Find the loop that makes your growth compound again

A focused sprint for product and growth teams whose first run of growth has flattened and who want a repeatable way to grow. We map how your product actually creates value, brings users back, and pulls new ones in, and you leave with one growth model the team can run experiments against.

The problem this solves

Growth that used to arrive on its own has flattened, and the first run came from luck and timing, so there is no method to run again. When the team sits down to prioritise, the call goes to whoever argues hardest, because no one shares a picture of how the product actually grows. Ask four people on the team how it grows and you tend to get four answers (we have run that exact exercise, and the spread is the problem). Experiments do get run, but they do not build on each other, so the learning never compounds.

What we actually do

A short, focused sprint with your cross-functional team. We start with how your product creates value, then map the loops around it: what brings users back, and what makes existing users pull new ones in. We run the mapping the way it works in practice, with each person sketching their own picture of the main loop before the group converges, so the quieter models in the room surface instead of the loudest voice setting the map. From the loops we build one growth model on your own numbers, so you can see which behaviour connects to users staying and where the largest lever sits.

What you leave with

• One growth model, drawn on your data, that names your loops and shows which behaviour connects to long-term retention.
• A prioritised hypothesis backlog, with each idea tied to a loop, so you can tell an experiment that strengthens a core mechanism from one that sits outside your loops.
• A shared language for growth across product, data, and design, so the next prioritisation call starts from the model rather than from conviction.

Proof

Tradera, Sweden’s largest marketplace for second hand, had data and experiments but no picture of which behaviours drove long-term growth. We shaped a growth team, ran a growth loops workshop, and built a growth model on their data, alongside an experimentation capability the team kept. Daniel Wiskman, Tradera’s product manager, summed up the work:

“We were most happy with the experiments of course. But I’m also very pleased you helped us discover GrowthBook (an open source A/B Testing Tool).”

Read the full Tradera case.

The same approach at Plick, another secondhand marketplace, produced a loop model and a hypothesis backlog the team ran against: 48 experiments in six months, revenue up 13 percent from the ones that worked, and around two-thirds of ideas shown not worth building before they shipped. Read the full Plick case.

Who it’s for, and how it runs

This fits product and growth teams that grew fast once, sit on plenty of data and ideas, and want a repeatable way to grow rather than another round of guessing. It works well for teams that are strong on product but have not yet treated growth as its own area.

The format is a sprint: short by design, run with your cross-functional team (product, data, design), part workshop and part building the model on your numbers. We can start from the data you already track.

Book a call

If your growth has flattened and you want one shared model of how the product grows, with a backlog of experiments to run against it, email hello@scilla.studio and we will book 20 minutes to look at where your next lever sits.