OpenAI Gives Sales Teams ChatGPT for Sales: 30+ Plugins Connect CRM and Calls
Sales reps spend hours pulling customer data from CRM systems (databases of clients and deals), call recordings, and spreadsheets. OpenAI has launched ChatGPT for Sales: a Sales plugin and 30+ partner plugins, which means connections to Salesforce, HubSpot, Gong, and Clay. Through them, the chat prepares a meeting brief and writes a follow-up email after a call. Start with one workflow and measure the time saved.
AI-processed from OpenAI Blog; edited by Hamidun News
Before an important meeting, a sales manager assembles the customer picture piece by piece. The deal history sits in the CRM (software that stores clients, contacts, and deal stages), the agreements remain in the call recording, fresh news about the company has to be searched for on the web, and the presentation gets built in a separate editor. OpenAI decided to bring all of this into one window and launched ChatGPT for Sales. It is a dedicated mode for sales teams: a Sales plugin from OpenAI itself and more than 30 new and extended partner plugins. A plugin here is a connection to a third-party service through which ChatGPT reads data and performs actions on the user's behalf.
What's inside
The Sales plugin contains ready-made workflows for daily work: selecting the accounts (client companies) worth working with, preparing for a call, and writing a follow-up, meaning an email with a summary and next steps. Around it sit partner plugins organized by stage of the funnel:
- Accounts and deals: Salesforce, HubSpot, Close CRM, monday CRM.
- Prospecting: Apollo, Clay, Common Room plus ChatGPT's web research; outreach is planned in Outreach and Salesloft.
- Calls: notes from Gong, Otter.ai, Granola, or the built-in Meetings plugin.
- Materials and training: Confluence, Guru, Seismic; presentations in Gamma, Figma, Canva, and ChatGPT Slides. Customer objections can be rehearsed in ChatGPT Voice.
- Contracts: Docusign, Ironclad, PandaDoc, ServiceNow CPQ.
- Analytics: Backstory looks for risky deals, while Hex and the Data plugin explain what changed in the funnel.
How it works in practice
OpenAI describes four workflows. The first: build a priority list of accounts from CRM data and research in ZoomInfo or Clay, ask why each one fits, and immediately turn the findings into an email with a specific reason to reach out. The second: before a meeting, request a briefing on the account history, past calls, and the buyer's priorities, then build a presentation around it in Gamma or Canva.
The third: after a call, ChatGPT writes the follow-up, records the agreements, and prepares edits to the customer record. A person reviews the proposed changes, and through connectors such as Attio, the next step can be saved to the CRM directly from the chat. The fourth: risk analysis. ChatGPT reviews the CRM history and conversations with the customer and explains what is slowing the deal down.
A separate piece is dot, a persistent agent in ChatGPT that OpenAI introduced in September. It uses the same plugins and context: it prepares meeting briefs, tracks changes to deals, and brings suggestions to a human for review. Combined with Rox, dot can be asked to check a specific deal every morning, flag changes, and prepare a meeting brief or a series of emails.
Numbers, limitations, and what to do
The announcement contains only one concrete metric. The co-founder of NeoBid says that with HubSpot and ChatGPT for Sales, the company cuts manual work on researching real estate clients, enriching CRM data, and preparing personalized emails by about 65%. This is the customer's own estimate, not an independent measurement. The other testimonials, from the Michael J. Fox Foundation, Pallet, Datasite, and Sculpted, are qualitative and contain no numbers.
There are caveats too. Availability of partner plugins and features depends on the plan, region, and workspace settings. Administrators of ChatGPT Business and Enterprise can enable Sales: in Admin Console, they open the workspace, select the Plugins section, and either make the plugins available to members or install them by default. Before doing so, it is worth checking which applications the plugins can access and which actions they are allowed to take. For CRM data and contracts, this is an essential part of the setup.
The practical takeaway for product managers and founders: the market for sales tools is shifting from standalone applications to an agent layer on top of them. Salesforce, HubSpot, Gong, and Clay remain data sources, while the chat becomes the interface. If you build a sales product, think about how it will look as a ChatGPT plugin. If you lead a sales team, start small: pick one workflow, such as a pre-meeting brief or a post-call email, measure the time before and after, and only then turn on automatic morning deal checks through dot.
- Nov 21, 2026FDE job postings grew 4.6x in a year: what a forward deployed engineer actually does at the client's site
- Nov 21, 2026A swimming team captain handed ChatGPT three scheduled tasks and cut her routine work by a multiple
- Nov 21, 2026OpenAI GPT-6 Cyber Found About 35 Flaws in iPhone, Mac and Other Apple Devices
- Nov 21, 202622 Claude Agents Spent Six Weeks on Black Hole Problems — and Solved Two
Key Questions Answered
What's inside?
The Sales plugin contains ready-made workflows for daily work: selecting the accounts (client companies) worth working with, preparing for a call, and writing a follow-up, meaning an email with a summary and next steps. Around it sit partner plugins organized by stage of the funnel: Accounts and deals: Salesforce, HubSpot, Close CRM, monday CRM. Prospecting: Apollo, Clay, Common Room plus ChatGPT's web research; outreach is planned in Outreach and Salesloft.
How it works in practice?
OpenAI describes four workflows. The first: build a priority list of accounts from CRM data and research in ZoomInfo or Clay, ask why each one fits, and immediately turn the findings into an email with a specific reason to reach out. The second: before a meeting, request a briefing on the account history, past calls, and the buyer's priorities, then build a presentation around it in Gamma or Canva.
What are the key details on Numbers, limitations, and what to do?
The announcement contains only one concrete metric. The co-founder of NeoBid says that with HubSpot and ChatGPT for Sales, the company cuts manual work on researching real estate clients, enriching CRM data, and preparing personalized emails by about 65%. This is the customer's own estimate, not an independent measurement. The other testimonials, from the Michael J. Fox Foundation, Pallet, Datasite, and Sculpted, are qualitative and contain no numbers. There are caveats too.
Delegate Deep Research, Coding & Multi-Step Workflows to Nomad Agent
Deploy autonomous task execution with Nomad Agent (agent.nomadnet.ai) — automated web research, data synthesis, code generation, and MCP tool orchestration.
The AI world, distilled — once a week
Seven stories that actually mattered, hand-picked. No noise, no reposts, no press releases.
Done! Check your inbox for a confirmation.