THE COMPLETE GUIDE
What is Rapid Experimentation?
The systematic approach to testing business ideas in days, not months. Created at Google, taught at Stanford, used by enterprises worldwide to save millions by killing bad ideas early.
"The Pretotyping course and rapid experimentation saved our organisation $200,000 before lunchtime on day 2! This should be BAU in every leading organisation."
DEFINITION
Rapid experimentation is a systematic approach to testing business ideas quickly and cheaply using real customer data, before committing significant resources to building them. Rather than spending months on business cases and market research, teams run lightweight experiments in days to weeks that reveal whether customers actually want what you're planning to build.
The methodology uses pretotyping — a set of tools and techniques created by Alberto Savoia at Google in 2010 — to gather real behavioral data from potential customers. The goal is simple: invest only in ideas that customers have proven they want, and kill the rest before they cost you millions.
THE PROBLEM
Why Rapid Experimentation Matters
90%
of new products fail
$30M+
saved by Exponentially clients
4,000+
experiments run since 2020
Most enterprise innovation fails not because of bad execution, but because teams build the wrong thing. They spend months on business cases, market research, and stakeholder buy-in — then launch a product nobody wants.
Alberto Savoia calls this "The Law of Market Failure" — the uncomfortable truth that most new ideas will fail in the market, no matter how well they're built. The failure rate is so high that building without validating first is essentially gambling with company resources.
Rapid experimentation replaces opinion-driven decisions with customer data. Instead of debating whether an idea is good, you test it. Instead of funding a 12-month roadmap, you validate the riskiest assumptions first. The cost of being wrong drops from millions to hundreds.
COMPARISON
Rapid Experimentation vs Other Approaches
Rapid experimentation answers a fundamentally different question than traditional innovation methods.
Traditional R&D
Research for months
- • Build it and hope customers come
- • Relies on market reports and surveys
- • Expensive if the idea is wrong
- • Typical cost to learn: $500K–$5M
A/B Testing
Optimize what exists
- • Compares variations of existing products
- • Requires a product to already exist
- • Optimizes, doesn't validate demand
- • Answers "which is better?" not "should this exist?"
Rapid Experimentation
Test before building
- • Test ideas before writing any code
- • Real customer behavioral data in days
- • Kill bad ideas cheaply and quickly
- • Typical cost to learn: $0–$1K
THE FRAMEWORK
How Rapid Experimentation Works
A structured, repeatable process from ideation to validation. Not gut feel. Not brainstorming. Real data from real customers, gathered as quickly and cheaply as possible.
Ideas
"How do you decide which ideas are worth exploring?"
Structured capture from anyone in the organisation — new products, improvements, AI features, venture ideas. Tap into customer insights, market signals, and cross-functional perspectives. The Exponentially Platform provides systematic capture so nothing gets lost in email chains or spreadsheets.
Prioritise
"How do you focus on what's viable, valuable, and urgent?"
Impact vs Effort analysis. High Value + Low Effort = test first. Low Value = don't pursue — save your resources. Align with strategic objectives and calculate opportunity cost. Not every idea earns an experiment, and that's the point.
Experiment
"How do you test ideas with real customers and real data?"
Define success criteria before you start. Design experiments using pretotyping methods — Fake Door, Mechanical Turk, Pinocchio, and more. Run 2–4 week testing cycles. Gather real customer behavioural data and review evidence with stakeholders. Not surveys. Not focus groups. What customers actually do.
Action
"How do you make a data-driven decision?"
Evidence drives one of three paths:
Stop — Didn't meet criteria. Save the investment. Document learnings. Move to the next opportunity.
Pivot — Shows promise with adjustments. Design follow-up experiments. Test new assumptions.
Scale — Exceeded criteria. Build with confidence. Validated ideas flow into your existing delivery pipeline.
Both outcomes are wins. You either find your next growth driver or save millions by stopping early. That's the power of making decisions with data, not opinions.
This framework is built on pretotyping, a methodology created at Google by Alberto Savoia. Leslie Barry has spent eight years adapting it for enterprise teams, developing the training, tools, and infrastructure that make rapid experimentation repeatable at scale.
An initial sprint runs 2–4 weeks. In that time, a team can test several ideas and run multiple experiments — enough to prove the approach works in your environment. From there, organisations like PEXA (9 months, 50 pretotypes), ANU (12 weeks, 45 experiments), and Tabcorp (18 months, $12M saved) have embedded rapid experimentation as an ongoing capability.
ENTERPRISE READY
Built for Enterprise Scale
Rapid experimentation doesn't operate in a vacuum. It works within your existing governance, compliance, and delivery frameworks — not around them.
Governance & Compliance
All experiments operate under a governance framework with built-in approval workflows and audit trails. Legal review applies to experiment categories, not individual tests — so compliance doesn't become a bottleneck. We've operated in regulated industries including property exchange (PEXA), education (ANU), and health insurance (HCF).
Plugs Into Your Existing Process
Rapid experimentation covers ideation to validation. It doesn't replace your product management framework — it's the missing first step that ensures you only build things customers want. Validated ideas flow into your existing delivery pipeline, whether that's Agile sprints, SAFe, or your own stage-gate process.
