Seasoned Customer Success Executive Breaks Down Where AI Changes the Game — and Where It Falls Short

Today, I work with B2B SaaS organizations globally, typically mid-market to enterprise, helping their Customer Success, Account Management, and broader post-sale teams drive revenue from their existing customer base. Not just retention, but real expansion and measurable growth. I work with teams ranging from 10 to 100-plus CSMs across EMEA, the US, and APAC.

Those teams usually manage a mix of customer segments, from high-touch enterprise accounts to scaled mid-market books of business. The common thread across all of them is that they are asked to do more than just manage relationships. They are asked to own revenue, influence renewals, and identify expansion opportunities.

My role helps those organizations make that shift, from reactive, service-led Customer Success to proactive, commercially-driven teams that can confidently tie what they do back to revenue.

Why CX leaders must be intentional about when to use AI

Many teams are getting the split between AI tasks and human tasks wrong. They are either over-relying on AI or avoiding it completely.

The line is very clear. AI should inform decisions and accelerate execution, but it should not own the relationship or the judgment.

Here is how I approach it. I rely on AI for:

• Issue and risk prioritization: AI excels at spotting patterns across large data volumes: product usage drops, support spikes, changes in engagement, and even tone in customer communication. It surfaces at-risk accounts much faster than a human scanning dashboards.
• Personalization at scale: Not for relationship building, but for preparation. AI helps draft tailored emails, call agendas, follow-ups, and even stakeholder-specific messaging based on role. Instead of a CSM staring at a blank page, they start with something 80% complete.
• Insight generation for conversations: Before a renewal or value review, AI synthesizes everything into a point of view. It identifies what changed, where value appears, where friction exists, and where expansion potential lies. This level of synthesis used to take hours.
• Internal efficiency: Summarizing calls, updating CRM notes, identifying next steps — it automates all of that. CSMs should not spend their time as admin operators.

AI should make your team faster, sharper, and more prepared. But moments that define customer relationships, trust, and revenue outcomes still need a human in the driver’s seat…The teams that win are not choosing between AI and humans. They are intentional about where each plays.

Share This Quote on:

• Share on Twitter
• Share on LinkedIn
• Share on Facebook

Anika ZubairOpens new window

CEO of The Customer Success Pro

Here's where I am intentional about keeping it explicitly human:

• Commercial judgment: Deciding how to handle a renewal, whether to push, where to concede, and how to structure a deal. This requires context, experience, and reading the room. AI can inform it, but it should not make that call.
• Relationship building and trust: Customers sense automation. Executive conversations, difficult discussions, and moments where you challenge a customer or reframe their thinking must come from a human.
• Service recovery and escalations: When something goes wrong, hiding behind automation is the worst approach. Empathy, ownership, and strong communication matter most then.
• Journey design at a strategic level: AI can suggest patterns, but deciding how your customer experience should feel, where to lean in, where to automate, and how to differentiate—that is a leadership decision.

AI should make your team faster, sharper, and more prepared. But moments that define customer relationships, trust, and revenue outcomes still need a human in the driver’s seat.

The teams that win are not choosing between AI and humans. They are intentional about where each plays.

Why AI is not a shortcut to transformation

And remember: AI will amplify whatever your team already is.

If your team is strategic, commercially sharp, and customer-focused, AI will make them faster, more consistent, and more impactful. But if your team is reactive, task-driven, and operating like support, AI will just make them a faster version of that.

It is not a shortcut to transformation.

Why AI in CX has huge upsides — and downsides

AI's results have been largely — but not completely — good:

• Time savings are significant: Teams see a 30 to 60% reduction in prep time for QBRs, renewal briefs, and account reviews. What used to take two to three hours now finishes in 20 to 40 minutes.
• Conversation quality has increased: This matters most to me. CSMs walk into calls with a point of view, not just data. You hear it immediately. More executive-level language, more outcome-driven conversations, less feature talk.
• Earlier risk identification: AI helps surface churn signals weeks, sometimes months, earlier than before. This gives teams a real chance to intervene instead of reacting when it’s already too late.
• Revenue impact: Teams improve renewal confidence and reduce discounting because they can articulate value. On the expansion side, they spot opportunities earlier, increasing pipeline creation from existing customers. I've seen a range of 10 to 25% improvement in expansion pipeline quality, not just volume.
• Less admin drag: Call notes, CRM updates, and summaries are handled faster, giving time back to the CSM.

