Philip Topham’s Post

Building customer confidence in your AI agent won't happen overnight But if you don't work at it, customers won't even attempt to use it before asking to "talk to a person", here are a few things that might help: → Display human v AI wait times When they request to talk to a person display current wait times before they make the handover e.g. "the current wait time is 15 mins for an agent, do you want to continue?" It then at least shows the trade-off making them (marginally) more likely to pursue self-serve → Give AI responses while they wait When they request a human handover, consider still allowing them to get AI responses while they wait and having the AI continue to try and troubleshoot. This will build trust over time and may solve the issue before they speak to someone → Show them it works! Displaying somewhere the % of conversations the AI is able to resolve with a message like: "56% of our customer's questions are resolved by our AI" A bit of a social proof trick here to show the majority of users are doing it and it's working → Ensure your bot gives direct answers Prompting the bot to give more direct answers (with the customer being told what to do rather than how to do it) may also help give more actionable responses. Any other ideas on how you can build customer's confidence in AI support agents? #customersupport hashtag#customersuccess hashtag#customerservice hashtag#csleaders hashtag#aiincustomersupport hashtag#aisupport

AI capabilities can create process efficiencies, enable knowledge retention and so much more if strategically approached. Before jumping in understand your end to end business processes and how AI can be a tool. Be cautious of putting all your eggs in one basket! Once you enter some tech ecosystems it is designed to make it difficult to leave. Include that in your tech matrix assessment!

Inspire by recent conversation with Peter Tu, creating a small serious of my thought toward AI and Customer Support. Post #1 - Will Customer Support Survive in the age of AI ? Ever since ChatGPT hit the scene, there’s been a looming debate: is human customer support on its way out? If you've ever been using the tool and only being able to get a subpar answer, you know how frustrating it can be. Even with the latest generative AI, the issue remains. In a recent conversation with a Peter Tu, we talked about whether AI can really replace human support. He shared how his four-year-old daughter now turns to AI for answers when she have questions. It made me realize, it’s not the technology decide whether customer support remain or not, it’s up to customer to decide. This fundamentally change my view on the recent AI boom. AI tool might be smart, but it takes time and effort to master at prompt engineering, and find the right way to engage with the tool, happen to be something majority of the customers don’t have. Until AI can truly "read between the lines," the human touch in customer support remains irreplaceable. What do you think? Can AI ever truly replace the human touch in customer support?

**Is AI Making Customer Interactions Too Creepy? Let’s Talk!** Here’s something that keeps popping up when I chat AI with others: “How much does AI actually change how businesses talk to customers...and is it crossing the line?” Picture this. You walk into a shop you’ve never been to before, and instead of a friendly “How can I help you?” they already know your size, style, and what you were eyeing online. Feels convenient at first, right? But then you think...how did they know that? That’s AI in action. Good AI predicts what you need before *you* know you need it. - Think Netflix with those eerily spot-on recommendations for your next binge. - Or Spotify, crafting playlists that just “feel” like they know your mood. But when AI gets this clever, the line between helpful and invasive starts to blur. Now, here’s where it gets sticky. What if that same AI is working off incomplete—or worse—*biased* data? What if it suggests products based on assumptions it made about you from a few clicks? And imagine a workplace where AI systems subtly favour one group of people over another without anyone even realising it. It’s easy to see why people feel uneasy. How do we balance this? Here’s the gut check for businesses using AI: - **Be upfront**. No one likes the feeling of their life being sniffed out without permission. Let customers know what kind of data you’re using and why. - **Stay ethical**. Train those algorithms well. If AI is pulling in flawed or biased data, it’s your responsibility to fix it. - **Be accountable**. When something goes wrong, someone needs to own it...not shrug it off as “a computer glitch.” AI is powerful, but it needs people with strong ethics steering how it’s used. The future of AI in customer interactions? It should feel like a helpful assistant, *not a mind reader that never asks permission*. https://lnkd.in/gbfege_3

Want to find internal docs more quickly - and skyrocket customer service to grow CLV? Listen in to tomorrow's AI Today. Teaser time... You've heard all about retrieval augmented generation (RAG) - storing your chunked internal docs and knowledge bases as embeddings in vector databases for swift retrieval. But it has drawbacks; AI often returns data that, while including the right words, doesn't include them in the right context. That's where contextual retrieval processing comes in. Or Contextual RAG. Those chunks are now labelled with additional information (which you can generate with AI) so the context in your prompt is now matched to context in your chunks. Context = customer satisfaction. Using customer service chatbots? Using contextual RAG, you can serve up answers that more accurately align to the context of customer queries. More on this brilliant development on tomorrow's AI Today!

AI chatbots are revolutionizing customer support by making it more efficient, cost-effective, and scalable. Their 24/7 availability, instant responses, and ability to analyze data enable personalized and consistent service. But challenges remain: chatbots struggle with emotional context and complex queries, and their effectiveness depends on training quality. As we look to the future, the question isn’t if AI will transform customer service—it’s how far that transformation will go. https://buff.ly/4fjEjpi Would you trust a chatbot with your most important questions? #chatbots #ai #artificialintelligence #chatbotdevelopment #technology #digitaltransformation #chatbotcompany #customerservice #bot

This is another great article from Laetitia Vitaud For me, part of the solution is to ask the positive questions. So if you are thinking about #AI for your business ask yourself: 1. How can AI improve the customer/ user experience? 2. How can AI make work more fun for my people?

Generative AI, led by innovations like ChatGPT, is transforming various job sectors. Notably, customer support is at the forefront of this revolution. The rise of chatbots powered by generative AI is reshaping the landscape, prompting a shift from human agents to automated systems, a transition that predates the emergence of language models (LLMs). With these advancements, the role of humans in staffing customer support lines and helpdesks may see significant changes in the near future. How will this impact professionals in the field? #AI #CustomerSupport #Automation https://lnkd.in/dD9hR3bQ

Have you started incorporating artificial intelligence into your customer experience efforts? 🤖 In our webinar How Ignoring Future Tech Will Leave Businesses in the Past, CustomerGauge CEO and Co-Founder Adam Dorrell teamed up with Owner and Director of Genroe Adam Ramshaw to uncover what AI is good for, its limitations, and how people can get true value from it through five AI + CX use cases. Adam Ramshaw shared our first use case on automating customer support — he believes training AI to do this is a fantastic starting point and something everyone should be doing. Because while AI isn't changing everything (yet), it can help industry professionals be more efficient now 💪 He also walked through additional use cases including: 🔎 Customer feedback analysis 🗣 Extracting themes and drivers from feedback 🤓 Account planning 🔮 Consulting the "oracle"— spotting trends to make improvements Adam says, "For a long time, this stuff has been deep in universities. It's never really caught on or been able to be operationalized I think, for me, until ChatGPT and OpenAI. People in all sorts of businesses are now able to use this tool. You don't have to have expertise in AI to use it, you just use it like another tool. But it's been a new dawn just in the last six months and I'm shocked at how fast it's changed." If you're interested in learning more about all five ways AI can streamline CX tasks, head to the comments ⤵