The numbers are starting to tell a different story about AI in sales. Outreach reported that AI credit consumption grew 12x during the first half of 2026, while AI annual recurring revenue jumped 480% year over year in its fiscal second quarter. Revenue teams are not just asking AI for information anymore - they are letting it handle pieces of the work.
Engagement with Kaia, the company's AI meeting and conversation intelligence product, grew 40%. The scale of usage suggests people are coming back and using these tools as part of their regular work, not just running a few demos and moving on.
The gap between knowing what to do and actually doing it
Sales has never had an information problem. Managers know deals are getting stuck. Reps know they should follow up with prospects. Revenue leaders know accounts go quiet because somebody missed a signal or simply got busy. The problem is what happens next. Knowing a deal needs attention is different from somebody actually doing something about it, especially when every rep has dozens of other things competing for their time.
This is where agentic AI starts to make practical sense. Outreach has been building agents that research prospects, personalize outreach, prepare sellers for meetings and spot changes in deals. Its Omni agent lets users ask questions across accounts, opportunities, prospects and conversations, then move directly into an action like drafting an email. Agent Studio lets companies build workflows around things sales teams already do: following up with inbound leads, going back to closed-lost opportunities and flagging deals that have stalled.
The value is in catching something that would have been missed and doing something about it while there is still time. SolarWinds used an Outreach win-back agent that produced a 45% reply rate, reactivated more than 100 accounts and reopened $200,000 in pipeline. Resi has logged more than 1.4 million Kaia recordings and reported a 35% win rate among its mid-market account executives. Those examples connect the technology to things companies already measure, which is more useful than another estimate about how many hours AI might save.
AI moves into the workflow, not another tab
A lot of workplace AI still feels like an extra stop. Open a chatbot. Ask a question. Copy the answer. Go back to the software where you were doing the actual work. It can be helpful, but it is still another tab. The next version looks different. Outreach has been building connections through Model Context Protocol that allow AI systems to access revenue data and take permitted actions inside sales workflows. Its MCP tools can work with prospects, sequences and records, which makes it possible for intelligence gathered in one place to turn into action somewhere else.
That is probably closer to how most salespeople will eventually experience AI. They will not spend their days thinking about which model they are using. It will just be part of the software and processes already around them, handling the repetitive stuff that tends to fall through the cracks. That does not mean handing the entire sales process to a machine. Buyers still want to talk to people who understand their business. Sellers still need to know when to stop following the script and have an actual conversation. AI is simply getting better at handling everything surrounding those moments - and AI for Sales training is becoming more focused on these practical applications.
Usage might be the metric that matters
The AI race has spent a lot of time focused on capability. Which model is smartest? Which company has the newest agent? Which product can generate the most impressive demo? Those questions matter, but actual usage may tell us more about where this market is headed. A 12x increase in AI consumption suggests people are integrating these tools into their daily routines. That is a much harder thing to manufacture than excitement around a product launch.
Revenue teams are a good place to watch this happen because there is nowhere for the results to hide for very long. Eventually, better execution has to show up somewhere in pipeline, conversion rates, win rates or revenue. If it does not, the experiment gets harder to justify. For years, companies have been asking what AI can tell their employees. We may finally be getting to the more useful part of this story - what happens when AI starts helping them get the work done. The shift from asking questions to taking action is where Workflow Automation Training becomes directly relevant to revenue outcomes.
Why this matters for sales professionals
Salespeople who treat AI as a research assistant are already behind. The tools are moving past answering questions and into executing tasks: drafting follow-ups, flagging stalled deals, reactivating cold accounts. The reps who get value from these systems will be the ones who learn to trust them with the repetitive work that eats up selling time. The metric to watch is not which model scores highest on a benchmark - it is whether your win rate moves when you let the agent handle the follow-up you would have forgotten anyway.
Your membership also unlocks:
