Why Starting Small Is Often the Quickest Path to Improvement at Full Scale

It is one of the most common tensions in improvement work: leaders face an urgent challenge. Teams are eager to act. Funders want immediate results. The instinct is understandable: if you have some good ideas, why not implement them everywhere as quickly as possible?

Yet, at IHI, we have found through our work supporting large-scale improvement efforts across health systems, communities, and countries that one of the fastest ways to slow improvement is to attempt to go to full scale too soon.

Counterintuitive as it may seem, starting small is often the quickest path to achieving improvement at full scale.

If you are skeptical of this approach, consider your own organization’s experience. Are there instances when it moved to scale too quickly and paid the price? Consider two high-profile examples from other sectors.

• Segway: Segways were hyped as being the next generation of urban transport. Before launching, Segway built massive production facilities capable of churning out tens of thousands of units per month, believing that the vehicle would fundamentally restructure urban transit. However, because they did not run small-scale tests, they overlooked massive hurdles, such as street legalities, high price barriers, and consumer social stigma. The company ultimately had to end production of this product.
• Denver Airport Luggage Handling System: Denver attempted to launch the world’s first fully automated airport baggage handling system. Rather than testing at a small scale and then expanding, the project decided to launch simultaneously airport-wide. This delayed the airport opening by 16 months, at an extra cost of $400 million.

It is commonly accepted that pharmaceutical companies must start with small-scale tests when developing new medicines, and technology companies need to do the same when developing new software or apps. What can we learn from these sectors? 

The Pressure to Move Fast

When faced with a significant health-related challenge, leaders understandably want to accelerate adoption of promising practices. The challenge is that a promising idea is not the same thing as a scalable intervention.

An untested idea that seems like a game changer might face barriers to implementation at scale, or might not even be feasible. An approach that showed improvement at one site may fail at others. Workflows that seem straightforward may create unintended burdens in practice. Measures intended to demonstrate improvement may prove difficult to collect. Embedding the intervention within existing workflows may prove challenging. The extensive training and support that allowed one team to implement the practice likely isn’t scalable; providing thousands of teams with the same level of support isn’t feasible.

These realities are not signs of failure. They are signs that more learning is needed to understand what it will take to bring improvement to scale.

Small-Scale Testing as a Rapid Learning Strategy

At IHI, we view small-scale testing as a way to rapidly learn how to achieve is needed for improvement to succeed at scale.

Rather than investing heavily in broad implementation before key questions have been answered, teams start by testing small changes. By doing so, we can learn quickly, adapt interventions, strengthen implementation approaches, and build confidence in both the intervention and the supporting system required for success.

Small-scale testing helps organizations answer critical questions:

• Is implementation feasible?
• Does implementing the interventions result in improvement?
• What barriers will teams encounter? How can we address them?
• What infrastructure is required to support implementation? 
• Which elements must remain standard, and which can be adapted?
• How can the intervention be spread across the system?
• How can improvement be sustained over time?

Every unanswered question becomes a risk during scale-up. Small-scale testing allows organizations to identify and address those risks while the cost of learning remains relatively low.

Why Scaling Too Quickly Slows Progress

Organizations often assume that skipping small-scale testing will save time. In practice, skipping this step often produces avoidable and costly delays.

When interventions are implemented broadly before they are sufficiently tested and refined, we find:

• Costly rework and redesign
• Staff frustration and resistance
• Low reliability of implementation
• Variation in results across sites, or no measurable improvement
• Additional training or retraining needed
• Loss of stakeholder confidence
• Delays associated with correcting avoidable problems

What appears to be acceleration at the start of a project often results in slower progress later — or no meaningful improvement at all. 

In other words, organizations can invest time in learning before scale-up, or spend it troubleshooting problems, managing delays, and redesigning during scale-up.

Small-Scale Testing Does Not Mean Slow

One misconception is that small-scale testing requires lengthy pilot periods, but effective improvement teams do not run pilots that last for years. Instead, they conduct rapid testing and learning cycles. 

At IHI, we are continually working to shorten these cycles so teams can build confidence in the interventions and understand what will be needed at scale. This approach reduces risk, improves reliability, and enables teams to develop a scalable model more quickly. 

A Better Definition of Speed

The true measure of speed is how quickly an organization reaches sustainable improvement at scale — not how quickly it launches an intervention broadly. The organizations that achieve the greatest impact are not necessarily those that move to full scale first; they are the ones that learn fastest.

The irony is that while starting small can feel slower, it is often the shortest path to getting where we ultimately want to go: improving health outcomes at scale.

Paul Howard is the Lead Advisor of Innovation, Content, Design, and Methods at IHI. 

Photo by Heather Wilde on Unsplash

You may also be interested in:

• A Framework for Spread: From Local Improvements to System-Wide Change
• Evolving the IHI Scale-up Framework
• Lessons Learned from Ghana's Project Fives Alive! A Practical Guide for Designing and Executing Large-Scale Improvement Initiatives