Can you imagine your movie nights going back to TV Guide and the linear TV programs? No Netflix! No recommendations! No personalization!
How many times have you read or heard of AI technologies to be used for shopping and you said to yourself "This is not applicable to my retail business." or "AI is just another buzz word for something expensive I don't need."?
AI is a wide-ranging branch of computer science concerned with building smart machines capable of performing tasks that typically require human intelligence. This is what’s going to reshape retail in such a disruptive way and with such a level of convenience that soon we will not be able to remember how retail was without AI.
As a starting point, it is very important to get a basic understanding of how AI systems are built and embedded into every day technologies (products and services) by the most technologically advanced companies. These companies accelerate their competitive advantage through AI. And more importantly, some of these companies started democratizing their AI capabilities to other companies, including start-ups.
This is a time when AI is still a niche technology, but easier to start with nowadays more than ever before.
- STEP 1: An AI system is built using machine learning and other techniques.
- STEP 2: Machine learning models are created by studying patterns (deep learning) in the data.
- STEP 3: Data scientists optimize the machine learning models based on patterns in the data (looking to optimize key ML model performance metrics like Median Average Error, Median Average Percentage Error & Confidence Factor)
- STEP 4: The process repeats and is refined until the model’s accuracy is high enough for the tasks that need to be done.
- STEP 5: The service is operational, and the accuracy is continuously improved based on real process data and results (feedback).
Business performance has always been directly corelated to capabilities and how these capabilities create competitive advantage and market value.
AI systems can help any business to scale up and accelerate by performing tasks that typically require human intelligence.
The second important aspect is to have a basic understanding of the capabilities of AI systems:
- Predictive & Prescriptive Analytics: This capability helps companies predict trends and behavioral patterns by discovering cause-and-effect relationships in data.
- Speech Recognition & Natural Language Understanding: Speech recognition enables a computer system to identify words in spoken language, and natural language understanding recognizes meaning in written or spoken language.
- Sentiment Analysis: A computer system uses sentiment analysis to identify and categorize positive, neutral and negative attitudes that are expressed in text.
- Recommendation Engines: With recommendation engines, companies use data analysis to recommend products that someone might be interested in.
- Image & Video Processing: These capabilities make it possible to recognize faces, objects (incl. text) and actions in images and videos and to implement functionalities such as visual search.
The third aspect is to have a clear, simple and practical understanding of how AI technologies are applicable to your retail business.
There are 5 levels of AI based on their capabilities and their impact into your retail business:
L1: Classification: “Tell me what this thing is”. Knowing what something is, the first step in deciding what to do with it.
- You receive an email. The system tells you it is about a customer complaint
- You receive an invoice by email. The system recognizes it is an invoice and all its financial data
- You use CCTV cameras to recognize products on shelves
L2: Categorization: “Group these similar things”. By knowing the categories (groups), the system can better learn the relationship, similarities, and differences.
- You have 1,000,000 customers. Separate the young moms from all the rest
- You have 1,000,000 customers. Group them into hierarchies of their Lifetime Value combined with Life stage
- You have 1,000,000 customers. Label each customers with their loyalty value
L3: Assessment: “Tell me whether I should care about this thing”. The system adds contextual clues, generating a sense of urgency, ranking, and prioritization.
- You have 1,000,000 customers. The system tells you which are dropping off
- You have 1,000,000 customers. The system tells you which are loyal
- You have 1,000 suppliers. The system tells you which are breaching their SLA
L4: Recommendation: “Tell me what to do about this thing”. The system begins to incorporate AI/ML outputs into business workflows. And suggest actions and objects.
- You think you have 1,000 customers that are about to drop-off. Now what?
- You have 1,000 customers complaining at the same time. What should you do with each of them?
- You have 10,000 highly loyal customers. You want to find another 20,000 that are most likely to become highly loyal as well.
- You want each users on your mobile app to be displayed different content (ads, offers, products) based on their interests.
- You have to manage supply purchase. You want the system to tell you the optimal quantities for your new orders
L5: Prediction: “Will this thing happen?”. Prediction is the golden ticket of artificial intelligence. The holy grail. The system enables the retail business to become proactive.
- Will my sales in this store drop or increase in the next 30 days?
- Which customers are most likely to download our mobile app?
- Which suppliers are most likely to miss on their delivery SLAs?
- You are a store manager. You want to see the predicted traffic for the next hours in order to do labor allocation.
- Which customers are about to become parents?
- If you change the price of an item, how this will influence sales volumes?