Roles & Change Management
Works across Product Managers, Product Owners, SMEs, Design, Tech, Marketing, and dedicated Innovation teams. Anyone can submit an idea. We build internal champions and train teams to run experiments independently — so the capability stays after we leave.
ROI Tracking
Estimate build cost using t-shirt sizing. Track cumulative savings from stopped ideas. Measure actual revenue from validated launches. The Exponentially Platform automates ROI tracking across your entire experimentation portfolio — no spreadsheets required.
METHODOLOGY COMPARISON
Where Rapid Experimentation Fits
| Approach | Core Question | Speed | Cost to Learn | When to Use |
|---|---|---|---|---|
| Rapid Experimentation | Will customers pay for this? | Days | $0–1K | Before you build anything |
| Design Thinking | What do customers need? | Weeks | $5–50K | Understanding problems |
| Lean Startup | Can we build a viable business? | Months | $50K–500K | After initial validation |
| Agile | How do we build it well? | Sprints | $100K+ | After deciding to build |
Rapid experimentation isn't a replacement for Design Thinking, Lean, or Agile — it's the missing first step that happens before all of them. It answers the most expensive question first: "Does anyone actually want this?"
PROVEN RESULTS
Enterprise Case Studies
Rapid experimentation has been embedded in 50+ enterprise organisations since 2017. Here's what happens when teams start testing before building.
AGL
Pretotyping at scale: 1,000+ tests per year with a dedicated experimentation team that made rapid experimentation core to product development.
$7.5M saved →
PEXA
Embedded pretotyping company-wide, running 50 pretotypes in 9 months across all divisions of a regulated property exchange platform.
Company-wide adoption →
Tabcorp
Unlocked retail innovation by running faster, smarter experiments that validated ideas before committing major investment.
$12M saved in 18 months →
ANU
Ran 45 rapid experiments over 12 weeks to improve student experience, validating a chatbot that served one sixth of students.
45 experiments in 12 weeks →
RACV
Embedded pretotyping as a central method in their innovation process, killing bad ideas that were on the cusp of having millions spent on them.
150+ experiments run →
Reece
Built a cross-functional rapid experimentation program that enabled teams across the business to validate ideas independently.
Cross-functional program →
TRUSTED BY ENTERPRISES
Who Uses Rapid Experimentation?
Rapid experimentation was born at Google and is taught at Stanford University. Through Exponentially, it's been embedded in enterprise organisations across Australia and beyond.
Leslie Barry and the Exponentially team have embedded rapid experimentation in 50+ enterprise organisations since 2017, training thousands of innovators and helping teams save $30M+ by testing before building.
GET STARTED FREE
Start Experimenting Today
Free AI Idea Validator
Test any idea in 2 minutes. AI generates your Lean Canvas, hypotheses, and first experiment design.
Free Innovation Guide
Download our guide to embedding rapid experimentation in your organisation.
Free Online Course
The complete pretotyping course. Originally $400, now free. Complete in 30 days.
FREQUENTLY ASKED QUESTIONS
Rapid Experimentation FAQ
What is rapid experimentation?
Rapid experimentation is a systematic approach to testing business ideas quickly and cheaply using real customer data before building. It uses pretotyping methods to validate demand in days rather than months, so teams invest only in ideas that customers actually want.
How is rapid experimentation different from A/B testing?
A/B testing optimizes existing products by comparing variations. Rapid experimentation validates whether a product should exist at all, before any code is written. A/B testing asks "which version is better?" while rapid experimentation asks "does anyone actually want this?"
How long does a rapid experimentation sprint take?
An initial proof-of-value sprint typically runs 2–4 weeks. That's enough time to test several ideas, gather real customer data, and prove the approach works. Most organisations then move into a longer-term engagement to embed the capability across teams. Individual experiments can be designed and run in as little as 1–3 days. See our Tabcorp case study for a real example.
What tools do you need for rapid experimentation?
You can start with nothing — many experiments use ads, landing pages, or even physical materials. The Exponentially platform helps teams capture, track, and measure experiments at scale, with AI assistance for designing hypotheses and selecting methods.
Can rapid experimentation work in regulated industries?
Yes. PEXA (property exchange) and ANU (education) both used rapid experimentation successfully in regulated environments. The key is testing demand signals — whether customers want something — without building the regulated product itself.
What is the ROI of rapid experimentation?
Exponentially clients have saved $30M+ by stopping bad ideas early and generated $38M+ from validated winners. AGL alone saved $7.5M from running 1,000+ experiments per year. The ROI comes from not building the wrong thing.
Who is Leslie Barry?
Leslie Barry is the founder of Exponentially and one of the world's leading rapid experimentation practitioners. He trained directly with Alberto Savoia (creator of pretotyping at Google), guest-lectures at Stanford, was Head of Innovation at Sportsbet and ThoughtWorks, and has embedded rapid experimentation in 50+ enterprise organisations.
Ready to embed rapid experimentation in your organisation?
Talk to Leslie about pretotyping training, experimentation sprints, or implementing the Exponentially platform in your organisation.