Sales call prep brief: Calendar → HubSpot + Notion → Slack

Enable the Calendar API and configure OAuth consent

Google Cloud Console → APIs & Services → Library → search "Google Calendar API" → Enable. Then APIs & Services → OAuth consent screen → set App name and User type (Internal only works inside a Workspace domain; External + Testing mode is enough for a team running this on itself).

Create a Web application OAuth client

Credentials → Create Credentials → OAuth client ID → Web application, not Desktop app — the Desktop quickstart only covers a one-off interactive login, not a refresh token an unattended poller can renew silently. Build the authorization URL with access_type=offline, or the access token has no way to renew once it expires.

Scope to read-only events and poll a 30-minute window

Use calendar.events.readonly — Google's own scope list describes it as "View events on all your calendars", narrower than the calendar.readonly scope most tutorials default to. Calendar has no webhook for "a meeting starts soon" — reminders.overrides only ever fires to email or a popup — so poll every 5–10 minutes with a 30-minute window instead. Both timeMin and timeMax are exclusive bounds and need a timezone offset; singleEvents=true has to come before orderBy=startTime is accepted, or a recurring meeting comes back as one rule instead of the instance actually starting soon.

GET https://www.googleapis.com/calendar/v3/calendars/primary/events
  ?timeMin=2026-08-08T14:30:00-04:00
  &timeMax=2026-08-08T15:00:00-04:00
  &singleEvents=true
  &orderBy=startTime
  &maxResults=10
Authorization: Bearer {access_token}

The scope value

The full scope URI to request in the OAuth consent flow — the scope checkbox or `scope=` parameter needs this exact string, not the shorthand name.

https://www.googleapis.com/auth/calendar.events.readonly

Name the native shortcut before you build around it

HubSpot's own Breeze Assistant already writes a call brief: open the meeting in HubSpot's calendar view and click Prepare, and it pulls the contact's tickets, calls, emails, and CRM activity into a summary — free on every subscription. Two gaps this workflow closes: Prepare only runs when a rep remembers to click it on that specific meeting, and it only ever reads HubSpot's own objects — it has no way to see the account notes your team keeps in Notion.

Create a private app token

HubSpot → Settings → Integrations → Private Apps (also reachable via "Legacy apps" — HubSpot renamed the section but kept the old path working). Scopes: crm.objects.contacts.read, crm.objects.deals.read, and crm.objects.companies.read for the fallback lookup below. Read-only — this workflow never writes to HubSpot.

Search contacts by email

POST /crm/v3/objects/contacts/search with an EQ filter on email, using the email of each attendee the Calendar step flagged as external — email isn't an enumeration property, so matching is case-insensitive and there's no need to lowercase the address first. One request per external attendee; the search endpoint caps at 5 requests/second per account, tighter than the general private-app limit, so queue attendee lookups rather than firing them concurrently when a call has several outside guests.

curl -X POST "https://api.hubapi.com/crm/v3/objects/contacts/search" \
  -H "Authorization: Bearer $HUBSPOT_PRIVATE_APP_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "filterGroups": [{
      "filters": [{ "propertyName": "email", "operator": "EQ", "value": "lee@northwind.co" }]
    }],
    "properties": ["email", "firstname", "lastname", "company"],
    "limit": 1
  }'

Pull the deals, or fall back to a company match

Take the matched contact id into GET /crm/v4/objects/contacts/{id}/associations/deals for its deal ids, then POST /crm/v3/objects/deals/batch/read for dealname, dealstage, amount, and closedate on each. Sort what comes back by closedate and keep the two soonest-closing deals for the brief, with a line noting how many others are open — a full deal list blows the brief's length on any account with real history. If the email search in the previous step came back empty (common on a brand-new contact who isn't in the CRM yet), fall back to POST /crm/v3/objects/companies/search filtering domain EQ on the attendee's email domain, and say in the brief that the contact isn't in HubSpot yet rather than leaving that section blank. (URLs here use /crm/v4/; HubSpot has moved to date-based versions like /crm/objects/2026-03/contacts, and v4 is scheduled to stop being supported on 2027-03-30 — swap in the dated path before then.)

curl -X GET "https://api.hubapi.com/crm/v4/objects/contacts/{contactId}/associations/deals" \
  -H "Authorization: Bearer $HUBSPOT_PRIVATE_APP_TOKEN"

curl -X POST "https://api.hubapi.com/crm/v3/objects/deals/batch/read" \
  -H "Authorization: Bearer $HUBSPOT_PRIVATE_APP_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "inputs": [{ "id": "8642091" }],
    "properties": ["dealname", "dealstage", "amount", "closedate"]
  }'

Create the connection at the current URL

The old notion.so/my-integrations now redirects — create it at app.notion.com/developers/connections instead (Developer portal → Connections → New connection; needs workspace-owner permission). Under Capabilities, check only Read content; this workflow never writes to Notion.

Add the connection to every account-notes page

A new connection starts with access to nothing. Open each customer or account page or database to be matched against → "•••" → Add connections → select it by name. There's no workspace-wide grant — repeat per page or database.

Search by company name, not by keyword

POST /v1/search matches only page and data-source titles, as a plain substring with no fuzzy tolerance — it does not search page body content. Use the company name matched from HubSpot as the query, falling back to the attendee's email domain, stripped of its suffix, when no HubSpot company matched. This depends on your team naming account pages consistently with the CRM company name — a page titled with a regional suffix the CRM doesn't carry will not turn up.

{
  "query": "Northwind",
  "filter": { "value": "page", "property": "object" },
  "sort": { "direction": "descending", "timestamp": "last_edited_time" },
  "page_size": 5
}

Create the app with two Bot Token Scopes

api.slack.com/apps → Create New App → From scratch → OAuth & Permissions → Bot Token Scopes: chat:write and users:read.email. Install to Workspace and copy the xoxb- token. chat.postMessage's channel parameter accepts a Slack user id directly and opens the DM itself, so there's no im:write scope or conversations.open call to add.

Resolve the rep's own email, not every attendee's

This only DMs the rep running the call, so call users.lookupByEmail once, on the email of the attendees[] entry where self=true — that is always the rep, because this polls the rep's own primary calendar. Do not use the organizer field for this: organizer reflects who created the event, and on a call the prospect scheduled through a booking link, organizer is the prospect's email, not the rep's. Skip the multi-attendee resolution a broader meeting-brief workflow would need, since the brief never goes to the external contact. Slack answers every call, success or failure, with HTTP 200; check the response body's ok field rather than the status code to see whether the DM actually sent.

{
  "channel": "U024BE7LH",
  "text": "Call prep: Northwind — 3pm today",
  "blocks": [
    { "type": "section", "text": { "type": "mrkdwn",
      "text": "*Northwind* — call in 15 min\n\n*Deal:* Northwind Expansion · Decision Maker Bought-In · $42,000 · closes 2026-09-15\n\n*Account notes:* <https://notion.so/...|Northwind account page>" } }
  ]
}