Company Buyer Intent API: Account-Level Scoring and Timeline Intelligence

Company-level buyer intent with composite scoring, buying stages, and timeline trends. Raw account data for your own models, not black-box predictions.

The Gap: Contact Intent vs. Company Intent

In May, we shipped the Buyer Intent API for contact-level signals. Named people at known companies, actively researching specific topics. That answers "who should I email today?"

But for account-based teams, there's a prior question: "Which accounts should I prioritize this week?"

Contact intent gives you individual researchers. Company intent aggregates those signals upward: it tells you which organizations have multiple people actively researching your topics, how intense that activity is relative to baseline, and whether the account is trending toward a buying decision or still just browsing.

This is the layer that feeds account scoring models, triggers SDR alerts, and powers the "hot accounts" list in your platform. And now it's available as raw, filterable API data.

What Shipped

Two new endpoints, both live today:

• Companies endpoint – composite scoring + buying stage per organization, for a set of topics
• Company timeline endpoint – week-over-week intent trend for a specific company/topic pair

Plus a unified export flow that covers both contact and company data with the same status-polling pattern.

How Composite Scoring Works

The companies endpoint doesn't just count how many people at an org researched a topic. It computes a composite score that factors in:

• Breadth – how many distinct contacts at the company are researching relevant topics
• Depth – are they researching general categories ("cloud infrastructure") or specific products ("Datadog APM pricing")?
• Recency – signals from the last 7 days weigh more than 30-day-old activity
• Seniority mix – a VP + 2 directors researching scores higher than 3 junior engineers

The result is a normalized score (0-100) per company per topic set. But here's what matters: we give you the components, not just the number. You can see the underlying contact count, the seniority breakdown, and the recency distribution. Build your own weighting on top. Ignore our score entirely if you want. The raw data is always there.

Buying Stages: Awareness, Consideration, Decision

Each company gets classified into a buying stage based on behavior patterns:

The stage is derived from the same underlying data you can inspect yourself. It's a convenience label, not a black box. If your scoring model disagrees, override it. You have full access to the signals that generated the classification.

Querying the Companies Endpoint

curl -X POST "https://signals.autobound.ai/v1/buyer-intent/companies" \
  -H "x-api-key: YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "topic_ids": ["b2b_5141", "b2b_1150", "b2b_2201"],
    "filters": {
      "employee_range": "201-500",
      "country": "US",
      "min_score": 40
    },
    "page_size": 25
  }'

Response:

{
  "companies": [
    {
      "domain": "acmecorp.com",
      "name": "Acme Corp",
      "composite_score": 78,
      "buying_stage": "decision",
      "contact_count": 5,
      "seniority_breakdown": {
        "vp": 1,
        "director": 2,
        "senior": 2
      },
      "top_topics": [
        {
          "topic_id": "b2b_5141",
          "topic": "Datadog APM",
          "is_product": true,
          "contact_count": 3
        },
        {
          "topic_id": "b2b_1150",
          "topic": "Observability Platforms",
          "is_product": false,
          "contact_count": 4
        }
      ],
      "employee_count": 340,
      "industry": "Software",
      "first_seen": "2026-05-28",
      "last_activity": "2026-06-13"
    }
  ],
  "count": 25,
  "total": 142,
  "next_cursor": "eyJsYXN0X..."
}

Notice what you get back: the score, the stage, AND the breakdown. You can see that Acme has a VP researching product-specific topics as recently as two days ago. That's why it's classified as "decision." Your platform can surface this context directly to end users, or use it to trigger automated workflows.

Timeline Intelligence: Track Intent Over Time

A single snapshot tells you who's active now. But account-based motions need trend data. Is this company heating up or cooling down? Did they just enter the buying cycle, or have they been in consideration for 6 weeks?

curl "https://signals.autobound.ai/v1/buyer-intent/companies/acmecorp.com/timeline?topic_ids=b2b_5141,b2b_1150&weeks=8" \
  -H "x-api-key: YOUR_KEY"

Response:

{
  "domain": "acmecorp.com",
  "name": "Acme Corp",
  "topic_ids": ["b2b_5141", "b2b_1150"],
  "timeline": [
    { "week_start": "2026-04-21", "composite_score": 12, "buying_stage": "awareness", "contact_count": 1 },
    { "week_start": "2026-04-28", "composite_score": 18, "buying_stage": "awareness", "contact_count": 1 },
    { "week_start": "2026-05-05", "composite_score": 34, "buying_stage": "consideration", "contact_count": 3 },
    { "week_start": "2026-05-12", "composite_score": 45, "buying_stage": "consideration", "contact_count": 3 },
    { "week_start": "2026-05-19", "composite_score": 52, "buying_stage": "consideration", "contact_count": 4 },
    { "week_start": "2026-05-26", "composite_score": 61, "buying_stage": "consideration", "contact_count": 4 },
    { "week_start": "2026-06-02", "composite_score": 72, "buying_stage": "decision", "contact_count": 5 },
    { "week_start": "2026-06-09", "composite_score": 78, "buying_stage": "decision", "contact_count": 5 }
  ]
}

This is the data that powers "this account just moved to decision stage" alerts. Acme went from a single person casually browsing in April to five people (including a VP) actively comparing products in June. That's a buying signal you can act on.

