The era of AI hype is behind us. After collaborating with over 40 B2B companies in the past 18 months and deploying more than 20 AI agents across SaaStr's own operations, we can offer a clear, experience-backed perspective on where artificial intelligence truly delivers results today—and where it falls short.

Here's an honest breakdown of AI's current impact on various business functions:

Outbound Email-Based SDRs: 90%+ Replaceable (But Not Effortlessly)

Today, it's possible to replace the vast majority—potentially 100%—of your outbound Sales Development Representatives (SDRs) with AI. However, there's a critical nuance often overlooked: human oversight and continuous training are indispensable, every single day.

Companies successfully leveraging AI in this area aren't simply "turning it on." They are deeply committed, training platforms daily, orchestrating responses in real-time, and dedicating 2-3 hours per day to refining prompts, analyzing outputs, and adjusting targeting. This rigorous, hands-on approach allows them to replace most traditional email-based SDR tasks. Conversely, businesses adopting a "set it and forget it" mentality are often the ones lamenting that "AI doesn't work for sales." It's also important to note that AI cannot replace in-person cold calling.

Inbound BDRs: 95%+ Replaceable (Right Now)

Replacing Inbound Business Development Representatives (BDRs) with AI is, surprisingly, simpler than with outbound roles. A properly trained AI agent can qualify inbound leads faster and more effectively than 95% of human BDRs currently can.

This efficiency stems from AI's strength in pattern matching. It excels at asking the right questions, identifying crucial signals, and routing leads appropriately. AI agents maintain consistent performance, never forget to inquire about budget, and respond in seconds rather than hours. Companies still relying on human BDRs for initial lead qualification are, quite simply, leaving revenue on the table.

Customer Support: 50% Replaceable (With Strategic Investment)

You can replace approximately 50% of your customer support team with AI today, provided you make the necessary investment. Furthermore, a well-trained AI agent can deflect an even greater number of issues.

Success in this domain also demands a deep commitment. Continuous platform training, daily orchestration, and, critically, a highly responsive human-in-the-loop escalation strategy are essential. The most effective implementations we observe see companies handling 50-60% of Tier 1 support tickets entirely with AI, reserving complex issues for human agents who can truly problem-solve. This approach leads to faster response times, increased customer satisfaction, and reduced support costs, but it absolutely requires substantial investment—there's no room for a half-hearted effort.

Marketing Managers: Many Roles Are Redundant

This insight often makes people uncomfortable, but it's a reality: current AI tools like Claude and ChatGPT, without requiring specialized integrations, can manage a significant portion of marketing operations. Advanced visual AI tools are capable of creating compelling collateral, and platforms like Gamma can equip sales teams with superior materials than many marketing departments produced just 18 months ago.

What remains unchanged, however, is the need for orchestration—and a lot of it. Companies reducing their marketing teams by 50% aren't just "using AI"; they have a highly talented marketer orchestrating multiple AI tools, meticulously reviewing outputs, and upholding stringent quality standards. While you may no longer need five people managing campaigns, you absolutely need one exceptional individual managing AI that performs the work of five.

Customer Success: 60%+ Not Needed (Awaiting Better Platforms)

This is perhaps the most controversial area. Customer Success Managers (CSMs) who primarily "check in," conduct Quarterly Business Reviews (QBRs) without adding tangible value, or whose main contribution is merely "relationship management" are becoming obsolete in the age of AI. Many of these roles can be transitioned out today.

While we still need more advanced AI platforms for customer success, even with existing tools, over 60% of current CSM roles may be unnecessary. The CSMs who will thrive are those who actively solve problems, drive product adoption, and identify expansion opportunities. Others are at risk. Investing resources into First-Day Experience (FDE) specialists, especially those trained to onboard customers with AI agents, often yields better returns.

Account Executives: Still Need 70% (For Now)

The majority of Account Executives (AEs) remain essential. This dynamic is expected to shift for inbound sales within the next 24 months, but we haven't reached that point yet.

For field sales and in-person selling, the timeline is much longer—perhaps five to ten years, or even more. The human element is still critically important for complex enterprise deals. Relationships, the ability to read a room, adapt on the fly, and navigate intricate political dynamics are skills AI has not yet mastered.

AI Won’t Replace Sales Reps So Much as ‘Deflect’ Them: What Support’s 70% Deflection Rates Tell Us About Sales’ Future

Engineering: Zero Net Headcount Reduction (But 20-40% Productivity Boost)

Here's a surprising finding: we are not observing AI leading to any net reduction in engineering headcount within companies. Instead, AI is delivering a widespread 20-40% boost in productivity, and sometimes a tenfold increase for prototypes and proof-of-concept development.

However, this increased productivity fuels an "arms race." As everyone becomes more efficient and software ships faster—features that once took months now take weeks—companies need more exceptional engineers than ever to keep pace with competitors who are also accelerating. In product development today, the winners aren't those cutting engineering teams; they are the ones investing further in top-tier engineers, empowering them with advanced AI tools.

The Real Lesson: AI Requires Sustained Commitment

The overarching pattern across all these functions is clear: AI doesn't replace humans through magic. It enables human replacement when companies commit to a strategic, ongoing effort:

  1. Daily Orchestration: Someone must actively manage, train, and refine AI systems every single day.
  2. Deep Integration: Simply bolting on an AI tool won't yield results. Workflows must be redesigned around AI capabilities.
  3. Human-in-the-Loop Escalation: The most effective AI implementations seamlessly know when to hand off tasks to human experts.
  4. Continuous Training: AI systems degrade without constant attention. Successful companies treat AI like a junior employee requiring daily coaching.

Companies that "try AI" and declare it ineffective are typically those failing to implement these critical steps. Conversely, businesses quietly replacing over 50% of certain roles with AI are treating it as the serious operational initiative it is—with dedicated resources, daily attention, and genuine investment. This commitment is the fundamental difference between AI hype and tangible AI results.

Related Posts