It’s no surprise that customer support is one of the first industries that is waking up to AI adoption. The incredible capability of AI to analyse and generate answers has everyone hopeful. Sam Altman, the CEO of OpenAI, has usually held back when he’s been asked about AI having a major impact on roles and industries, but even he had to admit that customer service could have a “huge impact on it.”
Image Credit – David Paul Morris for Bloomberg via Getty Images
Is AI just hype?
We’ll find out, but more likely than not, AI is going to stick. Yes, yes, you and I can be forgiven for being skeptical especially after the most recent wave of crypto, web3 and metaverse fads that came and went.
But the key difference here is obvious – you can see how AI can be useful. Look, if you’re reading this, you’ve mostly used it to. You must’ve been equally mind-blown, confused and wondered about how ChatGPT is coming up with these responses? You and I have both tried to see “if it can do that again”.
The proof is in the pudding, we’ve actually been able to see how AI works and what it can do. Now it’s just time for people to start building and adopting it. Will it change the way we’ll all work? I don’t know. Maybe. However, I’m certain it’ll amplify what we all do in some form or fashion.
Where AI can help customer service
AI is going to shine in resolving customer questions instantly, ultimately allowing support teams to focus on more complex and time intensive tasks.
For Shopify stores, you’ll see AI really helping in the following categories
- Common Questions & FAQs – A large category of e-commerce tickets are around order tracking, returns, cancellations and other post purchase questions. This pattern is prime for AI to resolve these questions immediately.
- Automated Workflows – Cancellations, order tracking, returns can all be initiated by AI, and can be brought to a point where support teams have to approve or reject the action.
- Product Questions – The AI can understand your products and help instantly with questions around them. The amazing part is since AI can perform semantic examinations, it doesn’t need keywords from your product to find the correct answer.
With a combination of your Shopify products, collections, help centre data and other information, you can create a vast and robust knowledge base. This will allow the AI to provide accurate answers and workflows.
What AI-powered CS tools will look like
At Macha, we’re working towards amplifying customer support specifically for Shopify stores which integrate with CS tools like Gorgias and Zendesk.
We’ve taken this approach for the following reasons
- AI for Shopify allows us to create workflows that are specific to e-commerce like order tracking, returns and policies
- It allows us to anticipate and reduce hallucinations around e-commerce, since we can build appropriate guardrails
- Integrating with Shopify helps us create workflows and interfaces that are specific around e-commerce.
We’re approaching baking AI into customer support in two broad ways – Issue prevention and issue resolution. I explain what they mean below.
Issue Prevention With AI
As the title says, we can instantly resolve a large number of questions before a ticket is even created. This happens via a chatbot interface. Common questions and FAQs can easily be intercepted and resolved at this stage before it makes it to a support teams ticket queue.
We’ve taken great care to approach this with empathy. We know everyone hates being stuck in a chatbot loop, trying to crack a code to make it to a human. To address this, we build experiences where we make customer comfortable and aware of what they can expect from the AI. And more importantly, make them aware that at any point, they can reach out to a human support agent if they’d like. This helps in drastically increasing customer trust.
Issue Resolution With AI
You’re already using CS tools to handle tickets. You probably have setup macros for your teams. However, you’re still receiving a large amount of repetitive tickets which agents have to identify and select the appropriate macro.
The scenario above is exactly where AI is going to add significant values. For tickets like that, Macha works towards resolving them immediately, tagging and categorising them appropriately so that your team doesn’t have to and can work towards seeing if that cancellation request is within your policies time frame.
Do I need to use AI for my support now?
My opinion is obviously biased, so don’t put much weight to what I say. But I do feel that it’s inevitable. And why not? I think it’s amazing that AI can help democratise support experience regardless of size and scale of companies.
Support heads will change focus and start asking the question – “Why does my support team need to answer that”. We’ve already taken a step in the direction of automation by introducing FAQs in our live chats, macros for our agents. This is another step in that direction.
I’m super excited to be in this space and helping brands embrace AI for their customer support!