How Google Cloud Platform and artificial intelligence support the e-commerce industry

Today, around 16% of retail sales are made via the internet. It is predicted that by 2023 e-commerce will take over 22% of the market. The sales sector is becoming more and more digitised – even stores that only sell on a stationary basis benefit from the support of new technologies in logistics, product lifecycle management or demand forecasting. One of the branches is artificial intelligence, which can take over some of the employees’ duties or completely relieve them in some tasks.

Google Cloud Platform services portfolio includes AI solutions dedicated to the e-commerce industry. They support, among others, customer service, product image search, or intelligent recommendations.

What is artificial intelligence?

Artificial intelligence (AI) is a branch of computer science that involves the creation of machines capable of self-learning. Intelligent machines are supposed to imitate the way people think and act, solve problems, draw conclusions and develop their knowledge. Technology is supposed to support people and take over tasks from men – from simple to very complicated.

Artificial intelligence is used, among others, in:

• healthcare – AI can support self-diagnosis of patients, thus minimising the workload of medical facilities,
• the financial sector – AI, by analysing current and historical data and relationships between them, helps to make business decisions,
• robotics – Roomba 980 vacuum cleaner model uses AI to scan room dimensions, identify obstacles and remember the layout to optimise cleaning,
• social media – e.g., Twitter tracks and removes posts containing hate speech or supporting terrorist activities (in the first half of 2017, Twitter banned 300.000 accounts connected to terrorist organisations – 95% of them were identified by artificial intelligence).

Examples of the use of artificial intelligence in e-commerce

The e-commerce sector is also experiencing many benefits from the development of artificial intelligence. Technology lets you relieve employees of simple, repetitive tasks, giving them more room for valuable, creative activities or solving complex problems. AI also helps to read customer needs on a large scale, conducts analytical activities and supports data-based decision making.

Examples of activities that AI can carry out in e-commerce:

• presenting to the client highly personalised article suggestions that may interest him,
• enabling the client to search for a product based on a photo or graphic,
• giving answers to customer questions via chat or hotline,
• management of communication channels, e.g. directing hotline or chat calls to the most appropriate service agents,
• reading the intentions of the interlocutor and supporting service agent during a conversation (e.g., by presenting hints or useful materials),
• forecasting sales trends,
• tracking potential problems in the functioning of the store or the sales process,
• process optimisation service and sales by analysing factors such as customer location, demand, level of satisfaction with service or product, product availability, shipping time or package dimensions,
• performing simple actions such as accepting returns or complaints,
• leading inventory of products or giving a support in this field.

Artificial intelligence on the GCP platform – solutions for e-commerce

Google is aware of the requirements that customers place in front of online stores. The attractive price and fast shipping stopped playing the leading role a long ago. Now, apart from the speed of operation, stability and availability of the website, what also counts is the ease of making a purchase, personalisation of the offer or flawless service.

Many stores worldwide – small and large – use Google Cloud Platform solutions to provide their customers with the best possible experience. The GCP cloud ensures stability, security and scalability. Additionally, it gives the possibility of conducting advanced analytics or support for service and sales thanks to machine learning and artificial intelligence.

The portfolio of Google Cloud Platform includes services that, using artificial intelligence, allow you to provide customers with positive experiences, give the team relief and at the same time increase conversion in the online store.

Product search by image – Vision Product Search

Vision Product Search is a service that allows you to search for specific or visually similar products among the assortment. The customer takes a picture of the article or posts a picture of the product, and the system redirects him to the card for this or a similar item. Thanks to this, customers can immediately find a product without manually searching the web or store page and opening multiple subpages or tabs. This is an excellent convenience for the customer, making it easier to reach the product also affects the level of conversion.

IKEA uses this GCP service. Susan Standiford, Chief Technology Officer at IKEA, says that Vision Product Search can further simplify and shorten the purchasing process. Customers have direct access to IKEA products wherever they are. This translates into a better customer experience and greater brand openness (which is one of the company’s core values).

Personalised product recommendations – Recommendations AI

Customers browsing online e-commerce shelves expect to receive recommendations of similar or complementary products to their basket. GCP’s Recommendations AI service automatically creates highly personalised product suggestions. The mechanism considers the customer’s path, subpages viewed, products added to the basket, changes in the assortment, product availability, prices, promotions, or special offers. Recommendations AI works in real-time and adjusts the proposals to the client’s identified needs on an ongoing basis. By constantly analysing the behaviour and reactions of recipients, the mechanism learns and creates better and better matches.

Kathy DePaolo – VP of Engineering at Disney, shared her observations on the use of Recommendations AI in reading customer needs during the Google Cloud Next ’19 conference.

Customer service support – a group of Contact Center AI services

Service specialists often have to answer simple, repetitive questions – for example, about the delivery, payment method or order status. This slows them down, distracts them from handling more complex requests, and is quite simply frustrating. On the other hand, customers get stuck in lines at the hotline, are sent away from the agent by the agent or wait hours for a simple query to be answered. Contact Center AI is a group of Google Cloud Platform services, with the help of which you can build your own tools to support service and increase customer satisfaction. Contact Center AI, along with other Google Cloud Platform services (e.g. Text-to-Speech, Speech-to-Text or Natural Language), allows you to create, among others:

• tool automating the order of calls and messages between available service agents, directing the client to the most competent specialist,
• own communicator with a chatbot or IVR system that can run automated customer service using natural language (it is possible thanks to the Dialogflow platform),
• tool for building and development of a self-service system with a virtual agent available 24/7; chatbot can drive personalised interactions, provide the customer with answers to simple questions or decide to redirect a more complicated problem to a customer service agent,
• support system for agents working on the hotline; the system can transcribe the conversation in real time, read the intentions and attitudes of the interlocutor using semantic links, and provide the agent with the necessary materials or tips on how to conduct a conversation step by step to solve the problem and ensure a high level of customer satisfaction.

AI in the service of conversion and customer satisfaction

Contrary to common fears, artificial intelligence will not take our jobs. Yes, it will take over some of the duties, thus giving us space to perform complex, more satisfying tasks. AI supports decision making that is based on numbers and results. In many areas, it analyses data much better than humans – that’s because it can see the broader perspective. Within seconds, it processes data layers, searches for numerous connections, diagrams and concludes.