Back to Work

Relationship Intelligence

Intro Map Pipeline

Given a target list and a source network, this pipeline finds the highest-quality warm introduction path to every target — scored on reach, relevance, responsiveness, and relationship strength.

500+

Connectors scored

4

Scoring dimensions

2.3

Avg hops to target

7-sheet

Excel deliverable

The Problem

Cold outreach has a 1–3% response rate. Warm introductions convert at 30–50%. The math is clear: if you can find someone who knows your target and is willing to introduce you, you're 10–50x more likely to get a meeting. But manually mapping introduction paths through a large network is prohibitively time-consuming. For a 100-target campaign, you'd spend weeks cross-referencing LinkedIn connections, mutual contacts, and shared affiliations.

How It Works

Step 1: Network Ingestion

The pipeline starts by building a complete map of the source network. LinkedIn Sales Navigator exports provide the primary dataset: first-degree connections with their own connection lists, shared connections, group memberships, and interaction history. This data feeds into the Identity Graph (separate case study), which resolves duplicates and enriches profiles.

Step 2: Target Matching

For each target on the prospect list, the pipeline searches the Identity Graph for paths. A path is any chain of connections from the source network to the target, with each link representing a real relationship (shared connections, co-investor history, same company, shared group membership).

Most useful paths are 1–2 hops: “You know Alice, Alice knows Bob (your target).” Three-hop paths exist but are generally too weak for a genuine warm introduction.

Step 3: Connector Scoring

When multiple paths exist to a target (common for well-connected targets), the pipeline ranks connectors on four dimensions:

DimensionWhat It MeasuresHow It's Computed
ReachDoes the connector actually know the target?Connection degree, shared connections count, interaction recency
RelevanceDoes the connector operate in a relevant context?Industry overlap, role similarity, shared deal history
ResponsivenessWill the connector actually respond to a request?Prior interaction frequency, message response patterns, engagement recency
Relationship StrengthHow strong is the source's relationship with the connector?Connection duration, mutual connections, interaction depth, shared context

Each dimension is scored 0–100 and combined into a weighted composite score. The weights are configurable per campaign — a fundraising campaign might weight Relevance highest (investor context matters), while a sales campaign might weight Responsiveness highest (you need the intro to actually happen).

Step 4: Outreach Sequence Generation

For the top-ranked connectors, the pipeline generates personalized outreach sequences:

  • Introduction request message (customized with shared context between source and connector)
  • Forwardable blurb about the source (designed for the connector to forward to the target)
  • Follow-up cadence (3-touch sequence over 10 days)
  • Alternative paths if the primary connector doesn't respond

The Deliverable

Output is a 7-sheet Excel workbook designed for immediate action:

  1. Target List: All targets with match status (path found / no path / multiple paths)
  2. Ranked Connectors: Best connector per target with composite score breakdown
  3. Full Path Details: Complete hop-by-hop paths for each target
  4. Connector Profiles: Enriched profiles of all recommended connectors
  5. Gap Analysis: Targets with no viable path — candidates for cold outreach or network building
  6. Data Quality: Source freshness, confidence scores, and resolution statistics
  7. Methodology: Scoring weights, data sources used, and coverage metrics

Integration with Identity Graph

The Intro Map Pipeline is the primary consumer of the Identity Graph. The graph provides the relationship data; the pipeline provides the intelligence layer on top. When the graph adds new data sources or resolves more entities, the pipeline automatically benefits — more paths become visible without any pipeline code changes.

Results

  • 500+ connectors scored across initial campaigns
  • Average 2.3 hops to target (within warm-intro range)
  • Path coverage: 70–80% of targets had at least one viable introduction path
  • Campaign response rates on intro-assisted outreach: 35–45% (vs. 2–3% cold baseline)

Stack

PythonNeo4jLinkedIn Sales NavigatorAirtableOpenAIpandasopenpyxl