Keep, Build On, or Rebuild
Classify an existing system before deciding how radically to change it
- Difficulty
- Easy
- Time to result
- ~days to results
- Steps
- 6
- Confidence
- 99%
This decision framework places an existing product, protocol, or market into one of three action categories. Keep applies when the current system delivers a strong user experience and has no urgent structural deficiency. Build on applies when the foundation works but specific constraints can be improved with new technology. Rebuild applies when the system's incentives, business model, interaction patterns, or user outcomes are fundamentally unsuited to the future. The assessment starts with present user value, not enthusiasm for a new technical paradigm. It then distinguishes repairable limitations from structural failures and identifies which new capabilities address those failures. The result is a bounded innovation posture that avoids wasteful imitation, unnecessary disruption, and the reflexive creation of decentralized copies of effective products.
Origin
Extracted from Marketing Against The Grain when Kip introduced a game for deciding how Web3 should affect payments, search, social media, and streaming video.
Core principles
- 01Not every existing system needs disruption
- 02Working foundations should be preserved
- 03Specific deficiencies call for extension rather than replacement
- 04Fundamental failures justify rebuilding
- 05New technology should solve a real problem
How to run it
- 1
Define the system
Choose the existing product, protocol, channel, or market to evaluate and specify its core user job.
Pro tip Assess one bounded system at a time.
Watch out An overly broad category can contain components that deserve different classifications.
- 2
Inventory what works
Document the capabilities and experiences users already value and would lose in a careless redesign.
Pro tip Use observed behavior and retention as evidence.
Watch out Do not mistake market dominance alone for user satisfaction.
- 3
Diagnose deficiencies
Separate minor limitations from fundamental failures in incentives, experience, ownership, access, or business model.
Pro tip Ask whether the problem can be corrected without replacing the foundation.
Watch out Do not label a system broken merely because it is not decentralized.
- 4
Choose the action
Select keep for a strong system, build on for a sound foundation with addressable gaps, or rebuild for structural failure.
Pro tip State the classification in one sentence tied to user value.
Watch out Avoid hybrid labels unless you can identify which components receive each treatment.
- 5
Match technology to need
Introduce new technology only where its capabilities directly solve the diagnosed problem.
Pro tip Specify the mechanism, such as faster settlement, ownership, permissioning, or portability.
Watch out Recreating a Web2 product with tokens is not inherently innovation.
- 6
Test the future experience
Compare the proposed result with the current experience and verify that it is materially better for target users.
Pro tip Prototype the riskiest assumption before committing to a rebuild.
Watch out A structurally elegant replacement can still lose if its user experience is worse.
In the wild
Existing payment infrastructure works well for many purchases, so it should not be discarded. However, cross-border transfers, taxation, recurring payments, settlement speed, and instant fund transfers remain difficult. Cryptocurrency infrastructure can extend the working foundation by solving those bounded problems.
→ Payments are classified as build on rather than keep unchanged or rebuild completely.
Google has monopoly-related problems and sometimes uses publisher content in ways marketers dislike, but its core search experience remains extremely effective for ordinary users. A replacement that does not outperform that experience is unlikely to gain adoption.
→ Search is classified as keep for now despite acknowledged deficiencies.
Social platforms face interconnected problems involving anonymity, accountability, advertising dependence, noisy interactions, and outdated modes of socializing. Because these issues concern the model itself rather than one missing feature, the proposed action is to rebuild the social web.
→ Social media is classified as a fundamental rebuild opportunity.
Common mistakes
Rebuilding because technology is fashionable
Creating a Web3 version of a successful Web2 product without a new use case wastes effort and often worsens the user experience.
Ignoring structural failures
Incremental features cannot repair a system whose incentives or business model cause the central problem.
Discarding a strong foundation
A complete rebuild can destroy capabilities users already value when a targeted extension would have been sufficient.
Is it for you?
Best for
It is best for founders and marketers deciding where emerging technology should preserve, extend, or replace an existing system.
Not ideal for
It is not ideal when the team lacks enough evidence about current user behavior and system performance to make a classification.
From the transcript
“Let's play a game. And this game is keep, build on, or rebuild.”
“And let's take things that are in the current world that we use and decide whether we should they're great, let's just keep them as…”
“Like, we need to build new things.”
From the episode
Is Web 3 Coming?