HubSpot + TrackLayer: server-side CRM events for B2B attribution

B2B attribution usually fails after the browser session, not during it. The ad click, landing page, and form submit are still important, but they do not tell the whole story once an account enters a real sales process. Buyers research across devices, get routed through SDR qualification, move between lifecycle stages, and often convert into revenue weeks after the first session. When only the browser event is measured, campaigns are optimized toward the easiest lead to capture rather than the lead that becomes a qualified opportunity. HubSpot is where that later truth becomes visible, because the CRM sees lifecycle state, ownership, deal creation, stage progression, and closed-won outcomes.

Feeding those CRM moments into TrackLayer closes the gap between demand generation and revenue operations. TrackLayer can preserve the click IDs, hashed identifiers, domains, and account metadata from the acquisition side, then join them to HubSpot stage changes and deal events from the CRM side. That gives paid media teams cleaner feedback loops for LinkedIn and Meta, gives ABM programs a better account-level view, and gives finance a clearer way to tie media spend to pipeline and revenue instead of vanity lead counts.

Build the bridge in layers. First prove the HubSpot webhook can reach TrackLayer. Then normalize identifiers and event IDs. Then map lifecycle stages into canonical events. Then attach deal value and stage changes. That sequence makes debugging much easier because each replay answers one clear integration question.

HubSpot Workflow → Send a webhook

POST https://edge.tracklayer.io/v1/ingest/crm
Authorization: Bearer tl_live_xxxxxxxxx
Content-Type: application/json

{
  "source": "hubspot",
  "event": "hubspot.lifecycle_changed",
  "occurred_at": "{{ date_now }}",
  "contact": {
    "id": "{{ contact.hs_object_id }}",
    "email": "{{ contact.email }}",
    "lifecycle_stage": "{{ contact.lifecyclestage }}",
    "lead_status": "{{ contact.hs_lead_status }}"
  },
  "company": {
    "id": "{{ company.hs_object_id }}",
    "domain": "{{ company.domain }}",
    "name": "{{ company.name }}"
  }
}

import crypto from "node:crypto";

function normalizeEmail(value: string) {
  return value.trim().toLowerCase();
}

function buildEventId(contactId: string, stage: string, changedAt: string) {
  return crypto
    .createHash("sha256")
    .update([contactId, stage, changedAt].join(":"))
    .digest("hex");
}

export async function forwardHubSpotEvent(payload: {
  contact: { id: string; email: string; lifecycle_stage: string };
  company?: { domain?: string };
  occurred_at: string;
}) {
  const eventId = buildEventId(
    payload.contact.id,
    payload.contact.lifecycle_stage,
    payload.occurred_at
  );

  await fetch("https://edge.tracklayer.io/v1/events", {
    method: "POST",
    headers: {
      Authorization: "Bearer tl_live_xxxxxxxxx",
      "Content-Type": "application/json",
    },
    body: JSON.stringify({
      event_id: eventId,
      event: "lead_stage_changed",
      source: "hubspot",
      occurred_at: payload.occurred_at,
      user_data: {
        email: normalizeEmail(payload.contact.email),
      },
      account_data: {
        domain: payload.company?.domain ?? null,
      },
      custom_data: {
        hubspot_contact_id: payload.contact.id,
        hubspot_lifecycle_stage: payload.contact.lifecycle_stage,
      },
    }),
  });
}

const lifecycleToCanonicalEvent: Record<string, string> = {
  subscriber: "lead_captured",
  lead: "lead_captured",
  marketingqualifiedlead: "qualified_lead",
  salesqualifiedlead: "sales_qualified_lead",
  opportunity: "opportunity_created",
  customer: "customer_won",
  evangelist: "customer_expanded",
};

function toCanonicalEvent(stage: string) {
  return lifecycleToCanonicalEvent[stage.toLowerCase()] ?? "crm_stage_changed";
}

{
  "event": "deal.stage_changed",
  "source": "hubspot",
  "event_id": "deal_84721_contractsent_2026-04-23T11:04:19Z",
  "occurred_at": "2026-04-23T11:04:19Z",
  "user_data": {
    "email": "buyer@acme.io"
  },
  "account_data": {
    "domain": "acme.io",
    "company_name": "Acme",
    "hubspot_company_id": "3029911"
  },
  "custom_data": {
    "hubspot_contact_id": "102948",
    "hubspot_deal_id": "84721",
    "hubspot_pipeline": "sales_pipeline",
    "hubspot_deal_stage": "contractsent",
    "tracklayer_stage": "proposal_sent",
    "owner_id": "55192",
    "currency": "USD",
    "value": 36000,
    "billing_period": "annual"
  }
}

curl -X POST https://edge.tracklayer.io/v1/replay \
  -H "Authorization: Bearer tl_live_xxxxxxxxx" \
  -H "Content-Type: application/json" \
  -d '{
    "event_id": "deal_84721_contractsent_2026-04-23T11:04:19Z",
    "destinations": ["linkedin", "meta"],
    "reason": "destination_timeout"
  }'

Expected result:
{
  "ok": true,
  "deduplicated": false,
  "canonical_event": "proposal_sent",
  "replayed_destinations": 2
}

The mapping table is where CRM semantics become destination semantics. HubSpot can keep its own lifecycle vocabulary, but the canonical event name should remain stable so TrackLayer can route the same CRM event into ad APIs, reporting exports, and internal analytics without rewriting the meaning every time a team updates pipeline labels.

Account-based marketing needs more than person-level conversion counting. One stakeholder may click the ad, another may become the named contact in HubSpot, and a third may be attached to the deal that eventually closes. If those records stay isolated, the campaign looks weaker than it really is. By carrying company domain, HubSpot company ID, account owner, target-account status, and deal associations into TrackLayer, the event stream can show that paid media is influencing movement inside the account even when the individual contact changes. That is the practical ABM win: better visibility into account progression, not just more individual leads.

Revenue attribution needs explicit logic, because one HubSpot deal amount can represent very different economics. For SaaS, MRR is often the cleanest optimization signal for near-term efficiency, especially when contracts start monthly and expansions happen later. ARR is more useful when the sales motion is annual and the buying decision is made on total contract value rather than the monthly equivalent. Contract value can be the right reporting metric for services or multi-year deals, but it should be handled carefully if revenue is recognized over time or if churn and downgrades are common.

The operational rule is to send the value that matches the optimization objective and keep the alternate values in the event envelope. A team might optimize paid social on first-year ARR, monitor MRR payback internally, and reserve full contract value for finance views. TrackLayer can store all three values while exposing only the destination-safe one to Meta or LinkedIn. That keeps bidding aligned with the business outcome you actually care about instead of whichever number was easiest to pull from the deal record.

HubSpot contact merges matter because deduplication changes the identity truth that attribution depends on. A buyer may appear first as a webinar lead with a personal email, later as a demo request with a work email, and eventually as the primary contact on a deal. When HubSpot merges those records, TrackLayer should ingest the merge event as an identity graph update rather than as a new conversion. The old contact IDs become aliases, the surviving record becomes the durable node, and historical events can resolve to the unified identity without double-counting the person across campaigns or destinations.

This is especially important for sales-assisted pipelines where enrichment tools, list uploads, and manual CRM cleanup are common. If merge history is ignored, matched audiences fragment, SQL and opportunity counts drift, and the same account can look like multiple net-new wins. When merge history flows into TrackLayer, identity stitching gets stronger over time instead of noisier. The result is a cleaner person-and-account graph that keeps ABM, lifecycle attribution, and warehouse reporting anchored to the same reconciled record.