Onboarding is where the product argues its case

Two offerings, one front door

TurboScale manages browser automation grids so engineering teams don’t have to. The original offering deploys into the customer’s own cloud; the new cloud-hosted tier runs on ours — cheaper per parallel, faster to start, narrower coverage. Different budgets, different buyers, same front door.

The market context that motivated the second tier is withheld: market-shift figures territory; what matters for the design is the consequence: the onboarding now had to route two different intents without asking new users to understand our architecture before they’d run a single test.

Naming the aha moments

The redesign started by writing down, precisely, what success is. For this product it’s two moments: a team runs a test on the trial grid (proof the product works), and later on their own grid (proof it works for them). Time-to-first for each, and the share of teams reaching them, are the funnel — everything else in onboarding either serves those moments or is in their way.

The decisions that shaped the journey

The fork moved to where intent is known. Cloud choice appears at the framework-selection step — reusing the selector pattern users already meet in the flagship product — not on the landing page. A new user shouldn’t choose infrastructure before they’ve seen the product run.

Grid creation became a milestone, not a buried step. For the self-hosted path, creating your grid IS the second aha’s doorway. It left the integration guide and became its own explicit stage with live status and a success state — the journey’s emotional peak, treated like one.

Every step that isn’t an aha got a skip. Trial-grid exploration, console tours, sample builds: valuable, optional. The guide’s spine is the shortest honest path to a passing test.

The journey survives interruption. Grid creation takes real time and pulls users into a different surface; the redesign added an explicit resume affordance so returning users land back at the next step — with the setup’s state, not a blank guide. Designing the re-entry mattered more than decorating the first entry.

The constraint underneath

The journey crosses two systems owned by different teams, and some steps could not move between them within the release. The design treats the seam honestly: each transition is explicit about where the user is going, and the resume mechanism exists precisely because the seam guarantees tab-switches. Pretending the architecture away would have produced a prettier flow diagram and a worse product.

Where the funnel goes from here

The instrumented aha moments give the team a shared scoreboard — withheld: funnel figures stay internal, but every future onboarding argument now starts from the same two numbers. The next design question is already visible: whether the cloud-hosted tier earns its own trial-grid equivalent, so its first aha gets as cheap as the original’s.