Exports: Flat-Rate, Unlimited Rows

For bulk workflows (weekly syncs to your data warehouse, batch scoring jobs, full-funnel analysis), use the export endpoint:

curl -X POST "https://signals.autobound.ai/v1/buyer-intent/contacts/export" \
  -H "x-api-key: YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "topic_ids": ["b2b_5141", "b2b_1150", "b2b_2201"],
    "filters": { "employee_range": "201-500", "country": "US" },
    "include_company_scores": true
  }'

Pricing: 50 credits base + 5 per topic. Unlimited rows. Whether you get 200 results or 200,000, the cost is the same. For 3 topics, that's 65 credits total.

Integration Patterns

1. Feed Account Scoring Models

Pull company scores weekly. Combine with your existing account signals (technographics, firmographics, prior engagement). Use the composite score components (breadth, seniority mix, recency) as features in your own ML model. The raw breakdown gives you way more signal than a single opaque number ever could.

2. Trigger Alerts on Stage Transitions

Poll the companies endpoint daily. When an account moves from "consideration" to "decision," fire a webhook to your orchestration layer. The timeline endpoint lets you confirm it's a real trend, not a single-day spike.

3. Combine Company + Contact for Full Context

Company intent says "Acme Corp is in decision stage for observability tools." Contact intent says "Sarah Chen (Director of Platform Engineering) and James Park (VP Engineering) are the ones researching." Together, that's a qualified account with named contacts and topic context. Your sales team knows who to call and what to say.

# Step 1: Find hot accounts
curl -X POST "https://signals.autobound.ai/v1/buyer-intent/companies" \
  -H "x-api-key: YOUR_KEY" \
  -d '{"topic_ids": ["b2b_5141"], "filters": {"min_score": 60}}'

# Step 2: Get the people at those accounts
curl -X POST "https://signals.autobound.ai/v1/buyer-intent/contacts" \
  -H "x-api-key: YOUR_KEY" \
  -d '{"topic_ids": ["b2b_5141"], "company_domains": ["acmecorp.com"], "seniority": "director"}'

4. Power "Hot Accounts" in Your Product

If you build a GTM platform, this is the data behind a "Trending Accounts" widget. Your customers configure topics relevant to their business. You query the companies endpoint on their behalf. Surface the accounts, the scores, and the timeline charts. Your product gets stickier because the data refreshes constantly.

How This Differs from 6sense / Bombora

The core difference is philosophical: we give you building blocks, not finished predictions.

• Bombora gives you a "surge score" per company per topic. You can't see what's underneath it. You can't weight seniority differently. It's a number. Take it or leave it.
• 6sense gives you buying stage predictions per account. Better than a raw number, but still opaque. Their model decides what "decision stage" means.
• Autobound gives you the full decomposition: contact count, seniority breakdown, topic specificity, recency distribution, product vs. category split. We ALSO give you a composite score and stage label for convenience. But you can ignore our labels and build your own model. The raw data is always accessible via API.

For OEM partners who integrate intent data into their own platforms, this matters. Your customers have different definitions of "in-market." Your scoring model should reflect YOUR understanding of your market, not ours. We provide the ingredients. You cook the meal.

Numbers

• 37,000+ B2B topics (same taxonomy as contact intent, with is_product distinction)
• Billions of intent signals processed
• Company scores refresh weekly
• Timeline data available up to 12 weeks back
• Export pricing: 50 credits + 5/topic, unlimited rows
• 99.9%+ API uptime through May and June

What's Next

Company intent is heading to full GA with composite scoring across all 37K topics. We're also building real-time webhook delivery for stage transitions, so you don't have to poll. When an account moves to "decision," you'll know within minutes.

Resources

• Buyer Intent API docs (full reference)
• Contact-level Buyer Intent API (companion post)
• June 2026 What's New roundup
• API Explorer (try it live)
• Full signal catalog
• Book a demo

Questions or integration help? Reach out directly. We're shipping fast and building what partners actually need.