Salesforce's Hunter is an outbound sales agent built to work deals over weeks, with GA listed for November. Set the guardrails before it starts making a dead deal look alive.
Hunter is Salesforce's outbound sales agent, announced September 11, 2026, that works a pipeline from research to outreach over weeks. It's the first agent on Agentforce's long-horizon runtime, with memory, durable execution and dynamic steering. It's in pilot with GA listed for November 2026. Teams should define what counts as deal progress and require human approval for first contact, pricing and stage changes.
At Dreamforce last month, Salesforce introduced a set of "job-ready" agents, and one of them is aimed squarely at sales reps. Salesforce's September 11 announcement describes Hunter as "your outbound sales agent," built to work a pipeline "from research to outreach, collaborating with sellers over weeks and months." It's in pilot now, with general availability listed for November '26.
That's not quite this week's news. It's about five weeks old. But with GA a few weeks away, it's the right time for sales teams to think about it, because Hunter is a different kind of tool from the email writers and call summarizers most reps have already tried.
Most AI sales tools so far have worked one task at a time. Write this email. Summarize this call. Research this account. You ask, it answers, the session ends.
Hunter is the first agent Salesforce says runs on a new "long-horizon runtime," meant to let agents pursue goals across days and weeks rather than finishing one interaction. Salesforce's own example: a seller asks Hunter to rescue their at-risk deals before the end of the quarter. Hunter turns that into a measurable goal, builds a plan, works out which tasks, tools and context it needs, and sets guardrails for when it can act on its own and when it needs the seller's approval.
Salesforce lists three pieces underneath:
The same announcement includes a customer figure: Salesforce says 60% of Perk's sales pipeline is built by its outbound sales agent, Hunter. That's a vendor-reported number about one customer, so read it as a sign of what's possible, not what your team will see.
There's a second agent in the same lineup worth knowing about. Piper, the inbound pipe gen agent, works across websites and inboxes to engage, qualify and convert inbound leads, and it's listed as GA now.
An agent that works a deal for weeks is going to make a lot of small decisions on your behalf. Which contact to email next. When to follow up. What to say after a buyer goes quiet. Whether a deal counts as "at risk."
Salesforce says those decisions happen within guardrails that define when Hunter acts on its own and when it needs approval. That's the right design. But somebody on your team has to set those guardrails, and the defaults you pick will matter more than the model.
Here's the question I'd start with: who owns the next step?
We've written before that a deal with no agreed next date isn't really in your pipeline. An agent that keeps sending polite follow-ups can make a dead deal look alive. The activity log fills up, the CRM shows touches every few days, and the forecast still has it in the column. Nobody's actually agreed to anything.
So before you turn on a long-running agent, decide what counts as progress. I'd suggest something strict: a deal moves forward only when the buyer agrees to something with a date on it. A reply that says "thanks, will circle back" isn't progress. Another automated touch definitely isn't.
If your team gets access, these are the rules I'd want written down before the pilot starts.
Approval before anything goes to a buyer for the first time. Research, drafting and internal notes can run on their own. The first message to a new contact, or anything that mentions price or terms, waits for a person.
A cap on follow-ups without a reply. Pick a number. After that, the agent stops and flags the deal for a human decision: call, change the angle, or close it out. Endless polite chasing hurts your reputation with a buyer more than silence does.
No stage changes by the agent. Let it suggest that a deal is at risk or ready to move. Let the rep actually move it. Your forecast should reflect judgment, not activity.
A weekly review of what it did. Durable execution means plans keep running while you're busy. That's the value, and it's also the risk. Ten minutes a week reading through the agent's plan and actions on your top deals is cheap insurance.
One person who can switch it off. Sounds obvious. Decide who it is before you need them.
The honest version: an outbound agent that works for weeks will take over a lot of the work reps have always complained about. Account research, first drafts, reminders, keeping track of who said what three weeks ago. That's good. We've argued for a while that reps lose too much of the week to admin.
What it shouldn't take over is the judgment about whether a buyer is actually interested. That comes from conversations, and from noticing what a buyer isn't saying. A long-horizon agent is good at persistence. Persistence is only a virtue when you're persisting on the right deals.
So the reps who'll do best with Hunter, or any tool like it, are the ones who use it to spend more time on the calls that matter, not the ones who let it run a hundred sequences and hope something sticks.
I like the direction. Sales work really does stretch over weeks, and tools that forget everything between sessions have always been a poor fit for it. Memory and durable plans are what a sales assistant should have had from the start.
But "it keeps working on the deal" is exactly the feature that can quietly damage a forecast. If Hunter goes GA in November as planned, the teams that get the most from it will be the ones that decided ahead of time what progress means, what needs a human, and what the agent is never allowed to do on its own. That's a one-page document. Write it before the pilot, not after the first awkward email.
SalesAssist installs the complete follow-up and outreach systems behind 7-figure businesses — so your reps spend 100% of their time closing.
Hunter is an outbound sales agent in Salesforce's Agentforce lineup that works a sales pipeline from research to outreach, collaborating with sellers over weeks and months.
HubSpot's Deal Progression reads calls and emails and suggests CRM updates a rep approves with a click. Use it for notes and drafts. Keep stage, amount, close date and next step with the rep.
"Follow up" isn't a next step. If the buyer hasn't agreed to a date, the deal belongs in your hopes, not your forecast.
The email that gets a reply is shorter than the CRM template. One fact from their mouth, one question they can answer, then stop.