What are the core principles of growth-driven design, and what evidence supports them?

The method begins with a strategy stage. The team studies its customers and sets objectives for the site, then turns an organized wishlist of ideas into an implementation plan. The reasoning is that a design decision involves a judgment about what a visitor needs, and talking to buyers first gives the team something to base that judgment on.

When Lean Labs redesigned the website for Integrate, the team worked through buyer personas and mapped the value proposition first. It then tested messages and calls to action on the existing site, simplified the menu and offered a preview of the platform. Within 60 days of launch, Lean Labs reported that the rate at which visits became contacts had risen 271%.

Several things changed at once on that site and there was no control group. So the case can't tell you how much of the change came from the message work. It does show a team doing the audience work first, and the result one agency reported afterward. Our answer on creating a growth-driven design strategy goes through the research itself.

The second principle is to launch sooner with a smaller scope. The method calls the first version a launchpad. It describes the launchpad as a foundation for collecting data about real users and improving from there, and it says the launchpad is not the final product. Building one still involves page plans, content prototypes, design, development and migration. The page also says speed should not compromise the value the site gives its users.

The argument for this principle is about when information arrives. Before launch, the team has its research and whatever the existing site's data shows. It doesn't yet have visitors using the new pages. A smaller first scope can shorten the build, which brings that behavior data forward.

The method aims for a launchpad in 60 to 90 days, and Lean Labs advertises the same window in its own program. The example projects displayed on the method's launchpad page took between 40 and 120 days, so the window is a target and some showcased projects ran past it. We look at the timing question separately in how long it takes to launch a growth-driven design site.

After launch, the method runs a repeating cycle. The team plans around one focus metric and builds the highest-priority changes. Then it learns from the data and experiments and passes the learning on. The method's site models this in 14-day sprints and Lean Labs runs 30-day cycles. Those are two implementations of the same cycle.

In 1993 Jakob Nielsen reported on four interface design projects, each of which went through several versions that were tested with users. Across those four projects, the median improvement in his usability measure from first version to last was 165%, and the median gain per iteration was 38%.

Some changes made usability worse, and he tested each new version for that reason. And the revisions were aimed at specific difficulties users had been observed to have. Growth-driven design relies on a similar loop: observe a problem, change what the team suspects is causing it, then check whether the change helped.

Nielsen's measure was a composite of usability scores from tests with eight to ten people per round. The products were software interfaces such as a home banking system and a cash register. So the study shows measured usability improving over tested revisions in four projects from that period. It does not predict leads or revenue on a marketing site, and it did not study growth-driven design.

Measurement has a gap of its own, which Nielsen described in a later essay on A/B testing. A live test compares versions on the outcome you chose to track. It does not by itself explain why visitors behaved as they did. His answer was to combine experiments with direct observation. Lean Labs' program describes a review that takes in more than one kind of input: page behavior and form data alongside CRM lifecycle stages and feedback from the sales team. Our answer on turning user insights into hypotheses and experiments shows how an observation becomes a test.

The last step of each cycle is transfer. The team takes what it learned about the audience and gives it to marketing, sales and service. Suppose a pricing page test shows that buyers respond to a clear setup timeline. Sales now has a reason to try raising setup time earlier in its calls, and marketing has a topic to try in its next email. Each of those is a new hypothesis for its own channel, because a result on one page doesn't prove the same response elsewhere.

The method's own site publishes outcome figures. Its overview page says agencies using growth-driven design on HubSpot websites reported 14.34% more leads after six months and 12.56% higher revenue over six months than on WordPress websites, based on a 2017 survey. That is a comparison of platforms, reported by agencies about their own clients. It doesn't isolate any of the principles above. Our answer on growth-driven design statistics goes through the published figures one at a time.

A team can also produce evidence of its own, and the Integrate account shows one habit worth copying. Lean Labs reported that total leads rose 174% against the same period before launch. It also said that figure included an eBook co-published around the same time, and that leads were up 45% with that resource excluded. Record your focus metric before launch and write down what else changed in the same weeks. Your before-and-after number will still lack a control group, but it will describe your own site and name the other changes that could explain it.