If you skim headlines, you’d think “AI = fire support, ship a bot, profit.”
The actual data tells a very different story:
- When done well, AI-powered CX engines can boost customer satisfaction by 15-20%, increase revenue by 5-8%, and cut cost to serve by 20-30%.
- Gartner expects agentic AI will autonomously resolve 80% of common service issues by 2029, reducing operational costs by ~30% but they frame this as redesigning the service workforce, not deleting it.
- Zendesk’s CX Trends work calls this the era of “intelligent CX”: the top “Trendsetter” companies use AI to free agents for high-value interactions, and more than two-thirds of CX orgs believe generative AI will help them provide warmer, more human-feeling service when it’s implemented right.
- Forrester describes the future contact center as AI offloading every repetitive, predictable task, summarizing cases, drafting responses, routing intent, so humans can focus on the complex, emotional interactions that require empathy and personalization.
On the Customer Success side:
- Gainsight’s State of AI in Customer Success 2024 says “automation is merely the starting point”: more than half of CS orgs already use AI, but the point is to make CSMs more efficient, effective, and creative, not redundant.
- The 2024/2025 Customer Success Index and related surveys show ~91% of CS leaders believe AI will have a moderate or significant impact on CS strategy, and around half of teams integrated AI in 2024 to remove blind spots and automate routine tasks so CS can focus on strategic work.
- Forrester’s In The Age Of AI, Reinvention Is The Future Of Customer Success argues CS is being reinvented around AI. AI agents surface churn risk and expansion opportunities, but humans still own relationships, outcomes, and the value story.
In other words:
The highest-ROI AI CX stories are not “we fired support.”
They’re “we re-architected CX around human + AI, and now our teams only do higher-value work.”
This post is about that.
1. The AI Upside for Service Is Real (If CX Leads It)
Let’s start with the upside, because it’s big.
McKinsey: “Next Best Experience” isn’t hype, it’s math
McKinsey’s recent work on AI-powered next best experience engines shows that, when implemented well, they can:
- Enhance customer satisfaction by 15-20%
- Increase revenue by 5-8%
- Reduce cost to serve by 20-30%
That’s not just “better routing.” It’s:
- Using AI to predict what a customer needs next
- Surfacing the right offer, answer, or action at the right time
- Coordinating across channels (site, app, service, email, CSM) so the experience feels joined up
But those numbers only show up when:
- The journeys are designed by CX, not just wired by IT
- Data, rules, and exceptions are shaped by people who know customers, not just what’s easy to implement
Gartner: Agentic AI will handle the basics, but not design your workforce
Gartner’s March 2025 prediction is blunt:
- By 2029, agentic AI will autonomously resolve 80% of common customer service issues
- That will drive about a 30% reduction in operational costs
Crucially, their note isn’t “so fire everyone.”
They frame it as:
- A massive redesign of the service workforce
- AI agents taking on more routine and mid-complexity interactions
- Humans shifting into:
- Exception handling
- Supervising and tuning AI
- High-stakes, high-emotion work
That redesign is exactly the kind of thing CX and support leadership should own.
Zendesk: “Intelligent CX” = AI + humans, not AI vs humans
Zendesk’s CX Trends 2025 and related stats point to a clear pattern:
- We’re entering the era of “intelligent CX” = AI + automation + data analytics reshaping how customers interact with businesses.
- CX Trendsetters(the leading cohort) use AI to:
- Free agents for high-value conversations
- Deliver more personalized, real-time experiences
- More than two-thirds of CX orgs believe generative AI will help them provide warmer, more familiar service, even at large scale.
They’re not talking about AI as a wall between customers and agents. They’re talking about AI as infrastructure for a better, more human-feeling experience.
Forrester: AI offloads the boring work so humans can do the human work
Forrester’s 2024 view on AI in customer service reads like a job description for AI as a teammate:
AI will:
- Automate simple interactions
- Capture customer intent
- Route inquiries to the right agent
- Guide agents with next best actions
- Summarize cases, take notes, draft replies, create knowledge content
In their words, AI will “offload every repetitive, predictable task from agents so they can focus on the more complex interactions that require empathy and personalization.”
So the vision from the big analysts isn’t “no agents.”
It’s AI-first service with human-led experiences.
2. On the CS Side: AI Is Reshaping, Not Replacing
Customer Success is going through its own AI wave, and the pattern is almost identical.
Gainsight: “Automation is merely the starting point”
Gainsight’s State of AI in Customer Success 2024 is explicit:
- 50%+ of CS orgs are already using AI.
- They frame their philosophy as “Human-First AI”:
- AI alongside humans, not instead of them
- AI as a way to remove grunt work so CSMs can focus on strategy and relationships
- Key use cases:
- Auto-summarizing calls and QBRs
- Surfacing health signals and churn risk
- Identifying expansion opportunities
- Drafting playbooks and outreach that CSMs can tune
The punchline in the report:
Automation is merely the starting point.
