CRM & Artificial Intelligence: A few Fantastic Use Cases

They call it Artificial intelligence, machine learning, natural language processing, predictive learning, predictive analytics… irrespective of the buzzword you select to call it, this technology is not new. People are talking about AI for a few decades.

But, thanks to cloud computing, big data and (almost) infinite processing power, time is ripe now for AI to have some practical applications, especially in the area of CRM.

Sales, Service and Marketing: all three pillars of CRM can benefit from AI. Many organisations have already started using this to improve their decision making and to enhance their customer experience.

Today I attended an event from Google where they gave two fantastic use cases for this. It was not a CRM event, but I could see how well Google was able to use AI to solve some CRM problems:

1

A supermarket has a free home delivery option. Customers can buy stuff, provide their addresses and a van will deliver their goods within next 12 hours.

By doing this, they are capturing the exact address (latitude, longitude) of where the customer lives: with the permission of the customer.

So, after few months, they have many addresses in their database. They send few discount coupons to these customers. Some of those coupons say 20% discount, while other coupons say 30% discount.

Why? Was that random?

No. They analysed their data and learnt which customers ***may*** need additional discounts.

But, how? Did they read their customers’ minds?

Fortunately, such a technology doesn’t exist. So, the company did the next best thing: predict (or ‘guess’) who needs more discount, by following a machine learning trick:

• Take each customer’s address
• Using Google Maps, find the driving route they would take to come to their supermarket
• See if this route has a competitor supermarket on the way
• ? If yes, send 30% discount voucher: incentivize them more, to drive that extra distance

2

Customer service manager of a food delivery service sees his daily customer feedback report. It says 97% people have given a five-star rating for their service.

Wow. 97%. Time to Party!

But, that manager didn’t party. Because his system found that all those bad comments (3%) mostly came from one particular (geographical) area. It showed a heat map chart which made it clear.

The manager goes one level further down; he learnt that most of the bad feedback is about delayed delivery due to traffic problems in that area. Now, he can either change the “promised delivery time” for that area OR Change the delivery plan in that area so that customers are served from more than one kitchen.

Not just food delivery, any business whose quality of service is dependent on traffic status should start using predictive intelligence. For example, pizza shops needn’t have rigid, permanent “Service Boundaries”. Instead, each shop’s service boundary can shrink or grow dynamically if the traffic is too much/too less.

Consider this: I usually serve a 3KM radius around my shop, but, during peak hours, traffic moves very slow, So, I can only serve 2KM radius. Similarly, when the traffic is less (holidays, late night hours), I can stretch up to 5KM radius, maintaining the same level of service.

So, what happens when the service radius goes down from 3KM to 2KM? A sizable number of customers won’t be served?

Yes. Can mobile pizza vans start operating in some strategic locations during peak traffic hours? Can your CRM predict those locations and time slots based on live/ historical traffic data?

AI is fascinating and a right fit for enhancing, getting best value from your CRM. We should start asking questions, making solutions. Technology is very much available.

Image Courtesy: https://pixabay.com/en/customer-family-magnifying-glass-563967/