Introducing Shape Up at Balsamiq
I introduced Shape Up in summer 2025 and adapted it for our size and existing cadence. As PM, I keep the betting table stocked with strong pitches each cycle.
Overview
Balsamiq used to plan quarterly and merge discovery, design, and build into one long template. I introduced Shape Up as a set of principles in summer 2025 and adapted it for our size and existing planning cadence. Rolled it out gradually: pitch template, betting table, cycle structure. Over time the whole company started shipping inside it. As PM, I keep the betting table stocked with strong pitches each cycle.
My role
- Introduced the framework and made the case for adopting it.
- Wrote the operating model doc that explains how we adapted Shape Up to Balsamiq's size and to our existing EOS-based planning.
- Created the pitch template used across the company, including an AI-assisted variant to hand to Claude for pitch drafting.
- Established the betting table cadence and roster.
- Keep the betting table stocked each cycle by shepherding pitches from rough idea to review-ready.
The challenge
Balsamiq had been shaped by instinct for years. Quarterly planning drove direction. A long process template merged research, design, and build into one artifact that stalled easily. It worked when the company was small enough that everyone was in every conversation. It got harder to sustain as priorities multiplied and shipping cadence got fuzzier.
Objectives
- Give shaping a home so ideas get judged, not stalled.
- Give building a rhythm so cycles feel predictable, not open-ended.
- Do this without discarding what already worked (owner-designer product taste, existing planning cadence).
Strategy and execution
Start small, not with a rollout memo
Introduced Shape Up in summer 2025 as a set of principles to try, not a mandate. Ran small experiments with individual pitches and betting sessions before committing to the whole framework. Adoption took months. That was on purpose.
What I adapted, and why
Vanilla Shape Up assumes 60+ engineers, dedicated shapers, and a company where the CEO is happy to step back from build. Balsamiq is much smaller and founder-adjacent. Adaptations that stuck:
- Meshes with our existing planning cadence rather than replacing it. Longer-range direction still evolving.
- "Anyone can shape," in practice the owner-designer does most and is accountable.
- Added Classification (A/B/C) so changes that affect existing customers get judged deliberately. More recent addition.
- Added Strategic fit as a guardrail so pitches connect to company priorities before the betting table.
- Explicit DACI for build-time decisions so tradeoffs don't stall the cycle.
Artifacts I wrote or created
- Product operating model doc. Canonical description of how we adapted Shape Up for Balsamiq.
- Pitch template. The version everyone uses now, with a downloadable variant designed to hand to Claude for AI-assisted drafting.
- Classification doc. How we classify and ship changes that affect existing customers.
- Balsamiq product calendar. Shared calendar so the whole team knows which cycle we're in and when the next betting table is.
- Betting table roster and cadence. Meets each cooldown to align on the next three cycles.
Point of view: the template was the easy part
Rolling out a template is a Confluence page. Changing how a founder-led company plans is a culture shift. In a company shaped by instinct for years, making bets explicit was inherently disruptive.
The way I landed it was to make each explicit step feel like it made the leadership team's life easier, not harder: a way to step back from mid-build noise, not a way to lose control. That's the trade you have to sell.
Process work is invisible when it works. When it doesn't, it's the only thing anyone can see.
Results
- Balsamiq now ships in fixed 6-week cycles with 2-week cooldowns, on a rolling semester rhythm.
- Every pitch that reaches the betting table has the same shape: problem, appetite, solution, engineering estimate, classification.
- Betting table is a real cadence. Meets each cooldown, aligns on the next three cycles.
- Templates and playbooks that other PMs and non-PMs shaping their own pitches actually use.
- Balsamiq can now describe its own operating model in one doc. Before, it lived in people's heads.
Why it matters
Feature work compounds only if the company shipping it has a stable operating layer to build on. What I own now is that operating layer: the templates, the cadence, the decision framework. Every new project starts inside a shape someone can point to, instead of getting reinvented from scratch. That's a different flavor of PM work from designing a specific feature, and it's the kind of work that keeps paying rent long after the person who set it up moves on.