Announcing the Buyer Intent API: See Who's Actively Researching Your Market

Four OEM partners asked us the same question in three months:

"Can you tell us who's actively researching topics relevant to our customers — at the contact level, with real filters, accessible via API?"

All four already had Bombora or 6sense. All four wanted something fundamentally different: raw topic-level research data they could feed into their own scoring models, filter by attributes they actually care about, and deliver to their customers' AI agents programmatically.

Not a black-box buying stage prediction. Not an IP-inferred company name with a surge score. Named contacts at known companies, actively researching specific topics — filterable by seniority, department, company size, revenue band, and country.

We built it. The Buyer Intent API is live.

What Problem This Solves

If you build GTM tools, account intelligence products, or AI SDR platforms, your customers need to answer one question better than anyone else:

"Who should my team talk to this week, and what should they say?"

Every data source in your stack contributes a piece of that answer. Hiring signals tell you who's building teams. Funding signals tell you who has budget. News signals tell you who's active.

But the highest-leverage piece is: who is actively in-market right now? Who went from "not thinking about this" to "actively comparing solutions" in the last 30 days?

That's what intent data provides. And the problem with existing intent providers is that they give you either:

• Company-level signals with no contact attribution (Bombora)
• Opaque buying-stage predictions you can't decompose or customize (6sense)
• Contact-level data limited to a single publisher network (TechTarget Priority Engine)

The Buyer Intent API gives you contact-level, topic-specific, filterable research signals across the B2B web. Raw building blocks, not processed outputs. You own the scoring logic. You decide what "in-market" means for your customers.

How It Works

Three endpoints. Simple architecture.

1. Browse Topics

~38,000 B2B topics organized by category and subcategory. Searchable, filterable, paginated.

curl "https://signals.autobound.ai/v1/buyer-intent/topics?q=observability&is_product=true" \
  -H "Authorization: Bearer YOUR_KEY"

The critical field: is_product. This distinguishes between someone researching "Kubernetes" broadly (general category interest) and someone researching "Datadog Kubernetes Monitoring" (specific product evaluation). Product-specific research is high intent. Category research is early stage.

This distinction lets you build tiered scoring: weight product-specific topics 3-5x higher than general category topics. Most intent providers don't expose this — they just give you a number.

2. Query Contacts

Given topic IDs, return named contacts actively researching those topics. This is where the value lives.

curl -X POST "https://signals.autobound.ai/v1/buyer-intent/contacts" \
  -H "Authorization: Bearer YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "topic_ids": ["topic_k8s_datadog", "topic_observability_platform"],
    "seniority": "director",
    "department": "engineering",
    "size_bucket": "201-500",
    "page_size": 25
  }'

Returns:

{
  "contacts": [
    {
      "name": "Sarah Chen",
      "title": "Director of Platform Engineering",
      "company": "Acme Corp",
      "domain": "acmecorp.com",
      "seniority": "director",
      "department": "engineering",
      "topics_researched": [
        {
          "topic_id": "topic_k8s_datadog",
          "topic": "Datadog Kubernetes Monitoring",
          "is_product": true,
          "category": "Cloud Infrastructure"
        }
      ],
      "employee_count": 340,
      "revenue_bucket": "$50M-$100M",
      "country": "US"
    }
  ],
  "count": 25,
  "next_cursor": "eyJsYXN0X...",
  "query_time_ms": 1243
}

Filter dimensions available:

• Seniority: c_suite, vp, director, manager, senior, mid_senior, junior, entry
• Department: engineering, marketing, sales, product, operations, hr, finance, legal, it, executive
• Company size: 1-50, 51-200, 201-500, 501-1000, 1001-5000, 5001-10000, 10001+
• Revenue: <$1M through $1B+
• Country: full ISO 3166-1 support

3. Discover Filters

Returns all valid filter values dynamically. Build your UI from this endpoint — don't hardcode enums that may change.

curl "https://signals.autobound.ai/v1/buyer-intent/filters" \
  -H "Authorization: Bearer YOUR_KEY"

Why the Two-Step Pattern Exists

We deliberately split this into "browse topics first, then query contacts." This isn't an accident.

