H

How Successful Sales Teams Are Embracing Agentic AI

by Jaya Prakash (Jay) Kaza, Doug J. Chung, Candace Lun Plotkin, Siamak Sarvari, Jennifer Stanley and Maria Valdivieso

September 15, 2025

We Are/Getty Images

Summary. Agentic AI is revolutionizing sales by enabling autonomous personal agents to work alongside human sales reps, identifying, nurturing, and closing deals across channels. This technology not only automates routine tasks but also anticipates next steps, adapts to market changes, and continuously learns, thereby transforming every customer engagement into a competitive advantage. As organizations integrate agentic AI into their workflows, they can unlock significant value by redesigning processes, enhancing customer interactions, and driving growth.close

Imagine creating a perfect replica of your top-performing sellers—but instead of someone whose capacity for work is limited by time and geography, this replica can work alongside human sales reps continuously. These autonomous personal agents can identify, nurture, and even close deals by engaging customers across channels. Their power lies not just in executing tasks, but in thinking ahead: anticipating next steps, adapting to changing market conditions, integrating across systems, and continuously learning.

This is the true promise of agentic AI—a full-scale transformation of sales, turning every customer engagement into a competitive advantage.

While generative AI has moved from experimentation to real impact—19% of B2B sales teams are already seeing success with it, and 23% are in pilot or development phases, according to a McKinsey study—agentic AI is set to unlock the next level of value. The latest McKinsey global survey on AI shows that the most significant gains come from redesigning workflows, not just digitizing them. Agentic AI is already redefining workflows across the B2B sales process—from lead generation to deal closure, unlocking seller capacity and accelerating growth.

In this article, we'll examine how successful sales teams are beginning to use agentic AI. The examples below are drawn from both our research or our consulting experience.

From "Tell Me" To "Do It for Me"

Many sales teams already use AI to increase productivity. For instance, predictive AI can tell salespeople what the next best action will be for a given prospect—the type of outreach (a call or an email) and what kind of information (about a product feature or a promotion) has the highest probability of triggering a sale. Agentic AI goes a step further: Instead of advising the salesperson on what they should do next, the agent can perform the task autonomously. Examples of this shift can be found in lead generation and outreach, sales planning and customer engagement, and retention and growth.

Lead Generation and Outreach

AI agents are redefining frontline B2B sales by actively supporting reps—interpreting buyer signals, retrieving context-specific insights, and recommending next-best actions in real time. At one B2B tech company, an AI-powered business development representative was deployed to autonomously manage top-of-funnel engagement. The process began with lead scoring, analyzing behavioral signals to identify which warm leads were most likely to convert. For each, the agent generated personalized outreach emails, drawing on data such as purchase history, usage trends, and inferred interests to tailor messaging at scale. This data-driven personalization led to a 6% lift in response rates. When leads responded, the agent parsed the language, identified intent, and continued the conversation, answering questions, addressing objections, and moving the prospect toward a meeting. Once qualified, leads were handed off to a human seller, with the full conversation history logged in the CRM for context-rich follow-up. This end-to-end orchestration that combined intelligent prioritization, data-driven personalization, and real-time dialogue management is projected to generate $50 million in incremental annual revenue through increased conversion and freed-up seller capacity.

Sales Planning and Customer Engagement

In the selling phase, AI agents can help streamline implementation, coordinate stakeholders, and ensure compliance. Traditionally time-consuming tasks like quote and proposal generation can now be automated and tailored to buyer needs. McKinsey's latest B2B Pulse Survey highlights the value of these capabilities, especially in industries with complex product portfolios or large lead volumes such as construction, shipping, or chemicals.

AI agents also accelerate pipeline progression by scheduling meetings, syncing calendars, and logging interactions into CRM systems. At one North American wealth management firm, an agentic AI tool was deployed to generate customer summaries by synthesizing CRM data with external sources, reducing meeting prep time by over 30% and driving a 6% revenue uplift. Previously, sellers relied on fragmented notes and manual research, often missing key insights or relying on pattern recognition biased by past client behavior. In contrast, the AI agent drew from both structured CRM data and unstructured inputs, such as meeting notes, product usage, and market signals to surface nuanced client needs. By identifying patterns humans might miss and tailoring insights to each advisor's workflow, the agent helped unlock new opportunities and deepen client engagement.

Retention and Growth

Agentic AI is extending the value of digital self-serve by handling far more than basic FAQs, especially in post-sales operations, a domain often underutilized as a growth lever. Traditionally, when customers hit a roadblock beyond standard queries, the issue was escalated to a service rep. But one leading technology company discovered that agentic reasoning enabled their AI-powered customer experience agent (AI chatbot) to resolve 85% of queries, including second-order questions that required interpreting context, querying multiple systems, and applying logic to deliver a tailored answer. These are the kinds of questions that previously overwhelmed self-serve channels and defaulted to human support. Now, customer service reps are freed up to focus on truly complex escalations, such as cases involving multiple, often unrelated issues that cannot be mapped to a single or even adjacent FAQ and require nuanced human judgment. The result is a 65% reduction in handling times and improved customer satisfaction.

