What should a growth-driven design diagram or phase chart show?

A correct GDD diagram shows two things at once: a short left-to-right sequence of three stages, and a loop that the sequence resolves into and never leaves [1][3]. The three stages are strategy, the launchpad website, and continuous improvement. Strategy is the research phase that sets goals, studies the audience, and produces a prioritized wishlist of ideas. The launchpad is a focused first version of the site built from the highest-impact slice of that wishlist, published fast on purpose. Continuous improvement is the recurring optimization work that begins the day the launchpad goes live [1][3]. Draw the first two as bounded blocks with clear exits and the third as an open circle, because the shapes carry the argument: strategy and the launchpad finish, optimization does not [2].

The second thing the diagram must show is direction of flow, both forward and back. A forward arrow carries the wishlist from strategy into the launchpad, and inside continuous improvement a return arrow closes the loop so learning from one sprint feeds the next [2][6]. Get those two ideas onto the page, a finite runway feeding an infinite loop, and the chart already communicates the core of the methodology: a live site in two to three months, then measured improvement indefinitely, rather than a single large redesign that stops at launch [1][4]. The sections below describe each element in the order you would draw it, with the durations, labels, and hand-offs that make the picture useful instead of decorative. Every figure here comes from the methodology's owners or independent practitioners, so the finished chart can cite its own numbers.

Place three stages along a horizontal axis, left to right, in the order they happen: strategy, then launchpad website, then continuous improvement [1][3]. Size the blocks to their durations so the chart doubles as a timeline, and change the shape at the third stage to signal that the process stops moving forward and starts cycling.

Strategy, 10 to 14 days

Strategy is the research stage, and it belongs first because everything downstream depends on it [1]. On the chart it is a bounded block roughly 10 to 14 days wide [1]. Inside or beneath it, note what the stage produces: goals, audience and persona research, fundamental assumptions about the visitor, and a wishlist of ideas, often 50 to 150 of them, ranked by expected impact [3]. The reason to show strategy explicitly, rather than starting the picture at design, is that the wishlist it produces is the input to every later stage, so a chart that omits it hides where the priorities come from.

Launchpad website, 60 to 90 days

The launchpad is a real, live website built quickly from the top slice of the wishlist, and it is the second block [3]. Draw it 60 to 90 days wide, ending at a clearly marked launch point [3]. The label that matters here is that the launchpad is intentionally incomplete: it covers the highest-impact pages and defers the rest, which is why it ships in months instead of a year [3][6]. Mark the launch point as a starting line rather than a finish line, because the whole argument of the diagram lives in what happens after this block, not at it [4].

Continuous improvement, ongoing

Continuous improvement is the third stage, and it is the one that changes shape [2]. Instead of a block with an exit, draw an open circle or ring that the launch point flows into and never leaves [2]. This is the single most important visual decision on the chart: the first two stages have ends, and this one does not, which is what makes the picture growth-driven rather than a rebranded project plan [4]. The next section describes the steps inside that circle.

Inside the continuous-improvement circle, draw four nodes connected by arrows in a ring: plan, build, learn, and transfer, with the arrow from transfer returning to plan [2]. Each node is a real step in a time-boxed sprint, and giving each one its own arrow is what turns a vague loop into an instruction a team can actually run [2].

Plan

Plan is where the team picks one focus metric and pulls the highest-impact ideas from the wishlist into the sprint [2]. On the chart, this is the node to connect back to the strategy stage's wishlist, because planning is where old strategy meets new data. Label this node with the focus metric, demo requests or pricing-page conversion rate for example, to show that a sprint aims at one measurable outcome rather than general tidying [2].

Build

Build is where the planned changes get made and shipped to the live site [2]. Keep this node modest on the chart relative to plan and learn, because in a working GDD process the thinking and the measuring carry more weight than the production, which is deliberately lightweight and fast. The build node is where the sprint cadence is most visible, so it is a reasonable place to annotate the two-week to one-month sprint length [2][3].

