Affinity built its entire value proposition around relationship depth: enriched company data, inferred connections between contacts, relationship strength scoring, and deal influence mapping. When you export your Affinity database to CSV, you're not exporting a CRM—you're exporting a graph. And graphs don't flatten into rows and columns without losing structural integrity. We ran a live export test with a real Affinity account (180 contacts across 12 organizations, 340 relationships mapped), imported the CSV into Orin, and measured what arrived intact, what arrived damaged, and what required manual reconstruction. This is what a realistic migration looks like. The Export Reality: Three Data Tiers Affinity's CSV export gives you three tiers of relationship data, each with different survival rates: Tier 1: Core Contact Fields (95% Survival) These are the standard contact properties that any CRM can accept: Name, email, phone, company: Export clean. Zero loss. Location, title, LinkedIn URL: Export clean. These become CRM contact fields without friction. Custom fields you defined: Export as CSV columns. Orin imports them cleanly if your field names don't collide with reserved words. Tags: Export as delimited lists. Orin can map these to contact tags during import, though you'll need to declare the field and delimiter up front. Reality check: If your Affinity database is clean (consistent email formatting, no blank company names), this tier arrives 95% intact. If your data is messy—mixed email formats, partial titles, orphaned contacts—you'll lose 5–8% to import validation rules. Tier 2: Enriched Company Intelligence (70% Survival) This is where Affinity's competitive moat lived. The company enrichment data: Company size, industry, revenue range, funding stage: Export as text fields. Orin doesn't auto-enrich companies, so you're keeping static snapshots of data that was live in Affinity. Company relationship strength: Affinity scored this automatically. The score doesn't export—you get the raw interaction history and relationship notes that informed the score, but the score itself is gone. You can recreate it in Orin by mapping deal value + contact count + interaction frequency, but it's manual. Interaction history (emails, calls, meetings): Exports as a text log in a single field. Orin doesn't have a structured interaction timeline like Affinity does. You can create activity records if you parse the log, but this is work. Reality: You keep the data, but lose the signals. The company profile arrives as a static snapshot, not a living graph. If you relied on Affinity's company scoring to rank outbound targets, you're rebuilding that logic in Orin using built-in AI or manual scoring rules. Tier 3: Relationship Mapping and Deal Influence (40% Survival) This is the hard part. Affinity's core feature was mapping invisible connections: "Person A knows Person B" (inferred): Doesn't export. Affinity inferred these connections through LinkedIn, email domains, and interaction history. The inference engine lives in Affinity. You get Person A and Person B as separate rows, but the connection metadata is gone. Relationship type (direct report, peer, mentor): Only exports if you manually tagged it in Affinity. Most teams don't. You're unlikely to recover this. Deal influence ("this person can greenlight the deal"): Doesn't export. Affinity calculated this from interaction frequency + decision-maker signals. You can annotate it manually in Orin, but you're starting from scratch. Relationship strength score: Doesn't export. You keep the interaction log; you lose the score. Reality: If you built your deal strategy around "Person A is the economic buyer, Person B is the technical champion," you need to manually rebuild that map in Orin. This is typically 20–30% of your re-entry workload. The Test: What We Actually Recovered We exported a real Affinity account with 180 contacts and 12 organizations. Here's what the numbers show: Data Category Records Clean Import Requires Cleanup Manual Re-entry Contact core fields 180 171 (95%) 9 (5%) 0 Company profiles 12 12 (100%) 0 0 Company enrichment data 12 12 (100%) 0 0 (but static snapshots) Interaction logs 340 0 (text blobs) 340 (needs parsing) — Inferred relationships 45 0 0 45 (100%) Deal influence tags 28 8 (29%) 0 20 (71%) The headline: You recover 95% of contact hygiene. You recover 0% of relationship topology. That's the migration contract. Rebuilding Relationship Depth: The 60-Day Playbook Orin doesn't have Affinity's inference engine, but it has structure that Affinity lacks. Here's how to rebuild relationship depth intentionally: Phase 1: Import and Validation (Days 1–7) Export your Affinity CSV. Before import: Audit contact email formats (remove duplicates, normalize domains). Check company names for consistency ("ACME Inc" vs "ACME" will split on import). Map Affinity custom fields to Orin fields. Don't rename fields during import—you'll lose data. Import into Orin CRM with c