How do you build a growth-driven design website on HubSpot CMS?

Growth-driven design has three stages: strategy, a launchpad site, and continuous improvement. The method's own site models them as 10 to 14 days of strategy, 60 to 90 days to launch, and 14-day improvement sprints. Those timings belong to that model. Lean Labs' growth-driven design program, which runs on HubSpot CMS, works in 30-day improvement cycles. Neither length is a rule. A cycle is the team's working rhythm, and you can set it to suit how long your changes take to build and review.

Nothing in those stages is a HubSpot feature. The stages tell a team what to decide and in what order. The platform affects how much work each decision takes to carry out. So the useful way to plan a HubSpot build is to take each kind of change you expect to make after launch and ask who will be able to make it, and how you'll know whether it worked. We describe the first version itself in what a launchpad website is and how to build one.

HubSpot's developer documentation describes a theme as a package of templates, modules and style fields that a developer assembles into an editing system. Editors then choose a template and work with the controls the theme gives them. Global style fields let them change site-wide styles without editing CSS.

That design draws a line through every later cycle. A change that rearranges or rewrites what the theme already contains can be editor work, where the theme's controls and the editor's permissions allow it. A change that needs a module or a layout the theme doesn't have is new developer work, and it has to be scheduled.

The launchpad build is the time to decide where that line sits. Take the changes your strategy work suggests you'll want, such as a comparison block on the pricing page or a shorter form, and ask the developer which of them the theme will let an editor make. Each one an editor can make is a change that doesn't have to wait for development time.

Suppose a B2B software company launches its site on HubSpot and picks demo requests from the pricing page as its focus metric. It records the starting point: how many people visit the page in a month, and how many of them submit the demo form. Then it looks for what stands in the way. The sales team reports that prospects keep asking how long setup takes, and nothing on the pricing page answers that.

Lean Labs' program shows where evidence like this comes from. The agency reviews page behavior, form results, CRM lifecycle data and feedback from the sales team together. It then ranks candidate changes by expected pipeline impact, by buyer relevance and confidence, and by effort. In our example, the sales feedback is what gives the company a buyer problem to solve. The page's visit and form numbers wouldn't have named that question.

The proposed change is a short setup timeline placed above the form. If the theme has a module that suits it and the editor has permission to use it, an editor builds the section. If the theme has no suitable module, a developer adds one first, and from then on editors can reuse it on other pages. We cover how to word the prediction behind a change like this in turning user insights into hypotheses and experiments.

A small edit can be made on the page itself. A rebuilt page, or a set of them, is better prepared out of sight. HubSpot's content staging, available on Content Hub Professional and Enterprise, puts website and landing pages on a sandbox domain. You can clone an existing page or start from a blank one, and you can replace several live pages at once when the work is ready.

Staging separates the pages. It doesn't separate everything those pages rely on. A change to a shared template, to global content such as a header, or to a HubDB table shows up on every live page that uses that asset, even when you made it while working on a staged page. So clone the shared asset first and change the copy. In the pricing example, text and layout edits that belong only to the staged pricing page stay on that page. A change to the site-wide header belongs in a cloned header until launch.

HubSpot's theme test mode previews a change to the theme without being able to publish it, so a theme-level change still needs its own publishing step.

When a staged page replaces a live one, the original page's analytics carry over, so the numbers before and after the change sit on one record. For a relaunch that goes beyond a few pages, see how to plan and execute a HubSpot website redesign.

HubSpot's page A/B test is also a Content Hub Professional and Enterprise feature. It serves two versions of a page at one URL and sends half the visitors to each. HubSpot recommends changing one variable at a time. That suits the example, because the timeline section is the only difference between the two versions. When the test ends, a person chooses the winner and the other version becomes a draft. HubSpot carries out that choice. Whether there was enough evidence to make it is the team's responsibility.

For that reason, the cycle's calendar and the test's evidence are separate questions. A 14-day or 30-day cycle sets when the team meets and plans. It says nothing about whether the test has seen enough visitors. On a quiet pricing page, the test can keep running into the next cycle while the team works on other pages.

The improvement cycle itself isn't tied to a subscription tier. A team without native testing can still ship the timeline section and compare the weeks after with the weeks before. That comparison is uncontrolled, since a campaign or a price change could move the number in the same weeks. Record the decision as a judgment.

The improvement cycle ends by learning from the result and passing it to marketing, sales and service. Suppose the timeline section raised demo requests. Sales now knows to answer the setup question early in a call, and marketing can use the same timeline on campaign pages. The module also now exists in the theme, so an editor with the right permissions can add it to the next page that needs it.

Then the list of ideas gets reordered. A result changes how confident the team is in related ideas, and that changes the order of the roadmap for the cycle that follows.