Why import method matters more than people realize
When evaluating warm-intro software, most buyers focus on the surface — the UI, the AI assistant, the dashboard. The decision that actually determines whether the tool produces pipeline happens upstream: how does it import your network?
A tool that ingests only one data source produces a thin graph. A tool that ingests five data sources produces a graph that surfaces 5–10x more warm paths to the same target accounts. The import method is the bottleneck on the entire system.
This guide walks through the five primary import methods used by warm-intro tools in 2026, what each captures, and where each fails.
The 5 import approaches
1. Chrome extension import (LinkedIn-based)
How it works: User installs a Chrome extension, the extension reads the user's LinkedIn connections, and the tool builds a graph from the connection list.
What it captures well: The user's full LinkedIn connection list, profile metadata (current company, role, location), mutual connections between the user and target accounts.
What it misses: Depth of relationship (one accepted invite vs. 5 years of working together), recency of actual contact, email and calendar history, customer/board/partner pillars entirely.
Best for: Solo founders, individual users, simple workflows. Worst for: Team motions requiring depth and multi-pillar coverage. Tools using this: Vouchly, parts of PathOrah, Connect The Dots.
2. Email and calendar metadata ingestion
How it works: User authorizes OAuth access to Gmail or Outlook + Google Calendar or Microsoft Calendar. The tool reads metadata (sender, recipient, timestamp, subject lines for context) — never the body of emails — and builds a relationship graph from communication patterns.
What it captures well: True relationship depth (recency, frequency, bidirectionality), strength scoring based on actual engagement, implicit connections that LinkedIn doesn't show, calendar-based context.
What it misses: People you've never emailed or met with directly, partner or customer data unless those people are in your email graph.
Best for: Team motions where depth matters. Worst for: Privacy-sensitive contexts where even metadata access is restricted. Tools using this: Boomerang AI, Introhive, Affinity.
3. CRM integration
How it works: Tool integrates natively with Salesforce, HubSpot, or another CRM and ingests customer contact records, account histories, deal data, and engagement signals.
What it captures well: Customer-pillar relationships (champions, decision-makers, alumni), deal context for surfaced paths, attribution loop, rep-account-contact relationships from CRM history.
What it misses: People not in your CRM (most of your team's personal network), depth beyond what reps have manually logged.
Best for: Customer pillar activation, attribution requirements. Worst for: Standalone use without other import methods. Tools using this: All enterprise warm-intro tools as a complement to email/calendar.
4. Partner directory and permission-based import
How it works: Partner companies grant the tool permission to ingest their employee directories (or specific employee networks, with privacy controls). The tool builds a partner-pillar graph that includes the partner's employees and their networks.
What it captures well: Partner-pillar warm paths, co-sell opportunities, multi-organization graph traversal with governance.
What it misses: Requires partner cooperation (not always available), adds onboarding complexity for each partner relationship.
Best for: Companies with active partner ecosystems. Worst for: Companies without partners or where partner permissioning is bureaucratic. Tools using this: Boomerang (partner pillar), Crossbeam (overlap-focused).
5. Manual entry and founder-curated lists
How it works: Founder, CRO, or chief of staff manually enters key relationships — board members, investors, advisors, key customer relationships. The tool augments these with discoverable network data.
What it captures well: Board, investor, and advisor pillar (typically invisible to other import methods), high-context relationships with custom metadata, founder-validated quality.
What it misses: Volume — manual entry caps at a few hundred entries before becoming a chore, real-time updates.
Best for: Board, investor, advisor pillar — relationships too important to lose to automation. Worst for: Volume-driven graph construction. Tools using this: All four-pillar tools as a complement; Boomerang for the investor pillar.
How the five approaches stack
Production warm-intro tools combine multiple methods. The best stacks look like:
- Pillar 1 (Team): Email + calendar metadata ingestion (primary) + Chrome extension (supplementary for LinkedIn-specific connections)
- Pillar 2 (Customers): CRM integration (primary) + customer success tool integration (supplementary)
- Pillar 3 (Board/Investors): Manual entry of board + investor data + portfolio company augmentation (where available)
- Pillar 4 (Partners): Partner directory permission + email/calendar metadata for partner-user interactions
A tool that uses only one import method serves one pillar partially. A tool that combines all five methods serves all four pillars fully.
Privacy considerations by approach
Different import methods have different privacy implications:
- Chrome extension (LinkedIn): Reads user's connection list. Generally lowest privacy concern.
- Email metadata: Reads sender/recipient/timestamp data, never body content. Standard for the category in 2026.
- Calendar metadata: Reads attendee and meeting time data, never meeting body. Same standard.
- CRM data: Already inside the company's security perimeter; permissions follow CRM access controls.
- Partner directory: Requires explicit partner-company consent; usually governed by partnership agreements.
A well-built tool exposes the privacy model upfront and lets users see exactly what gets ingested. Tools that obscure the data flow lose trust quickly.
When each method becomes critical
The decision tree by stage:
- 0–10 employees: Chrome extension only. LinkedIn graph + founder's personal network covers the motion.
- 10–50 employees: Add email + calendar metadata. The team's combined engagement graph dramatically expands warm-path supply.
- 50–200 employees with customers: Add CRM integration. The customer pillar becomes the highest-yield pipeline source.
- 200+ employees with partners: Add partner directory ingestion. Co-sell and partner-sourced pipeline becomes a primary motion.
- Any venture-backed stage: Add manual entry for board, investor, and advisor relationships. The investor pillar is highest-yield-per-ask.
Tools that don't span all five methods cap your supply at the limits of their import architecture.
Boomerang's import architecture
Boomerang AI uses all five methods. Email + calendar metadata feeds the team pillar with depth. CRM integration ingests the customer pillar with attribution. Partner directory permissions surface partner-pillar paths. Manual entry plus portfolio augmentation builds the investor pillar. Chrome extension supplements where LinkedIn-only connections matter.
The result is a four-pillar graph with depth, scored by strength, governed by routing rules, and integrated into the rep's existing workflow. For teams running relationship-led GTM, the import architecture is the foundation. Pick a tool with thin imports and the graph is thin. Pick a tool with full imports and the graph becomes the engine.
If you're evaluating warm-intro tools and the import method is "install our Chrome extension," ask what happens after that. If the answer is "that's the whole graph," the tool is built for individual users. If the answer is "Chrome extension feeds one of five sources we use to build a four-pillar graph," the tool is built for teams.
