Business challenge of a global Fortune 500 retailer
Our team was commissioned to build an AI chatbot for one of the world’s retail giants as they prepared to release a new lineup of top-class consumer goods. Consistently ranked on the Fortune 500 list, our end client is committed to enriching the rollout of new product lines with innovative solutions for all distribution channels. As chatbots in the retail industry have become instrumental in promoting new brands and boosting sales, the company wanted to implement this technology in their daily operations.
The new software was meant to serve as an eLearning solution to train salespeople at stores and outlets on how to educate their customers on the main features and benefits of the company’s products. By equipping sales managers with ample information on products, the training software would maximize their engagement, provide seamless assistance to buyers, and give the company a competitive edge by delivering enhanced customer experiences.
Considering the strict deadlines the company faced, they were looking to work with a proven partner with successful experience developing machine learning algorithms. They also needed someone who would quickly grasp all the specifics of their product, as chatbots in retail vary in purpose. Our deep expertise in eLearning solutions and, in particular, in implementing machine learning algorithms played a decisive role in their choice of service provider. They chose Intellias to take on this challenge, and after discussing the main requirements, the project began as quickly as the client had hoped.
Retail chatbot we delivered
Preparing for the launch of a new product, our end client required a retail chatbot to increase learner engagement among sales representatives and clearly explain to potential buyers all the advantages of their consumer goods.
To provide end-to-end product development, we gathered a team of skilled engineers that included a solution architect, five .NET developers, two QA specialists, and a project manager. Our team developed the backend and frontend for the chatbot, which included designing the UI/UX and implementing machine learning algorithms and modern technologies for natural language processing (NLP).
We started by defining the key features of the chatbot:
- A database containing product details
- A fast on-click personal guide to assist sales managers on the go
- An evaluation system to train and develop salespeople
We worked through the bot architecture, created the content, and suggested a set of functionalities to enrich the product with new standout features.
Natural language processing
The retail chatbot we were building had to understand human language and respond to questions about a product’s characteristics. We created an application with extensive AI capabilities that helps sales assistants get instant replies and provide ad hoc consulting to customers so that customers don’t need to flip through catalogs looking for products.
The chatbot recognizes inquiries by processing natural human language and understands questions even if they’re asked in a unique way. Machine learning algorithms find similarities in queries and adapt based on collected variations on the same question. If a customer asks about a product, the chatbot will recognize the topic and give relevant information or shopping advice. For example, customers can ask:
- Where can I buy it?
- What colors does it come in?
- Can I see some photos?
- How about spare parts?
- What new features does it have?
- How much does it cost?
The system processes requests and replies to the contents of queries, giving advice on how the customer should handle a specific scenario or even helping them make a choice.
The chatbot is an interactive eLearning solution that teaches salespeople in a conversational manner by asking the most popular questions about a product’s specifics, giving feedback on responses, and offering tips on what can be improved. It also features a test to monitor users’ learning progress and grade performance. With the data analytics we implemented, management can compare test scores with sales levels to assess how much eLearning helps to generate revenue.
Integration with messaging platforms
The chatbot interacts with sales managers through chat in popular messaging programs. It’s integrated into Facebook Messenger, Skype, Telegram, web apps, and other social media platforms so users can access it anywhere. Still, it’s a robust solution strictly protected from unauthorized access. Only employees of the company have access to this chatbot’s functionality.
Initially, the chatbot supported only English. But as the next step in advancing this software, our team scaled its architecture to support Spanish, Arabic, and Russian to reach more markets.
User data analysis
The chatbot application can generate insightful reports that can then be used to improve training, services, and the overall customer experience. It enables monitoring of the engagement of local staff across regions and analysis of individual and overall performance. Regional managers can access information on users, countries, and visit times, the number of sales reps entering the system, the pages they browse, their learning progress and test scores, and FAQs from customers.
Technologies used: .NET / Angular / MS Azure Text Analytics API / Cosmo DB / Power BI / Microsoft Bot Framework / LUIS (Language Understanding Intelligent Service)
Introducing AI-driven chatbots in retail has a tangible impact on businesses that place a high priority on customer satisfaction. Intellias implemented an innovative solution that works globally and is a powerful tool for launching and propelling new product brands. It allows our end client to speed up training of sales staff on freshly released goods and leads consumers to make more purchases.
The retail chatbot we built has proved its efficiency in simplifying the buying process and ensuring a unique and satisfying shopping experience. In turn, it has led to an increase in customer loyalty and brand exposure. The chatbot is also a handy way for regional and local executives to evaluate the performance of employees and give recommendations in an informal and friendly manner.
The solution we’ve developed allows our end client to:
- Reduce operational and service expenses
- Wow buyers with a new-age platform
- Increase engagement with customers and touchpoints
- Multiply reach, breadth, and depth of engagement
- Get rich analytics and ensure customer interaction
- Provide an instantaneous and accurate response without waiting on people
- Offer detailed explanations and personalized recommendations to buyers
- Improve marketing strategies and targeting