Intent data is only as good as your topic selection. If you query "Cloud Infrastructure" broadly, you'll get thousands of contacts who are vaguely interested in clouds. Not useful.

If you query "Datadog Kubernetes Monitoring" + "New Relic APM" + "Grafana Cloud" specifically (all is_product: true), you get a tight set of people actively evaluating observability platforms. That's a qualified prospect list.

The two-step pattern forces you to think about targeting before spending credits. The result is dramatically higher signal quality than a single broad query would provide.

How It's Different (for the VP of Product comparing options)

If you're evaluating intent data sources for your platform, here's the honest comparison:

Bombora: Company-level only. IP-based inference. You get "Acme Corp has a surge score of 82 on Cloud Infrastructure." You don't get who at Acme is researching, what specifically, or what seniority level. Your sales team still has to guess who to contact and what to say. Broad reach, low specificity.

6sense: Combines IP-based signals with their own pixel data. Outputs a "buying stage" prediction per account (Awareness → Decision). Better than raw surge scores, but opaque — you can't see the underlying data, you can't feed it into your own models, and you're dependent on their scoring logic matching your customer's definition of "in-market."

TechTarget Priority Engine: Contact-level and specific, but limited to people who read TechTarget's own editorial properties. That's a narrow slice. If your prospect is researching on Reddit, vendor docs, Stack Overflow, or industry forums — TechTarget doesn't see them.

Autobound Buyer Intent API: Contact-level. Topic-specific (38K topics with product/category distinction). Full filter dimensions. Raw data accessible via API. You own the scoring. Credit-billed so you control spend. Feeds directly into AI agents via MCP. The tradeoff: it's a building block, not a turnkey dashboard. You build the experience; we provide the data.

Billing

• 1 credit per contact returned
• Daily limit: 1,000 queries per workspace (resets at midnight UTC)
• Rate limit headers on every response (x-ratelimit-daily-limit, x-ratelimit-daily-remaining, x-ratelimit-daily-reset)
• Pagination via cursor for large result sets (max 100 per page)

The Compound Value

Intent data is most powerful when combined with other signals. A contact researching "B2B intent data providers" becomes dramatically more interesting when their company also:

• Filed SEC Form D for $18M last month (they have budget)
• Posted 4 Data Partnership roles in the last 3 weeks (they're building the team)
• Their CTO spoke on a podcast about "building a data moat" (leadership priority confirmed)

That compound picture — intent + funding + hiring + executive speech — is only possible when all your signals live in the same system, queryable from the same API key.

Via the MCP server, you can ask Claude: "Find me companies where we have buyer intent signals on data infrastructure topics AND a recent funding event AND they're hiring in data engineering." One query across three data sources, synthesized answer.

Integration Patterns

Embed in your GTM product: Your customers use your tool to find prospects. Add an "In-Market Signals" tab powered by the Buyer Intent API. Each customer configures their relevant topics; you query on their behalf.

Feed into AI agents: Your autonomous SDR agent needs to decide who to email this week. Intent signals provide the "who's active right now" input. Combine with enrichment data to generate personalized outreach.

Batch scoring: Weekly cron pulls intent contacts for your top topics → cross-references against CRM pipeline → surfaces net-new in-market accounts → creates prioritized task lists.

MCP integration: Connect the Autobound MCP server to your team's Claude Code or Cursor. Ask natural language questions about who's in-market. Get structured answers you can act on immediately.

Get Started

The Buyer Intent API is live for all customers with active credits.

# Browse topics relevant to your product
curl "https://signals.autobound.ai/v1/buyer-intent/topics?q=CRM&is_product=true&page_size=10" \
  -H "Authorization: Bearer YOUR_KEY"

# Pull in-market contacts
curl -X POST "https://signals.autobound.ai/v1/buyer-intent/contacts" \
  -H "Authorization: Bearer YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{"topic_ids": ["topic_crm_hubspot", "topic_crm_salesforce"], "seniority": "vp", "size_bucket": "501-1000"}'

Self-service API access opening later this quarter.

Resources:

• Buyer Intent API docs →
• Topic taxonomy browser →
• MCP server (query via AI tools) →
• Book a demo →