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

The AI Account Executive Is Arriving. Here’s What Support — Which is 18+ Months Ahead — Teaches Us About What Comes Next.

Customer support cracked the deflection code with AI. Companies are routinely achieving 60%+ deflection rates – meaning AI handles 6 out of 10 (or more) support tickets before a human agent ever gets involved.

Now, the first AI Account Executives and BDR platforms are starting to do the same.

The question isn’t whether AI will impact sales. It’s how quickly sales will follow support’s playbook.

Support’s Deflection Breakthrough

The numbers from leading AI support platforms show how radically AI has already changed support and post-sales:

• Decagon, which just raised $131M at a $1.5B valuation, reports businesses achieving average deflection rates nearing 70%, with companies like Duolingo pushing well above 80%. Bilt handles 70% of their 60,000 monthly support tickets with AI agents, saving hundreds of thousands monthly.
• Intercom’s Fin achieves 86% resolution rates, handling everything from account changes to complex troubleshooting. These aren’t simple FAQ bots – they’re AI agents that can access customer data, apply business logic, process transactions, and escalate intelligently.
• Gorgias’ AI for eCommerce Support already hits 60% routinely even with smaller SMBs with less time to train their AIs.
• Salesforce reports that companies using Agentforce achieve a “double digit percentage increase in customer satisfaction and deflection rates” with a 50% increase in case resolution. Their Service Cloud helps businesses “deflect 30% of cases” out of the box.
• Zendesk partners like Aisera report “70 percent ticket resolution” rates, while
• Klarna’s AI assistant was partially rolled back, yes — but it now handles two-thirds of their customer service –equivalent to 700 full-time human agents.
• Talkdesk has introduced Autopilot, their generative AI that “autonomously resolve complex use cases” rather than simply deflecting calls to live agents like traditional systems.

This same deflection model is now making its moves into sales.

The AI Account Executive Emerges.  The Goal Again?  Deflection

While support deflection is mature, sales deflection is just beginning. We’re seeing the earliest AI Account Executives handle:

Existing Customer Expansions: “I need 5 more licenses” – simple add-ons that don’t require relationship selling

Routine Renewals: Annual subscriptions with standard terms and minimal complexity

Inbound Product Inquiries: Qualified prospects seeking specific information or standard pricing

Demo Scheduling: Intelligent routing based on prospect needs and rep specialization

The technology exists today. Companies like Decagon and Intercom have proven AI can handle complex workflows autonomously. What’s happening now is the migration of this capability from post-sale support to pre-sale interactions.

Why Sales Will Follow Support’s Path (It’s Already Starting)

The parallels are striking. Five years ago, support teams worried AI would eliminate jobs. Today, those same teams report that AI has elevated their roles – handling routine inquiries while humans focus on complex problem-solving and relationship building.

Jesse Zhang, CEO of Decagon, puts it perfectly: “AI is often seen as destroying jobs, but at Decagon, we believe the opposite. Our AI agents are enhancing jobs, not replacing them.” Support reps have become AI managers, configuring and overseeing agents rather than answering every Tier 1 ticket.

Sales is heading down the same path, just 18-24 months behind.

And it’s a very helpful way to reframe what you want out of an “AI Account Exec”.  It’s not either or.  It’s about deflection.  How many routine sales can be handled by a great AI, instantly and in real-time, before the prospect wants to talk to sales?

The Early Deflection Opportunities

Smart sales leaders are identifying their “support ticket equivalents” – the routine sales interactions that could be handled by AI:

• Low ACV Deals at List Price:  If customer has already tried on their own, they may not even want to talk to human.
• Simple Upsells: Existing customers adding standard features or capacity
• Renewal Processing: Straightforward annual renewals with minimal negotiation
• Pricing Requests: Standard package quotes for qualified inbound leads
• Basic Product Questions: Feature comparisons and compatibility inquiries

Intercom’s Fin demonstrates the capability, with the ability to “handle complex workflows autonomously” taking resolution rates “from 50% average to potentially 80 or 90%.” Applied to simple sales scenarios, this suggests significant deflection potential.

What the Support Model Teaches Us

Support’s journey offers a roadmap for sales deflection:

• Phase 1 (Where sales is now): AI handles basic inquiries and schedules meetings
• Phase 2 (Coming soon): AI processes simple transactions and standard renewals
• Phase 3 (The future): AI manages entire deal cycles for routine purchases and “deflects” the need for sales to get involved until issues come ups.

The Economics Drive Adoption

The ROI is compelling. ClassPass achieved a 95% cost reduction in support conversations. For sales, every routine interaction handled by AI rather than a human rep represents immediate cost savings and rep time freed for complex deals.

HubSpot’s research shows sales teams using AI are 1.3x more likely to see revenue increases, with 40% of sales professionals already using AI tools. But we’re still in the early adoption phase compared to support’s mature implementation.

Looking Ahead: The Deflection Timeline

Based on support’s evolution, here’s the likely sales trajectory:

• 2024-2025: AI handles inbound qualification and simple expansions (5-15% deflection)
• 2026-2027: AI processes routine renewals and standard upsells (20-35% deflection)
• 2028-2030: AI manages full transaction cycles for commodity sales (40-60% deflection)

The key insight from support: deflection rates compound quickly once the technology proves reliable. Duolingo didn’t gradually increase from 10% to 80% – they saw explosive growth once the AI agents demonstrated consistent performance.

The Strategic Imperative

Sales leaders should start preparing now:

1. Identify deflection candidates: Which deals require minimal human judgment?
2. Measure baseline metrics: Track current time spent on routine vs. complex sales
3. Pilot AI interactions: Test AI handling of simple customer requests
4. Train for the transition: Prepare reps to work alongside AI agents

The Bottom Line

Support achieved 60-80% deflection rates by letting AI handle routine inquiries while humans focused on complex problem-solving. The first AI Account Executives are now applying this same model to simple sales transactions.

We’re not there yet – sales deflection is still in its infancy. But support’s playbook shows us exactly where this is heading. The companies that start measuring and optimizing for deflection rates now will have a significant advantage as AI Account Executives become mainstream.

The question isn’t whether AI will deflect a significant portion of your sales volume. Support already proved that’s possible. The question is how quickly you’ll adapt to this new reality.