Learn

Learn is where the data from the shipped change gets read against the focus metric [2]. Draw a return arrow from this node, because what the team learns here is meant to feed the next planning round, and a chart without that arrow implies the loop is decorative [2]. This node justifies the whole circular shape: without measured learning flowing back into planning, the diagram is only a straight line bent into a ring.

Transfer

Transfer is where the learning leaves the website team and moves to marketing, sales, and service [2]. It is the step most often dropped from GDD charts, and it is worth keeping, because it shows the value of the process is organizational rather than confined to a better web page [2]. Draw it as the fourth node with an outbound arrow pointing away from the loop toward the rest of the company, then continue the ring back to plan.

One honest caveat about the step count

The methodology's owner draws a four-step loop, while some independent guides compress it to three steps, plan, develop, and learn, running monthly [2][3]. Either version is defensible. Pick one, label it, and do not blend the two, because a five-step hybrid would misrepresent both.

Three annotations separate a useful GDD chart from a decorative one: the focus metric on the loop, the 80/20 marker on the launchpad, and the labeled hand-offs between stages [2][6]. Each is a small piece of text that encodes a real rule of the methodology.

The focus metric

Add a focus metric label to the sprint loop, naming the one outcome the current cycle is trying to move [2]. This annotation matters because it encodes the rule that each sprint targets a single measurable result, not general improvement, which is what keeps the loop accountable [2]. If the chart shows several sprints in series, vary the metric from cycle to cycle, form completion rate on one and average order value on the next, to show the focus shifting as the team learns.

The 80/20 launchpad marker

Mark the launchpad with the 80/20 rule: build only the 20 percent of wishlist ideas that produce roughly 80 percent of the impact [6]. The methodology's own wording is to prioritize and start with only the 20 percent that make 80 percent of the impact, so this marker quotes the process rather than interpreting it [6]. The label is what explains, on the face of the chart, why the launchpad can be both fast and worth launching: it is not a cut-down site, it is the high-impact core with the long tail deferred.

The stage hand-offs

Label the arrows between stages, not just the stages themselves [2][3]. The strategy-to-launchpad arrow carries the prioritized wishlist, so annotate it as such; the launch point that opens the loop carries the first live version of the site into measurement; and the transfer arrow inside the loop carries learning out to other teams [2][3]. Hand-offs are where a reader understands what actually moves through the process, so an unlabeled arrow wastes the most informative part of the diagram.

An optional, traceable outcome figure

If you add a results number, keep it traceable. The defensible figure is from the 2017 GDD community survey: agencies running GDD on HubSpot reported 14.34 percent more leads after six months than comparable WordPress projects, a number to attribute to that survey rather than present as a guaranteed result [1]. Skip any statistic you cannot trace to a live source, and do not invent screenshots or sample dashboards to fill space.

A GDD chart is not a waterfall timeline with a loop bolted onto the end, and three common drawings get it wrong in ways worth avoiding [2][4].

It is not a Gantt chart

A Gantt chart shows fixed tasks against fixed dates and treats the plan as known up front, which is the opposite of the GDD premise that the plan changes as data arrives [2]. If your diagram locks every sprint's contents in advance, it has quietly become the waterfall model it was meant to replace. Show the loop as repeating and its contents as chosen each cycle, not pre-scheduled.

It is not a funnel

A marketing funnel narrows from awareness to purchase and describes visitor behavior, while a GDD chart describes the team's work of building and improving the site [1]. The two get confused because both use stages, but a GDD diagram's stages are phases of work, not stages of a buyer, so mixing funnel language into the chart misleads the reader about what the picture measures.

The loop is not decoration

A ring with four labels but no return arrow is decoration, not a process, because the defining feature of continuous improvement is that learning flows back into planning [2]. If deleting everything after the launch point would not change the chart's message, the chart is not showing growth-driven design, it is showing a redesign with a circle drawn beside it [4].