Configure Lead Scoring in Attio

Set up the CRM data model for lead scoring: fit score, intent score, composite score, and tier fields on People and Deal records. Enable score-based automations and queries.

Prerequisites

• Attio workspace with admin access
• Attio API key or MCP server connected (see attio-mcp-setup)
• Scoring criteria defined (fit dimensions + intent signals + point values)

Steps

1. Create scoring attributes on People

Endpoint: POST https://api.attio.com/v2/objects/people/attributes

// Fit score (firmographic match)
{
  "data": {
    "title": "Fit Score",
    "api_slug": "fit_score",
    "type": "number",
    "is_required": false
  }
}

// Intent score (behavioral signals)
{
  "data": {
    "title": "Intent Score",
    "api_slug": "intent_score",
    "type": "number",
    "is_required": false
  }
}

// Composite lead score
{
  "data": {
    "title": "Lead Score",
    "api_slug": "lead_score",
    "type": "number",
    "is_required": false
  }
}

// Score tier
{
  "data": {
    "title": "Lead Tier",
    "api_slug": "lead_tier",
    "type": "select",
    "config": {
      "options": [
        {"title": "Hot", "color": "red"},
        {"title": "Warm", "color": "yellow"},
        {"title": "Cold", "color": "blue"}
      ]
    }
  }
}

// Last scored timestamp
{
  "data": {
    "title": "Last Scored",
    "api_slug": "last_scored",
    "type": "date",
    "is_required": false
  }
}

2. Write scores to a person record

Endpoint: PATCH https://api.attio.com/v2/objects/people/records/{record_id}

{
  "data": {
    "values": {
      "fit_score": [{"value": 72}],
      "intent_score": [{"value": 85}],
      "lead_score": [{"value": 79}],
      "lead_tier": [{"option": "Warm"}],
      "last_scored": [{"value": "2026-03-30"}]
    }
  }
}

3. Query leads by score tier

Endpoint: POST https://api.attio.com/v2/objects/people/records/query

{
  "filter": {
    "lead_tier": {"is": "Hot"}
  },
  "sorts": [
    {"attribute": "lead_score", "direction": "desc"}
  ]
}

To get leads above a score threshold:

{
  "filter": {
    "lead_score": {"gte": 80}
  },
  "sorts": [
    {"attribute": "lead_score", "direction": "desc"}
  ]
}

4. Batch update scores

When re-scoring all leads (e.g., after model weight changes):

for each person in people_records:
  PATCH /v2/objects/people/records/{person.id}
  body: {
    data: {
      values: {
        fit_score: [{ value: new_fit }],
        intent_score: [{ value: new_intent }],
        lead_score: [{ value: new_composite }],
        lead_tier: [{ option: tier_from_score(new_composite) }],
        last_scored: [{ value: today }]
      }
    }
  }

Rate limit: 100 requests/second on Attio API. For >1000 leads, batch in groups of 50 with 500ms delays.

5. Score decay automation

To implement score decay (reduce intent score for inactive leads), query leads where last_scored is older than 14 days and intent_score > 0:

{
  "filter": {
    "last_scored": {"before": "2026-03-16"},
    "intent_score": {"gt": 0}
  }
}

For each result, update: intent_score = current_intent_score * 0.5, recalculate lead_score and lead_tier.

Via Attio MCP Server

attio.create_attribute({
  object: "people",
  title: "Lead Score",
  api_slug: "lead_score",
  type: "number"
})

attio.update_record({
  object: "people",
  record_id: "...",
  values: { lead_score: 85, lead_tier: "Hot", last_scored: "2026-03-30" }
})

attio.query_records({
  object: "people",
  filter: { lead_tier: { is: "Hot" } },
  sorts: [{ attribute: "lead_score", direction: "desc" }]
})

Alternatives

| Tool | Method | Notes | |------|--------|-------| | Attio | REST API / MCP | Native custom attributes + query API | | HubSpot | Lead Scoring tool | Built-in scoring with Properties API | | Salesforce | Einstein Lead Scoring | AI-based or rule-based via Process Builder | | Pipedrive | LeadBooster + Custom Fields | Custom fields with Automations API | | Clarify | Schema API + computed fields | Flexible object model for custom scoring |

Error Handling

• 409 CONFLICT: Attribute already exists. Use GET /v2/objects/people/attributes to check first, then update the existing attribute.
• 422 UNPROCESSABLE_ENTITY: Invalid type or missing config. Verify the type string matches supported types.
• 404 NOT_FOUND: Record ID does not exist. Re-query the person by email before updating.
• 429 TOO_MANY_REQUESTS: Rate limited. Back off and retry with exponential delay.