Chatbots Have Been Failing Us for 20 Years—Can AI Finally Fix It?

We’ve all been there — desperately searching for a human behind a chatbox to solve our issues with the products or services we’ve acquired. But did you know that the first-ever chatbot for customer interactions was created all the way back in the early 2000s?

It was called SmarterChild. It understood little. It obeyed strict rules. It frequently got stuck. There was nothing natural about it. Users had to figure out the precise way to ask their question, or they would be trapped in an endless loop of repetitive, robotic responses. It was merely a template machine — no AI, no intelligence.

Today, things have changed dramatically.

AI: The Game Changer in Customer Success

AI is everywhere. It’s like that popular kid in school that everyone talks about. And as our lives get an artificially intelligent upgrade, it’s time to rethink business optimisation.

Our 2025 Customer Success goals at SDG have only grown more ambitious. We strive to provide highly personalised, human support that fosters strong relationships with our customers. At the same time, we are actively refining automation and optimisation to set the industry standard and make Customer Success yet another one of our competitive advantages.

“Hmm, contradictory. Why?” You may ask.

Because our loyal customers deserve our human attention, focus, and time. AI is a strategic investment that enhances our work, allowing us to provide precisely that!

Balancing AI & Human Support: A Learning Curve

AI implementation requires balance. And how do you achieve that?

✅ Frame the digital transformation process as a product.
✅ Be the product manager of automation and optimisation.
✅Experiment. Measure. Repeat.

The Numbers Game: Data is King

Data-driven decisions are essential for any successful business. Before deploying AI across your processes, you need a clear understanding of the current state of things.

💡 Benchmark first. Analysing key metrics will prevent unnecessary customer dissatisfaction and help you accurately measure AI’s effectiveness.

Start by asking:

▶ How many tickets do you receive in a given period?

▶ How many should remain untouched by AI and receive only human support? Why?

▶ How many customers require basic information that AI could handle, saving human time?

▶ How many tickets can realistically be automated? (Emails? Calls? Chats?)

▶ How many human work hours could be reallocated to high-value tasks?

Once you complete this analysis, you will have the numbers and KPIs to track AI’s impact — successes and failures alike.

Automation vs. Personalisation: Finding the Right Balance

Once you gather data, you’ll begin to identify trends in your customers’ inquiries. Most interactions can be automated, but the real question is:

💡 “Where do I want to preserve the value of personalisation?”

If Customer Lifetime Value (CLV) is a key metric in your business, AI cannot fully replace your Customer Success Managers (CSMs). They are the human touchpoints of your company, the very people who build long-term customer relationships. A company without visible human presence risks losing trust.

Where Should Personalisation Be Prioritised?

Depending on your industry, these common areas might require human-first support:

✔ Premium and loyal customers
✔ Serious complaints
✔ Refund requests
✔ Payment issues
✔ Paying customers
✔ Technologically challenged customers

The goal isn’t to remove human support but to optimise agent working hours. Why waste professional talent on repetitive, copy-paste responses when AI can handle those? By focusing human effort on unique or sensitive cases, Customer Success can drive higher satisfaction and increase CLV.

AI as a Co-Pilot: Expanding Customer Success Managers’ Capabilities

AI isn’t just about handling routine requests — it can also empower CSMs by acting as a co-pilot.

✅ Sentiment Analysis→ AI scans customer communications, detects dissatisfaction, and flags cases needing human intervention.
✅Real-Time Assistance for Agents→ AI suggests responses, relevant knowledge base articles, and past insights.
✅Call & Chat Summarisation→ AI transcribes and summarises conversations, freeing agents to focus on problem-solving.
✅Proactive Retention Strategies → AI identifies customers at risk of churn and prompts CSMs to act.

Rather than replacing agents, AI makes them more effective.

Experimentation: The Key to AI Success

Before launching AI initiatives, you need a roadmap that outlines:

• Which AI implementations to prioritise

• Expected impact

• How success will be measured

  
  

📌 Documentation is key! It avoids confusion, ensures alignment among stakeholders, and establishes clear benchmarks and business goals. No documentation = no organisation.

Test Before You Scale

To evaluate AI’s impact, divide customers into two groups:

1️⃣ Control Group→ Receives traditional human-only support.
2️⃣Experimental Group → Interacts with AI-enhanced processes.

For accurate results, ensure:
✔ Experiments run long enough to collect meaningful data.
✔ Key metrics remain stable (or improve).

👉 If automation performs as well as human support while saving time, it’s a win!

One by one, roll out AI experiments and adjust based on results. If an AI implementation underperforms, tweak it and try again. Stay Agile. Iterate. Improve.

Conclusion: AI as an Enabler, Not a Replacement

AI isn’t here to remove humans from Customer Success — it’s here to enhance their impact.

By automating routine tasks, surfacing key customer insights, and freeing agents to focus on meaningful interactions, AI helps businesses deliver a higher level of personalised service — without sacrificing efficiency.

✅ Use AI strategically
✅ Keep human support where it matters most \ ✅Create a customer experience that is both seamless and deeply personal**

Ultimately, the businesses that will thrive are not the ones that eliminate human touch — but the ones that make human interactions more impactful than ever before.

Written by Tony Jacobs, Director of Customer Success Department at Social Discovery Group