INNOVATIONOngoing practice

The Invent-Improve Spectrum

Balance incremental improvements with breakthrough reinvention

Problem it solves

stagnant innovation

Best for

Leaders in established organizations who need to manage a portfolio of innovation efforts ranging from incremental improvements to radical reinvention of value propositions and business models

Not ideal for

Pure startups with a single product where the only task is finding initial product-market fit rather than managing a portfolio

Overview

Why this framework exists

The Invent-Improve Spectrum is a strategic framework for established organizations to classify, manage, and balance their innovation efforts across a continuum from incremental improvement to radical invention. At the improve end, teams make incremental changes and tweaks to existing value propositions without substantially altering the underlying business model. At the invent end, teams design entirely new value propositions that may require radical business model changes.

The framework recognizes that each end of the spectrum requires fundamentally different mindsets, processes, risk tolerances, and success metrics. Improvement work is low risk, builds on high customer knowledge, focuses on execution, and targets modest revenue increases. Invention work is high risk, operates with potentially nonexistent customer knowledge, embraces failure as learning, and targets transformational growth. Applying improvement processes to invention work or vice versa leads to predictable failure.

Between the endpoints lies the extend zone, where organizations search for new growth engines that substantially extend the existing business model without requiring complete reinvention. Great companies manage a portfolio that covers the entire spectrum, making synergies and competitive conflicts between positions explicit. Critically, they invest in invention proactively while still successful, rather than waiting for a crisis to force their hand.

Core principles

5 total
  1. Improvement and invention require fundamentally different attitudes, processes, risk tolerances, and success metrics; never confuse the two.
  2. Outstanding companies reinvent themselves while they are still successful rather than waiting for a crisis to force change.
  3. Failure is a necessary part of the invention learning process but should not be tolerated in improvement execution.
  4. A balanced portfolio of value propositions and business models across the entire spectrum is the hallmark of resilient organizations.
  5. The exploration of new value propositions should be taken just as seriously as the execution of existing ones.

Steps

4 steps
  1. Map Your Current Portfolio
    Classify all current value proposition and business model initiatives on the invent-improve spectrum. Identify which are incremental improvements, which are extensions, and which are genuine inventions requiring new business models.
    Pro tipMost organizations discover their portfolio is heavily skewed toward the improve end. This feels safe but leaves them vulnerable to disruptive competitors.
    WarningProjects that leadership calls 'innovation' often turn out to be incremental improvements when mapped on the spectrum. Be honest about where each initiative truly falls.
  2. Assess Balance and Gaps
    Evaluate whether your portfolio provides adequate coverage across the spectrum. Identify gaps, particularly at the invent end where organizations typically underinvest. Make synergies and conflicts between initiatives explicit.
    Pro tipAsk whether any of your current cash cows could be disrupted by an initiative at the invent end. If so, that conflict needs explicit management rather than avoidance.
  3. Apply the Right Process to Each Zone
    Assign improvement projects to execution-focused processes with planning, clear metrics, and low failure tolerance. Assign invention projects to search-focused processes with experimentation, learning cycles, and high failure tolerance. Do not mix these approaches.
    Pro tipInvention projects need protection from the organization's execution-oriented immune system. Consider separate teams, reporting lines, or even physical spaces.
    WarningApplying improvement metrics like quarterly revenue targets to invention projects kills them before they have a chance to find fit. Use learning-based metrics for invention work.
  4. Invest in Invention Proactively
    Allocate dedicated resources to invention work while your current business models are still healthy. Use customer experiments rather than business plans as the primary tool for evaluating invention initiatives.
    Pro tipInvest in continuously experimenting with new value propositions and business models through many small bets rather than making big bold uncertain ones.
    WarningWaiting until a crisis forces reinvention means you are already behind. The best time to explore new models is when you do not yet need them.

Checklist

Saved in your browser

Examples

2 cases
Amazon's Three-Zone Portfolio

Amazon manages initiatives across the entire spectrum. Amazon Prime (improve) adds membership benefits to the existing e-commerce model. Kindle (extend) creates a new channel for digital content within the established model. Amazon Web Services (invent) designs an entirely new IT infrastructure value proposition for a new customer segment, requiring substantial business model expansion.

OutcomeBy maintaining active projects across all three zones, Amazon continuously strengthens its existing business while building new growth engines, ensuring that no single disruption could threaten the entire organization.
Taobao's Decade of Reinvention

Chinese e-commerce platform Taobao reinvented its business model three times in ten years. It started as a consumer-to-consumer marketplace, evolved into a platform for microentrepreneurs, and then leveraged its massive consumer base to attract big brands. Each evolution required progressively larger business model changes.

OutcomeTaobao's willingness to proactively reinvent while successful, embracing changes on its platform rather than defending the status quo, allowed it to grow from a simple marketplace into a complex commerce ecosystem that competitors could not replicate.

Common mistakes

4 traps
Applying Execution Processes to Exploration Work
Requiring detailed business plans, precise financial projections, and quarterly milestones for invention projects imposes false certainty on inherently uncertain work. It leads either to killing promising ideas too early or creating fictional plans that mask real uncertainty.
Waiting for a Crisis to Innovate
Organizations that only reinvent when forced to by declining revenue or competitive disruption start from a position of weakness. They have less time, fewer resources, and greater pressure than organizations that innovate proactively from a position of strength.
Calling Incremental Work Innovation
Rebranding routine product updates as innovation creates a dangerous illusion of strategic renewal. True invention involves exploring genuinely new value propositions and business models with unknown outcomes, not polishing existing products.
Failing to Manage Portfolio Conflicts
New value propositions at the invent end may cannibalize existing ones. This conflict is healthy and inevitable, but it must be managed explicitly rather than ignored or used as a reason to kill the new initiative.

Origin story

How this framework came to be

The framework draws from the concept of organizational ambidexterity, the idea that successful organizations must simultaneously exploit current business models and explore new ones. Osterwalder and team were influenced by Columbia Business School Professor Rita McGrath's work on transient advantages, which argues that sustainable competitive advantage is increasingly rare and companies must develop the ability to continuously create new advantages.

The spectrum was developed in response to a common challenge observed in workshops with established organizations: teams either focused exclusively on incremental product improvement or attempted to apply startup methodologies without accounting for the unique constraints and assets of their corporate context.

Source

Traced to primary
Source · BOOK
Value Proposition Design: How to Create Products and Services Customers Want (Strategyzer)
Alexander Osterwalder, Yves Pigneur, Gregory Bernarda, Alan Smith · 2014
Open source →

Related frameworks

Browse all Innovation →