New use cases are quickly emerging, such as onboarding agents that guide clients through setup and churn-prevention agents that proactively address customer risk. Unlike traditional AI models that simply predict churn, agentic AI can reason through the next-best steps to mitigate that risk. For example, when churn risk is flagged, the agent determines the top three to four actions—prioritized based on context, customer history, and business impact—and initiates or recommends those interventions. Once those actions are taken, traditional models can then be updated with the latest customer data to recalculate risk. This creates a continuous feedback loop where predictive AI and agentic reasoning work in tandem, moving from insight to intervention with far greater speed and precision.

Sales Teams Must Adopt New Metrics and Roles

As these examples demonstrate, Agentic AI is poised to transform sales as we know it, introducing changes to organizations' operating models and roles.

Organizations need to rethink how to measure performance and assess compensation structures, adopting metrics that reflect both human and AI contributions. Customer engagement metrics like sentiment scores, channel-specific interaction volumes, and cost-per-acquisition figures will help companies understand the holistic impact of their sales efforts. At the same time, new agent-specific metrics will emerge, including conversation quality scores, drop-off rates, accuracy in responses, and adherence to risk and compliance standards.

The operating model itself will demand new visibility. Leaders will need to track how work is distributed between AI and humans, such as what percentage of engagements are handled by agents, how often those interactions are successfully completed, and whether handovers between AI and human sellers happen seamlessly. As AI takes on more routine and transactional tasks, human sellers will increasingly be evaluated on the aspects of the role that require distinctly human strengths: building relationships, navigating complexity, and influencing key decision-makers.

Beneath the surface, technical performance indicators like response time, multi-turn conversation completion, hallucination detection, memory efficiency, and token usage will provide a window into how well AI agents are performing from a systems perspective.

Agentic AI will also transform sales roles. Some will become automated, others will be redefined, and entirely new ones will be created. The focus will shift from execution to orchestration and influence.

Account managers, for instance, will pivot from tactical execution to relationship cultivation and decision-maker influence—coordinating across humans and agents. Business development managers may see their traditional tasks—prospecting, qualifying, and follow-up—automated, allowing them to focus on agent oversight and process optimization.

These changes free salespeople from repetitive tasks and elevate the importance of human judgment, emotional intelligence, and trust-building—fostering a culture of continuous learning, coaching, and performance insights.

. . .

Agentic AI may still be in its early stages, but its potential represents the most significant productivity leap in sales since the dawn of CRM. Organizations that strategically position their sales functions to incorporate agentic capabilities will outpace those still mired in manual workflows. As with any powerful technology, the real opportunity lies not in replacing people but in reshaping how humans and machines collaborate—to elevate roles, increase performance, and redefine purpose.

Jaya Prakash (Jay) Kaza is a Senior Expert at McKinsey & Company's Washington DC office. He serves clients on AI-enabled growth strategies, commercial technology, AI adoption, Data, and IT strategy.

Doug J. Chung is the CBA Foundation Centennial Fellow, Professor of Marketing, and the Director of the Sales and Business Development Forum in the McCombs School of Business at the University of Texas at Austin. He teaches sales management and strategy in the MBA and executive education programs, and has worked with organizations worldwide to develop effective sales management strategies.

Candace Lun Plotkin is a partner in McKinsey & Company's Boston office and a leader in the Growth, Marketing & Sales (GM&S) practice.She is the Global Co-Leader of the firm's B2B ecommerce and omnichannel service line and co-leads the Annual B2B Pulse research on how B2B customers buy, across traditional and digital channels, and the impact of gen AI on their behaviors.

Siamak Sarvari is an Associate Partner in McKinsey's New Jersey office, in the GM&S practice, and a core leader of its practice on AI-enabled sales, developing assets and partnering with clients to improve sales productivity through both growth and efficiency measures.

Jennifer Stanley is a Partner in McKinsey & Company's London office, where she leads their B2B Growth, Marketing, and Sales practice int he UK. She has worked with B2B sales teams globally for 20+ years, including in tech, financial services, and industrials.

Maria Valdivieso is a Partner at McKinsey & Company's Growth, Marketing & Sales Practice based in Miami, where she specializes in helping B2B and consumer companies build advanced sales capabilities for above-market growth. Over the past two decades, she has led projects in the GM&S practice and research on sales strategy, commercial transformations, and gen AI/agentic in sales. She is also a co-author of Sales Growth: Five Proven Strategies of the World's Sales Leaders (Wiley, 2016).

Read more on AI and machine learning or related topics Sales and Sales team management

Recommended For You

Podcast

Why It's So Hard to Delegate — and How to Improve

Managing Up

How to Work for a Boss Who Always Changes Their Mind

Generative AI

Does Your Board Really Understand AI?

Compensation And Benefits

Research: Pay-for-Performance Doesn't Have to Stress Workers Out

Partner Center

Start my subscription!

Explore HBR

HBR Store

About HBR

Manage My Account

Follow HBR

Facebook

X Corp.

Instagram

Your Newsreader

About Us

Careers

Cookie Policy

Trademark Policy

Harvard Business Publishing:

Higher Education

Corporate Learning

Harvard Business Review

Harvard Business School

HBR Executive

Innovation starts with how you lead. Discover the habits and mindsets that fuel breakthrough thinking.

Start your subscription now.

David

617-331-7852
David@DavidCutler.net
Growth Actions: www.DavidCutler.net
Web3 Applied: www.TruthRefinery.com
Circular Partners: www.CircularLabs.io

Search This Blog

H

Popular Posts

Same But Different

The Circularity