Existence Check

Overview

Systematically check if a solution or similar solution already exists before investing effort in creation

Steps

Step 1: Clarify search target

Define precisely what we’re looking for:

1. State the core need in one sentence
2. List must-have characteristics (non-negotiable)
3. List nice-to-have characteristics (desirable)
4. Identify what would make something “close enough”
5. Define the threshold for acceptable adaptation effort

Step 2: Search for exact matches

Search for solutions that match exactly:

1. Search within current project/organization (local scope)
2. Search public repositories and package managers
3. Search commercial solutions and SaaS offerings
4. Search academic papers and research implementations
5. Document each match with: name, source, fit assessment

Step 3: Search for similar matches

If no exact matches, search for near matches (90%+ overlap):

1. Relax one must-have at a time and re-search
2. Search for solutions to slightly different problems
3. Look for solutions with different interfaces but same function
4. Check for solutions in different languages/frameworks
5. For each match, document: what’s missing, adaptation effort

Step 4: Search for partial and analogous matches

Broaden search further:

1. Search for partial solutions (50-90% overlap)
   • What addresses part of the problem?
   • What could be extended to cover our needs?
2. Search for analogous solutions from other domains
   • How do other fields solve similar problems?
   • What patterns exist that could be applied?
3. Search for component building blocks
   • What pieces exist that could be assembled?
   • What libraries/tools handle sub-problems?

Step 5: Evaluate findings

Assess all findings against our needs:

1. Rank all matches by fit score (exact > similar > partial > analogous)
2. For each viable option, estimate:
   • Time to use/adapt vs time to build from scratch
   • Risk of using existing vs building new
   • Maintenance burden of each approach
3. Consider combinations (multiple partial solutions)
4. Identify what’s truly missing that requires new creation

Step 6: Generate recommendation

Based on evaluation, recommend approach:

• use_existing: Exact or near-exact match found, use as-is
• adapt_existing: Good match found, adaptation effort acceptable
• combine_multiple: No single solution, but combination works
• build_new: Nothing suitable exists, must create
• hybrid_approach: Use existing for some parts, build others

Provide clear rationale for recommendation.

When to Use

• Before creating something new (feature, document, process, tool)
• At the start of research phase in any project
• When looking for reusable solutions to avoid reinventing the wheel
• Before writing code that might already exist in libraries
• When exploring market or competitive landscape
• Before designing a system that might have established patterns
• When considering building vs buying/using existing
• At strategy discovery to find proven approaches
• When time or resources are limited and reuse is valuable
• Before patent/IP work to understand prior art

Verification

• Multiple search sources were checked (not just one)
• Search scope was appropriate (not too narrow, not wastefully broad)
• Each match type was attempted before concluding “not found”
• Findings include source URLs or references for verification
• Adaptation effort estimates are realistic (not overly optimistic)
• Recommendation rationale addresses build vs use tradeoffs
• Cross-domain search was attempted for analogous solutions