This is a particularly tough time for online retail, and so finding solutions that are going to actually drive increased conversions online is essential. There is a lot of noise around generative AI tools like ChatGPT and how they could transform the shopping experience and do just that.
But how much of this is feasible, and how much is just noise and buzz? Plenty of retailers are talking about using the power of AI, but there haven’t been a lot of results published from these early experiments. Some sources state that adding live chat itself can increase conversion rates by 20%, and if this can be automated to the same level through AI then the impact could be significant.
Let’s look at the main use cases that generative AI can be put to and how they could drive conversions.
Virtual Shopping Assistants
When you encounter a great shop assistant – one who knows their product range inside out, and can match you to the right product for your needs – it’s an unforgettable experience.
This has always been one of the most important gaps between the in-store shopping experience and the online experience. Online you can get a real sense of the breadth of products and can more easily find the product that you want – but it does require you to know what to search for.
In-store, you can have the benefit of not only seeing products, but being able to talk to a product expert directly. That expert can be the difference in making the right purchase, or even making a purchase at all.
So how can you replicate this online, and how can AI tools help?
If you have a live chat with a customer service agent on your site, then it’s the same principle. A chat window would be available throughout the customer journey, but in particular it could be prompted to appear to help customers.
Rather than a human agent however, a chatbot run by generative AI would allow customers to type whatever question they might ask of a shop assistant, and the generative AI could respond.
Where it gets clever is with training the AI on proprietary brand information. So, using product descriptions, FAQs, policies and any other information that a store assistant could use, the AI can be trained to answer specific questions with the level of expertise of a store assistant.
This can help customers to find the right information they need in order to convert to a purchase. So for example, you could ask the chatbot if these shoes are the right ones for the race you are running, and by checking all the product information available, the chatbot could give a confident, and truthful answer.
The trick will be in how it’s deployed – and when retailers draw customers’ attention to it.
Upselling and Cross Selling with Recommendations
One thing a real shop assistant can do is give tailored recommendations based on what the customer is telling them. For example, if you are looking for clothes for a wedding, an assistant would instinctively know not to recommend casual clothes. And if you were looking to accessorise an existing outfit, or add a matching item, then the assistant could give you their opinion on which items would go well together.
Traditional, preset chatbots can’t handle this kind of context, but the new world of generative AI can.
As in the example above, this product knowledge could be used to train an AI, and certain items could be grouped together as suggested outfits. Similarly if the request was for a wedding, then items could be tagged as appropriate for a wedding, and the AI could reach for those as suggestions.
One advantage that an AI-powered chatbot could have over an in-store assistant is access to a shopper’s order history. There’s few things more annoying than being recommended a product you already have, or one you sent back because it wasn’t right for you.
By being able to see the order history, the AI could avoid these mistakes. For example: “I can see that you already have the slim fit jeans, so if you are looking for something else to go with this top, I’d suggest that this skirt would work.”
By making these recommendations, the AI can upsell or cross-sell to your customers directly, increasing average order value, and driving conversions.
Personalising the customer experiences with AI
One of the fears with AI is that you de-personalise the experience for shoppers by making them go through the cookie cutter conversations where everything is pre-set. Generative AI can solve that by creating multiple versions of the same message to create a sense of spontaneity in conversation – in other words making it feel like a real conversation with a human.
But what is much more powerful within personalisation is using integrations with a customer’s order history and other information within the CRM in order to tailor the conversation. If you knew that a customer is a VIP or frequent shopper, you’d speak to them differently than if they were a new shopper for example.
Taking it a step further, if you knew what a customer had ordered last week, you would take that into account when you spoke to them.
If your AI chatbot can be connected to these backend systems then it can have the necessary context to be able to create a personalised response that is highly contextual to that individual.
Round the clock, white glove service
One issue with providing live chat is that your agents may tend to work office hours, which is almost certainly not when most of your customers are shopping. Otherwise you are spending a lot of money to provide round-the-clock service.
It goes without saying that an automated chat function wouldn’t have that same drawback and would be able to provide the same level of service throughout the day and night.
Being able to provide consistent 24-hour customer service could allow you to drive conversions all throughout the day.
Freeing up customer service agents to become sales people
Fundamentally AI automates tasks. Right now a lot of the tasks it automates are responding to simple repetitive customer enquiries. These are the kind of enquiries that don’t take your agents a lot of time to answer, but they have a mountain of them to deal with every day.
These are the functional questions like “Where is my order?”, or “How do I return this?”. AI can deflect these tickets and resolve them away from human agents. This in turn frees up the agents to be able to actually sell to shoppers. Or to contact existing customers proactively, check how they are getting on with their purchase, and potentially drive new sales.
Accessing shoppers on ChatGPT
Since launch, ChatGPT has reached a million users faster than almost any other tech product in history. People are using it for a variety of purposes, but one of them is effectively to look up the answers to questions that otherwise they would have used Google or another search engine for.
This is a potential audience for your brand, and using the ChatGPT interface, you could inform users about your brand and include pictures of your product range.
So if a user asks a question like “Could you recommend a leather jacket with a slim fit?”, your products could be featured along with pictures and descriptions directly from your side.
This is a potential new way to acquire and convert shoppers, much like Instagram or TikTok before it. If a brand is an early adopter it could reap the benefits ahead of the competition here.
Can Generative AI drive conversions?
The short answer is yes it can. But it’s not as simple as switching it on and letting it run. The AI needs training on your brand, it needs to see your customer journeys, and then you will need to optimise it to make sure it appears at the right time and takes the correct actions at the appropriate stage.
Testing, optimising and measuring how it affects conversion will be key. But there is no reason why it cannot help sell directly on site, and also indirectly by freeing up your team to be able to sell directly themselves.
If you want to find out more about how you can deploy Generative AI in your business to drive conversions, solve customer problems, and free up your agents to drive more conversions get in touch with our team today.