These efficiency gains require a mindset shift — CX leaders need to move away from the belief that great CX requires time.

For years, we assumed high-quality customer experience took hours: deep prep, long QBR builds, manual analysis, and crafting every message from scratch. And to be fair, that used to be true. But AI completely challenged that.

How AI enhances how CSMs handle QBRs

Here's a big change I've made, thanks to AI.

A year ago, most teams manually pulled data, built slides, and tried to stitch together a “value story” from usage metrics, support tickets, and gut instinct. This process was slow, inconsistent, and very activity-focused.

I completely flipped that process with AI. I now use AI to generate a first draft of the customer narrative before opening a slide deck. Instead of asking, “What slides do I need?” they start with, “What is the business-impact story for this account?”

I feed it product usage trends, key milestones, support interactions, known customer goals, and call notes. AI then helps structure this into:

• The outcomes the customer tried to achieve
• What changed in their business
• Where value was created or is at risk
• Where clear expansion signals exist

The tools I use for this are custom GPTs, Claude, Fathom, Gong, and Planhat,

By the time the CSM builds the QBR or runs the call, they are not guessing. They walk in with a point of view. The results are significant:

1. The conversation quality immediately improved. CSMs sound more strategic because they have a narrative, not just data.
2. Prep time significantly dropped. What used to take hours now takes minutes. This means teams spend more time thinking and less time assembling.
3. Most importantly, it directly impacted revenue. Leading with a clear value story makes it much easier to handle renewals, push back on discounts, and naturally introduce expansion opportunities.

AI didn’t just make the process faster. It made Customer Success more commercial.

Why AI in CX has huge downsides

Anika Shares

The biggest issue is shallow thinking: When teams rely too heavily on AI outputs, they stop questioning the narrative.

Now, the bad — because a downside definitely exists if done poorly:

• The biggest issue is shallow thinking: When teams rely too heavily on AI outputs, they stop questioning the narrative. I’ve seen CSMs take an AI-generated summary at face value without applying their own judgment, which can lead to completely missing what is happening in the account.
• Generic communication: If not careful, everything starts to sound the same. Customers feel that. It quickly erodes trust if every email or update feels templated, even if technically “personalized.”
• False sense of strategy: This is a significant issue. Just because AI can generate a nice-looking “value story” does not mean the team is being strategic. I’ve seen leaders think they’ve leveled up their team, when in reality, they’ve just improved formatting — not thinking.
• Over-automation in the wrong moments: Using AI in escalations, tough renewal conversations, or sensitive situations can backfire badly. These are moments where tone, empathy, and judgment matter more than speed.

I've found that the winning teams use AI as a starting point, not the final answer.

Why AI falls short in delivering true CX personalization

The biggest area where AI has not delivered the CX impact expected is true personalization at the relationship level.

AI excels at making things appear personalized. It can tailor emails, reference usage data, and adjust tone by persona. On paper, it looks great. But customers — especially at the enterprise level — immediately feel the difference.

A big gap exists between, “This was generated for me” and “This person understands my business.” And that gap matters because trust builds on depth, not formatting.

Why enablement matters more than tooling

Enablement matters more than the tools themselves. That's something I wish I'd realized earlier.

When I first piloted AI with CS teams, I thought the hard part was picking the right tool and showing people a few prompts. It wasn’t. The real challenge was getting the team to think differently about their work.

To do this, I stopped talking about AI as a tool and started talking about it as a lever for revenue. I made sure my team had the right tools and the right mindset. And instead of asking them to use AI to write faster emails, I asked them to use AI to prepare for commercial conversations. What does this customer need to hear to renew? What are the expansion signals hiding in their usage data? What is the risk we are not seeing?

So, here's my advice: Do not treat AI as a tech initiative. Treat it as a capability shift.

The winning teams use AI as a starting point, not the final answer….Do not treat AI as a tech initiative. Treat it as a capability shift.

Share This Quote on:

• Share on Twitter
• Share on LinkedIn
• Share on Facebook

Anika ZubairOpens new window

CEO of The Customer Success Pro

Follow along

You can follow Anika Zubair's work on LinkedIn. And check out The Customer Success Pro.

More expert interviews to come on The CX Lead!