well. So this track spends its early weeks proving the workflow before anyone commits to a full-build number.

Discovery (Brief)

A first call about the workflow you want to fix. Outcomes, users, constraints, and who owns it in production. A scoped reply within one business day.

Research (Inputs)

Here, inputs are not brand files. They're how the work actually happens. Process and data mapping: who decides what, where work stalls, and what your existing systems can support. Your CRM, your database, your accounting tools.

Proof of concept (Design)

In about two weeks we put a working prototype of the core flow in front of real users. Not a Figma file. Software your team clicks through, even if the data behind it is mocked. Once people have used the flow for a day, they know what it needs to be in a way no design review can match.

We iterate until it feels right. Then we lock scope and price. With a proof of concept you've helped shape, the estimate is a number, not a range.

Development (Build & launch)

Disciplined execution on the locked scope. We ship in phases, so your team uses working features early instead of waiting for one big release.

Testing (Build & launch)

Every feature tested against real scenarios: roles, permissions, and the edge cases that only show up in production data. All of it before it reaches your users.

Launch (Build & launch)

Deployment, data migration, and training. We stay close through the first weeks of real use.

Monitoring (Grow)

Uptime, errors, and usage tracked after launch. The version people need after living with it is different from the one that shipped. New views and integrations on a monthly cadence close that gap, built on the same foundation.

Why only web apps get a proof of concept

A website's Design phase ends with approved pages in Figma. That's enough, because a website is a known shape. Visitors scroll, read, and click one button. The risky questions are about content, brand, and search, and Inputs already answered them. Once you've approved the pages, the next useful thing is the real build. An extra step there adds weeks and tells you nothing new.

A web app's Design phase has a harder job. The risky question is whether the workflow works for the people who'll use it every day. A static screen can't answer that. People approve mockups they'd never tolerate in daily use. Only working software exposes the missing field, the extra click, the report nobody can read. So the web app track uses Design to build the smallest thing that answers that question, before the expensive part starts.

Same phase, different job. That's the whole point of two tracks.

What each phase forces

Brief forces a written outcome. If we can't say what changes when this ships, we don't build it.

Inputs force the boring questions early. Missing copy, an analytics setup nobody understands, a data source nobody owns. All cheaper in week one than week five.

Design forces a decision before code. Approved pages for a website, a signed-off scope and fixed price for a web app.

Build & launch forces operational readiness. A site your editors can't update, or an app your team can't run, is a demo we happened to ship.

Grow forces honesty about what launched. Real traffic and real usage tell you what to build next.

Which track are you on?

If you're building a marketing site, landing pages, or a store, you're on track one. See our websites and web design services. If you're replacing spreadsheets with a dashboard, internal tool, or custom software, you're on track two. See web apps.

Not sure? Send us a short brief. We'll tell you which track fits within one business day.

Let's build something

Need a partner like us?

Tell us about your project. We respond within one business day.