Apollo, the sales data and outreach platform, used its annual ApolloNEXT conference on 30 September to unveil three products: Builder Studio, the Intelligence Layer and Messaging OS. Apollo's announcement presents them as one connected system for prospecting, deciding who to contact and sending the outreach. For small teams comparing AI sales agents, the practical question is what the new pieces do, when they arrive and how they will be metered.
What Apollo announced
According to the company's own release, the three products are:
- Builder Studio: a natural-language builder. Teams describe the tools and workflows they want, and Apollo says it turns them into working pages, databases and automations on its data. The release says early beta users are building custom workflows.
- Intelligence Layer: a context engine that includes an Intelligence Profile, a continuously updated view of each contact and account, and the GTM Harness, an AI engine meant to work out which accounts matter and what to do next.
- Messaging OS: a signal-based execution system that recommends who to reach, when and on which channel. It adds Email Campaigns, for marketing emails sent from Apollo, a Bulk Send API so agents can run campaigns, and Ad Audiences, which push Apollo lists to Google Ads.
On timing, Apollo says Builder Studio and Messaging OS will be available soon, while the Intelligence Layer is available now. Apollo's Intelligence product page describes the engine as goal-based and supervised, saying it stays supervised until the customer decides it can act independently.
The numbers are Apollo's own
The release makes several growth claims: more than 5 million users, a user base that doubled year over year, and more than 600,000 companies. It also says teams actively using its MCP connection grew more than fourfold between March and August, and that daily credit consumption through it is up nearly tenfold. These are company figures from a press release, and we did not find independent verification of them. The release also says Apollo wants to work inside tools customers already use, such as Claude, ChatGPT, Salesforce and Slack.
The launch follows Apollo's September announcement that it is a launch data partner for Salesforce's outbound agent, Hunter. That release says Hunter was in pilot, with general availability planned for November, and that Salesforce customers can reach Apollo's contact data through the agent without a separate Apollo subscription. Taken together, Apollo is positioning itself both as a destination platform and as a data supplier to other vendors' agents.
Why it matters for buyers: credits
The announcement does not give a price or credit cost for Builder Studio or Messaging OS. What buyers can read today is how Apollo meters its existing features. The pricing page FAQ says credits are pooled across a team, granted at the start of each billing cycle and reset each cycle. It states that unused credits do not roll over, and that this applies to purchased add-on credits as well as plan credits.
The same page says add-on credits are charged immediately and are recurring: they renew automatically each billing cycle until removed. To stop future charges, an admin drags the slider back to zero before the renewal date. It also says features labelled Unlimited may still be subject to a Fair Use Policy.
Apollo's credit usage help article, updated on 6 October, lists what can consume credits: email and phone data, enrichment, waterfall enrichment, AI research (including the AI assistant, AI custom fields and AI email writer), email infrastructure and, on some plans, phone calls. For calls, it says a thirty-second minimum applies to each connected call, with longer calls rounded up to the next thirty seconds. It also says usage can be broken down by surface, including the web app, MCP, API and CLI, and that a separate AI runs tab shows credits used per run.
That matters because the Intelligence Layer is built around agents that act on a goal. If the system plans and runs multi-step work on a schedule or when a signal fires, as the product page describes, consumption could be less predictable than a rep clicking a button. The product page says usage controls keep credit consumption in the customer's hands, and the pricing page says admins can set per-user credit caps. Neither page says how the new features will be charged.
What to check before relying on it
- Ask for the credit cost of each new feature in writing before it goes live on your account, especially anything that runs automatically.
- Set per-user credit caps and review the credit usage page after the first billing cycle with an agent switched on.
- Check add-on credits. They recur until removed, so a one-off top-up can become a standing charge.
- Understand plan rules. The pricing page says all seats on an account share one plan, downgrades take effect immediately with no refund for unused time, and cancellation takes effect at the end of the plan term. Moving from annual to monthly means waiting for the annual term to expire.
- Keep humans approving outreach. The product page lists opt-out management, do-not-call list checks and email validation as compliance tools. Rules on marketing email, calling and uploading audience lists to ad platforms vary by country, so speak to a qualified adviser rather than relying on a vendor feature list.
- Do not pay for beta promises. Two of the three products are described as coming soon. Compare what is available today.
Is it worth paying attention to?
For a small business, the pitch is consolidation: one system for data, sequencing, email campaigns and ad audiences instead of several tools. That could lower costs, but only if credit use stays predictable. Apollo's free plan, which its pricing page says has no time limit, is a low-risk way to see what is included before committing. Where outreach volume is small, a simpler tool may do the job, and an AI agent layer is optional rather than essential.
