What growth-driven design statistics exist, and what can each one establish?

The method's overview page says agencies that used growth-driven design on HubSpot websites reported 14.34% more leads after six months than on WordPress websites, and 12.56% higher revenue over six months. It attributes both figures to a 2017 survey. The report page identifies that survey as the 2017 State of Growth-Driven Design, Agency Edition, a 50-page report offered to agencies as material for their sales pitches.

The comparison in that sentence is between two platforms. So the figures don't measure growth-driven design against a traditional redesign. They are also agencies' reports about their own clients' sites, and an agency's account of a client's revenue is a different kind of record from the client's own books.

The survey supports one statement: in 2017, the agencies that answered reported a gap between those two groups of sites. It is not an estimate of what a team would see on either platform today. Our answer on whether HubSpot is the best platform for growth-driven design treats the platform question directly.

The method sets a goal of 60 to 90 days for a launchpad site, the focused first version that the team improves after launch. The examples displayed on the same page took between 40 and 120 days. Those are projects the publisher chose to show, so they tell you launch times varied threefold within one showcase. They don't give an average for the method.

A launch date also needs its scope beside it. In the method, a launchpad has a smaller initial scope than the full plan, with more work scheduled for later cycles. A fair speed comparison sets it against a project of equal scope. Lean Labs advertises the same 60 to 90 day window for the launchpad in its growth-driven design program, after a strategy stage of 10 to 14 days, and those are one provider's stated stages. We cover what moves the date in how long it takes to launch a growth-driven design site.

Lean Labs describes its redesign for Integrate as an inbound-driven redesign, and three numbers from that account aren't interchangeable. Within 60 days of launch, the rate at which visits became contacts rose 271%. Total leads rose 174% against the same period before launch, a total that included an eBook co-published around the same time. With the eBook's leads removed, the rise was 45%. Reporting both lead figures tells a reader that a campaign ran alongside the launch, and neither figure isolates what the eBook or the redesign caused.

A conversion rate and a lead total can move apart. A rate can rise while the count barely moves if traffic falls, and a count can rise on traffic alone. A percentage increase in a rate also depends on where the rate started. Suppose a site turns 1% of its visits into contacts. A 271% increase takes that to 3.71%. The 271% is a change relative to a starting rate the account doesn't give.

Media Junction's 2016 account of its work for the Pittsburgh Parks Conservancy reports a 32.4% month-to-month increase in conversions during the first quarter. That is a growth rate between months. It isn't a total for the quarter, and it isn't a figure for donations. The account is more useful for the revisions it describes. The team changed a menu label from Support to Give and made it more prominent. After launch it moved some card text because it read visitor behavior as people thinking the page ended there, and it reported more scrolling afterward. That reading of the behavior is the agency's interpretation.

Both cases are agencies describing their own projects, with a before and an after and no control group. They establish that these outcomes were reported for these sites, with the metric and the window each account names. They don't forecast another site's result. Our answer on what growth-driven design case studies teach looks at the two projects more closely.

Jakob Nielsen's 1993 report on iterative interface design bears on the reasoning behind the method's improvement cycles, though it predates them. It followed four software interfaces through several tested versions. The median improvement from first version to last was 165%, and the median gain per iteration was 38%.

The four results behind that median were far apart. A hypertext system improved 41% and a cash register system 87%. A home banking system improved 242%, and a security application 882%. The measure was a composite of usability scores from tests with eight to ten people per round, so it isn't a conversion rate or a revenue figure. Nielsen also found that some revisions made usability worse.

What the study establishes is that measured usability improved across tested revisions in four interface projects of that period. It supports the reasoning behind testing and revising. It doesn't measure growth-driven design, and it gives a marketing team no expected lift in leads.

Before planning with any figure, we'd check who reported it, what was counted, over what window, and against what comparison. That applies to a figure an agency or vendor shows you, and to the numbers you produce yourself.

Website visits, lead counts and conversion rates are different outcomes, and a plan built on one of them shouldn't borrow a benchmark stated in another. Pick the outcome your team will track, record it before launch, and note what else changes during the same period. Our answer on which metrics to track in growth-driven design covers how to choose that outcome.