Experiments power breakthrough discovery in both the Scientific Method and our adaptation of it for entrepreneurship:
Herein lies the key to breakthrough in both:
If your guess disagrees with experiment, it is wrong. It doesn’t matter how beautiful your guess is, how smart you are, who made the guess, or what their name is… it is wrong.
- Richard Feynman
But while experiments are highly effective at testing guesses, simply running experiments is also not enough. The output of your experiments can only be as good as the quality of your input guesses. Further, running experiments may not automatically lead to new insights. Many experiments invalidate a bad idea and leave you stuck.
This begs the question: “Where do good guesses or ideas come from?”
The answer is that good ideas can come from anywhere. The challenge, of course, is that truly good ideas are rare and often indistinguishable from bad ideas at the outset. You need a robust process that, on the one hand, allows you to source a wide diversity of ideas and, on the other hand, quickly lets you vet good ideas from bad ideas.
This is exactly what the LEAN Sprint is designed to do.
What is the LEAN Sprint
A LEAN sprint is a time-boxed iteration cycle for sourcing, ranking, and testing new ideas.
We have successfully applied LEAN Sprints internally in our team and across several dozen entrepreneurial teams at startups and large organizations.
If you come from a software or design background, you have most likely been exposed to the scrum/agile methodology. While LEAN sprints are heavily influenced by agile and scrum (and more recently by Google Venture’s Design Sprints), there are some key differences.
1. The goals are different
The goal of a scrum sprint is to demonstrate “build velocity.” The goal of a LEAN sprint is to demonstrate NOT “learning velocity” but “traction velocity.”
Much as it’s not enough simply to build a great product or feature, it’s not enough simply to demonstrate learning without business results.
Instead of pointing to the stacks of interview notes, point to the number of customers (or closest proxy of one) you signed up for this week versus last. Only when customer growth (or traction) can be correlated to learning is the learning actionable.
2. The participants are different
Scrum and agile are typically developer-only practices. On the other hand, LEAN sprints require the complete team, including your internal and external stakeholders.
3. Time-boxing does not dictate build or release cadence
I’m a big fan of embracing time constraints (or deadlines) for driving innovation and forcing action.
To achieve great things, two things are needed: a plan and not quite enough time.
― Leonard Bernstein
That said, I’m still an advocate for using just-in-time techniques like kanban and continuous deployment/delivery of value to customers. And the two are not mutually exclusive. Time boxes in a LEAN sprint are only used to force a decision and don’t have to drive the release cycle.
The 5 Stages of a LEAN Sprint
There are five stages to running a LEAN Sprint. These stages are modeled after the align-diverge-converge technique, developed at IDEO, for fighting group thinking while enabling original thinking. This technique uses meetings only for alignment and decisions — not free-form discussion or group brainstorming.
1. Expose Problems
The team aligns around a common understanding of your business model constraints or problems while staying disciplined not to bias their conversation with possible solutions. Much like scientists start with models, entrepreneurs need to build models similarly, which they then rigorously test through experiments.
We use three models for this purpose:
i. Lean Canvas
Your Lean Canvas business model captures the story of your business. This story uses a journey metaphor to describe your rough ballpark destination or goal.
ii. Traction Model
The traction model breaks this ballpark destination into a set of significant milestone markers that you then use to chart progress. While the overall goal is increasing traction, your strategy for getting to each milestone will differ.
Once you position yourself on this roadmap, your right next strategy emerges.
iii. Customer Factory
Finally, the customer factory model deconstructs the levers of traction and helps you identify the few constraints holding your business model back. These constraints represent the riskiest parts of your business model that require the highest prioritization.
2. Define Solutions
The team then diverges to generate solutions individually. These are captured on a 1-page Strategy Proposal I’ve previously written about.
Like the Lean Canvas, the Strategy Proposal is designed to fit on a single page to encourage deep thinking and rapid idea sharing.
3. Shortlist Solutions
These ideas or strategies are then shared, ranked, and shortlisted in another team meeting (sprint planning). During these discussions, we use a meritocratic voting method that emphasizes proposals based on empirical evidence versus sheer will (less grounded leaps of faith).
We are also less concerned about scoping efforts at this stage. There is an art to breaking any grand strategy (like building a feature or trying a new growth hacking technique) into small, fast, additive experiments. The first step is embracing a time constraint.
Instead of sizing our experiments into small, medium, or large buckets (like sprint stories), we force all experiments to fit within the same LEAN Sprint time-box. The right time-box for a LEAN sprint is determined by the stage of your product and the size of your team. I recommend starting with 2-week sprints and adjusting from there.
We don’t always get to completely validate an idea within a 2-week sprint window, but we often get to completely invalidate the idea in that time.
4. Test Solutions
The team diverges again to test their solutions (or strategies) using one or more experiments. My last post covered a set of ground rules for running good experiments. All experiments go through the three-step build-measure-learn cycle.
It’s okay to run more than one experiment to test a strategy (either in parallel or serially) within the sprint window, but all experiments have to complete by the end of the sprint.
5. Decide on Solutions
Experiment results are then analyzed against the models from earlier, and the next actions are decided in a sprint review meeting.
The cycle then repeats with the next sprint.
Putting it all together
Here’s what a typical 2-week LEAN Sprint cadence looks like:
LEAN Sprints aim to accomplish the following:
1. Goal — Break a big vision into smaller time-boxed goals.
2. Orient — Align the team around problems versus solutions.
3. Leverage — Source/rank a wide diversity of possible solutions or strategies for achieving the goal.
4. Experiment — Test these strategies additively using experiments.
5. Analyze — Reevaluate learning against the goal.
6. Next Action — Decide what’s next.
Want to learn more about LEAN Sprints?
I am rolling all our learning from running hundreds of LEAN Sprints into a book, PDF handbook, and the Lean Stack software.