Company A
The initiator of the match, including goals, capability boundaries, region, industry, products, and resource constraints.
Build matching systems from objects, relationships, evidence, and actions.
Company relationship matching is not simple search or black-box recommendation. It needs a clear business object model, explainable relationships, traceable evidence, and executable next steps.
This method applies to procurement leads, channel partnerships, supply-chain expansion, tender opportunities, regional intelligence, and partner discovery.
Object modeling comes before algorithms.
If the objects are unclear, collection, reasoning, and scoring will drift. Relpop fixes the business language first, then lets rules, models, and agents operate on it.
Company B
A potential customer, supplier, partner, channel, peer substitute, or policy-related organization.
Opportunity
A follow-up space triggered by demand, events, budgets, projects, policies, organizational changes, or upstream and downstream links.
Relationship
Purchasing, supply, competition, partnership, investment, regional, personnel, project, or event links between companies.
Evidence
Sources, timestamps, text snippets, structured fields, connection paths, and evidence strength.
Action
Contact priority, outreach angle, validation questions, risk notes, and observation plans.
Relationship scoring is not one score. It is a set of explanations.
Scores help ranking, but decisions need a clear answer to why this company, whether the evidence is enough, and what to validate next.
Match strength
Demand fit
Whether Company B's public demand, project movement, or business direction fits Company A's capabilities.
Relationship path
Whether there are indirect links through supply chains, customers, regions, projects, people, or events.
Time window
Whether the opportunity is actionable, expired, or not mature enough to pursue.
Trust constraints
Evidence quality
Source reliability, freshness, number of cross-checks, and conflicting information.
Risk notes
Qualification, region, competition, budget, compliance, delivery, and information uncertainty.
Action cost
Reach difficulty, missing information, validation cost, and expected follow-up route.
Recommended delivery rhythm
Prove the judgment with a small sample before expanding data and scenarios. Each stage should produce material that can be read, reviewed, and iterated.
Problem Definition
Clarify Company A's goal, the scope of target Company B, scenario boundaries, and success criteria.
Sample Report
Select 5 to 20 samples and output candidates, reasons, evidence, risks, and action suggestions.
Rule Capture
Turn useful judgments into relationship types, scoring dimensions, evidence requirements, and evaluation metrics.
System Evolution
Add collection, graphs, agents, private demos, and continuous monitoring only after the sample output works.