The real prize is unlocking a more creative, proactive, value-focused CS team.
CS Index & Forrester: AI + CS = outcomes, not tickets
Across the 2024–2025 CS Index studies and commentary:
- Around 91% of CS leaders believe AI will have a moderate or significant impact on CS strategy.
- Roughly half of CS teamsreport integrating AI in 2024, mainly to:
- Eliminate blind spots in accounts
- Automate low-value tasks
- Free CSMs for strategic work
Forrester’s In The Age Of AI, Reinvention Is The Future Of Customer Success adds:
- AI agents are the new CS teammates:
- They summarize meetings
- Monitor adoption and usage
- Flag risk and opportunity signals in real time
- Human CSMs:
- Focus on guiding customers to outcomes
- Build multi-threaded relationships
- Craft the value story for CFOs and execs
CS doesn’t fade in this future; it steps into its prime with AI doing the heavy lifting underneath.
3. What the Best Teams Actually Do Differently
If you put all this together, the highest-performing CX and CS orgs are not just “adding AI.”
They’re re-architecting how work gets done.
Here’s the pattern you can crib from.
1) Treat AI as infrastructure, not as a channel
Leading teams:
- Use AI to:
- Classify and route contacts
- Suggest next steps
- Summarize history and context
- Keep knowledge bases fresh
- Then layer that across:
- Chat, email, voice, in-product, CSM workflows
The key: CX owns the blueprint. IT and vendors supply the plumbing.
2) Start with repetitive work, not relationship work
They systematically ask:
- What are the repetitive, predictable tasksbogging agents and CSMs down?
- Password resets
- Basic how-to documents
- Status checks
- Routine follow-ups
- Where can AI:
- Fully automate (no human needed)?
- Co-pilot (draft + assist a human)?
- Just prepare context (summaries, intel) so the human is sharper?
Only after that do they touch:
- Escalations
- Sensitive or emotional issues
- Strategic account conversations
And even there, AI mostly preps the human, not replaces them.
3) Redesign roles and metrics around higher-value work
This is where CX and CS leadership really matter.
They:
- Rewrite role expectations:
- Agents as problem-solvers and advocates, not script readers
- CSMs as outcome owners, not ticket shepherds
- Adjust metrics:
- Less obsession with handle time for complex contacts
- More focus on:
- First-contact resolution for simple issues
- NPS/CSAT on complex ones
- Adoption, expansion, and NRR on the CS side
AI delivers the productivity dividend.
Leadership decides to reinvest it in better experiences and outcomes, not just fewer people.
4. The Wrong Lesson: “We Fired Support”
- “We automated 60-80% of tickets.”
- “We cut our contact center headcount by X%.”
The problem: Forrester, Gartner, Zendesk, and others are very clear that many AI CX deployments underperform, often because they’re:
- Launched as cost-cutting exercises
- Done without redesigning journeys or roles
- Not backed by strong data, knowledge, and change management
When that happens, you get:
- Bots that trap customers
- Frustrated agents and CSMs with worse tools and more firefighting
- Short-term OPEX gains, long-term churn, and reputation damage
In a world where retention and NRR are your actual profit engine, that’s a bad trade.
Where The Scale Crew Fits In
This “CX-led AI” picture is exactly where The Scale Crew is building.
With our expanded offering, we bring:
- Fractional CX/Support/Success leadership
- People who’ve actually run:
- Support orgs
- Customer Success teams
- CX programs tied to NRR, churn, and LTV
- People who’ve actually run:
- AI Readiness & Transformation
- A gated approach that:
- Starts with “Should you?” (what to automate, what to augment, what to leave alone)
- Aligns leadership on where AI belongs in the journey, and where humans must stay central
- Designs AI + human workflows for support and success, not just standalone bots
- A gated approach that:
We don’t show up and say “replace your team with AI.”
We show up and ask:
- Where is your cost to serve too high and your experience too mediocre?
- Where could AI take 30-50% of the drudge work out of CX and CS?
- How do we redesign roles so your humans:
- Only do the things that actually build loyalty, expansion, and advocacy
- Use AI as an under-the-hood advantage your competitors can’t easily copy?
If You Want AI Wins That Are CX-Led, Not Vendor-Led
We’ll help you see:
- Where AI could realistically deliver the 15-20% CSAT, 5-8% revenue, 20-30% cost-to-serve improvements McKinsey talks about
- How to make those gains CX-led, so your support and success teams do more of the work that actually moves NRR
- And how to avoid the trap of “we installed a bot” instead of “we redesigned how we serve our customers in an AI era.”
