You've built a relationship graph in Affinity. Three years of notes, connection chains, meeting history, and implicit relationships between people, companies, and deals. Then you export it. The CSV lands in your inbox. You open it. What stares back is flat—a row for each contact, yes, but the connective tissue is gone. The relationship to their founder. The deal they influenced two degrees away. The meeting note that explains why they won't take a call in Q4. It all evaporates. This isn't a bug. It's the architecture of Affinity colliding with the architecture of CSV. Affinity stores relationships as graphs—interconnected nodes where one person links to another, to a company, to a deal, to a conversation. CSV is a table. Tables don't have edges. If you're moving off Affinity, you need to know exactly which relationships die, why, and how to rebuild them fast in a tool that actually preserves them. This playbook covers all three. What Affinity's export actually loses Affinity's strength is that it stores relationships as first-class objects. Not fields on a contact record—actual edges in a graph. When you export to CSV, Affinity gives you this: Contact records: Name, email, phone, company, custom fields Company records: Name, industry, size, website Deal records: Title, value, stage, owner Interaction logs: Date, type, attendees (as a text field, not links) What you do not get: Relationship types: Who is connected to whom, and in what capacity (founder, investor, board member, customer). Affinity stores these as explicit edges. CSV has no way to represent them. Implicit network effects: Deal A involves Person X at Company Y. Deal B involves Person Z at Company Y. Affinity knows these are related through the company. Your CSV does not. You have to infer it yourself. Conversation-to-relationship mapping: Meeting notes in Affinity are anchored to relationships. When you export, notes become text blobs attached to a contact record. The context of why that meeting mattered to a specific deal vanishes. Influence chains: Affinity tracks who influenced a deal and in what order. A CSV export gives you a list of people connected to a deal, but not the sequence or strength of influence. Historical relationship depth: How long have you known this person? How many interactions? Which conversations shifted the deal? These are implicit in Affinity's graph. They're invisible in a CSV. Warm introductions and second-degree connections: Affinity excels at showing you who knows whom at a target company. The export does not preserve these paths. Why relationship graphs matter—and why losing them costs This isn't abstract. Here's what happens when your relationship graph flattens: Sales velocity drops. Your reps lose context. A contact record shows they work at Company Y, but your rep doesn't know if they're a founder, an investor, or an account manager—and the export won't tell them. They have to re-learn the org chart through cold outreach. Deal timing gets worse. You lose the implicit markers of deal health: the conversation that signaled a buyer was warming up three months ago, the coffee meeting that turned into a pilot, the email thread where someone else at the company expressed skepticism. Your rep now starts from zero. You stop finding warm intros. One of Affinity's superpowers is showing you "your network has three degrees of connection to the VP of Engineering at Target Inc." That disappears in a flat export. Your team defaults to cold outreach, which has a lower close rate and longer cycle time. Relationship data decays immediately. The moment you export, the relationships are static. In Affinity, relationships are living objects—you can add context over time, link new deals, update influence. In a CSV, you get a snapshot. The moment you move to a new tool, that snapshot is obsolete unless you manually rebuild it. The 60-day rebuild playbook The good news: you can rebuild relationship depth fast if you have a process. This playbook assumes you're moving to a CRM that actually preserves relationship graphs (like Orin , HubSpot, or Pipedrive). The process has five phases. Phase 1: Audit and map the export (Days 1–5) Before you do anything, understand what you've got: Open your Affinity export. Count columns. Note which are Affinity custom fields and which are standard (email, phone, company, etc.). Run a relationship density audit. For 20–30 of your highest-value deals, manually list the people involved and note the relationship type (founder, investor, customer, champion, influencer, skeptic, etc.). This tells you how much relationship context you're starting with. Flag warm intro chains. For deals that moved in the last 12 months, trace how they started. Who introduced you? Was it direct, or through someone else? Mark these manually in a spreadsheet. This is data that won't export; you're capturing it before it's gone forever. Map the new tool's relationship structure. Whatever CRM you're moving to has a different gra