How does growth-driven design work from strategy through launch and continuous improvement?

The method's strategy stage asks a team to understand its customers, set objectives and gather its ideas into an organized wishlist. The wishlist then becomes an implementation plan. An agreed goal gives the team something to rank those ideas against.

Suppose a home solar installer wants its website to produce more booked roof assessments. The team interviews recent customers and looks at where visitors leave the current site. Several customers say they wanted a rough price before they would hand over a phone number. The team now has a goal, a number to watch and a working assumption about what stands in the way.

Its wishlist holds a page explaining how pricing works, a savings estimator, a financing page, a shorter booking form and a gallery of finished installs, along with a dozen smaller ideas. Ranked against the goal, the pricing page moves near the top. The gallery moves down, because nothing the team heard points to it.

The method's overview models the strategy stage at 10 to 14 days, and Lean Labs' growth-driven design program uses the same span. Those figures describe one model and one agency's program, and research across several kinds of buyer can take longer. We cover the stage in more detail in how to create a growth-driven design strategy.

The method describes the launchpad as a foundation for collecting user data and improving the site, and it says plainly that this first release isn't the final product. Building it still involves page plans, content, design, development and migration. The method sets a goal of 60 to 90 days and shows examples that took between 40 and 120. It also says speed shouldn't cost the visitor anything of value.

The ranking from the strategy stage guides what goes in, alongside the pages a site can't run without. For the installer, the launchpad carries the home page, a page explaining how pricing works and the assessment booking page. The savings estimator stays on the wishlist. It would take weeks to build, and the team doesn't yet know whether visitors would use it.

The team gives up part of the site it had in mind and gets real visitors sooner. What those visitors do gives the team evidence for choosing among the remaining ideas. A smaller scope doesn't lower the standard, since each page that ships should be a finished page. We explain the scoping decision in what a launchpad website is and the schedule in how long a growth-driven design site takes to launch.

The method's continuous improvement stage repeats four steps: plan, build, learn and transfer. In planning, the team picks a focus metric and the ranked work it expects to move that metric. A cross-functional team builds the work. The team then studies the results, and it passes what it learned to marketing, sales and service.

For the installer, the focus metric is the share of booking-page visitors who book an assessment. The team takes one idea from the wishlist: show a typical price range above the booking form. It builds the change, publishes it and watches the number.

What the team can conclude depends on how it measured. Comparing the month before with the month after shows whether the rate changed. It doesn't show why, because a seasonal rise in interest or a new ad campaign would move the same number. A split test deals with that by showing two versions to visitors over the same period. HubSpot's page A/B test sends half of a page's visitors to each version, and it's available on Content Hub Professional and Enterprise. A split test still needs enough visitors to separate the two versions, so a cycle can end before the test has an answer.

Suppose more visitors book once the price range appears. That result fits what customers said in the interviews, and the transfer step passes both observations to the sales team. Reps can then bring up price at the start of the assessment call. If bookings don't move, the team has a reason to revisit its assumption about price and try a different idea in the next cycle. We go further into this step in turning user insights into hypotheses and experiments.

The method's overview models 14-day sprints. Lean Labs runs 30-day sprints and picks each sprint's work from page behavior, form results, CRM lifecycle data and sales feedback. The sprint is a period for planning and budgeting the work. A test doesn't have to finish inside it and can run across more than one.

The wishlist carries through the stages. Strategy creates it and the launchpad draws on the top of it. Each improvement cycle then returns to it. The installer's savings estimator is still on the list after launch. If bookings rose once the price range appeared, the team has more reason to try the estimator, because it addresses the same question customers raised in the interviews.

The improvement stage is the one to plan for before the project starts. Someone has to own the wishlist after launch. The team also needs time in each cycle to build a change, review it and choose the metric the next cycle works on. When you budget for a launchpad, budget for those cycles as well.