Reducing costs of Azure Databricks (FinOps in practice)
Big data is what powers companies. Lots of data mean lots of insights, which enables better decision-making. Having ...
We’ll use the example of the banking industry, but knowing your customer is key for every business. Why? Read on to find out!
Understand your customer – it matters
Based on Accenture research, 91% of consumers are more likely to shop with brands who recognize, remember, and provide relevant offers and recommendations. 80% of buyers want brands to understand them and know when to approach them and when not to.
This is very important since currently customers have a lot of options to choose from. Brands, or banks in this case, have to be very careful and very persistent in using the data well.
We all know that the data we share is being used. But customers, especially young people, want that data to be used wisely. They really value a pro-consumer attitude.
No time to read? Watch the video instead!
Irrelevant messaging worsens your customers’ experience
According to McKinsey, 65% of customers are frustrated by inconsistent experiences across channels. A situation when a person appearing in a digital web channel and the mobile channel is treated as two different people, still happens in today’s world. And consumers do not like it.
Because it shouldn’t happen! Companies should be able to connect this data and communicate with people, having all the communication history from different channels centralized.
What do we have next? As Accenture reports, 3 out of 4 customers are frustrated by the irrelevant content they are exposed to on websites. For instance, as a customer of a bank, I don’t want to get any messages about travel insurance if I am not going anywhere anytime soon.
The bank has all the ability to analyze my data and know whether these are my plans. Offering me something, which is completely irrelevant to my current circumstances makes me less loyal to the bank (or a different organization).
But having said that, we also know that managing customer data is not so easy. The information is there, but the banks and different financial organizations do not have the full ability to use it properly. Only 13% claim that they make the most out of the data, CDP Institute reports.
The biggest challenge they see is the inability to unify the data and lack of means to extract it. We all hear that we live in the era of big data. The data is flowing, and we know we can use it, yet we still don’t have proper ways to collect, unify and make use of it in the best possible way.
What’s the answer? Here comes our hero!
Customer 360 Overview lets us create a unified view of the customer data and to use the same channels to engage.
So, how can we use it to gain a competitive advantage?
We start with a not-so-small problem. As I mentioned before, consumers expect banks to provide comprehensive and personalized services and relevant communication. We know that communication gets less and less expensive in terms of the financial approach.
However, it is costly in a different manner, because whenever a customer receives something that is completely not related to them, they tend to get less loyal. That’s why we need to ensure that the messages we send are relevant and targeted.
What’s the solution?
Firstly, we want to gather all the data which, depending on the organization, will be different. For a bank, the main source of data will be obviously transactions.
This includes card transactions, card payments and transfers between customers. We should also include demographics, behavioral data, and much less traditional data, such as contacts with customer center, answers to campaigns or social media data (even if it’s pretty challenging to get those nowadays).
By analyzing all this information, we can create a general view of a customer and get to know them better. In time, you’ll be able to go beyond that and predict their behavior.
For example, if we have the transactional data, why not try to predict important life events? We may try to forecast whether a customer is going to buy a house soon, which would be an excellent opportunity to offer a loan.
What is the outcome of that?
We can create a very personal relationship. We can make use of the knowledge we have about our customers, to create focused communication and to reach those customers who seem to be the most relevant for our new offering.
You might wonder, how to do this?
And the answer is – Customer Data Platform!
What is the Customer Data Platform?
The physical representation of the elements of Customer 360 overview is a so-called Customer Data Platform. It consists of 3 main pillars.
How to build and use Customer Data Platform with Dynamics 365 Customer Insights?
Dynamics 365 Customer Insights is a real-time customer data platform from Microsoft, created to help its users drive customer-centric experiences. The key phases of building this solution are: ingestion, mapping and matching, conflation, enrichment, insights and action.
First, we need to ingest all the data we have – which means bringing customer and activity data from all sources. We can deal with structured data like traditional databases or unstructured like videos, images, etc.
Mapping and matching the data
The aim of this stage is to identify and understand customer data from different sources. We have some assets for that. Customer Insights also includes methods unify the data efficiently
What is very important here is source lineage. If risk departments ask us where the data comes from, we have to be able to quickly investigate that. This is especially important in the current times of high privacy and many risks associated with breaking it.
We have the data, we have connected and unified it. Now we should use AI to enrich it and generate new unobserved knowledge through machine learning models.
Data and the models are only helpful if we generate appropriate business insights on that. Understanding the data and the models is challenging, that’s why the graphical representation comes into play.
Finally, we need to take action through a personalized marketing campaign. Only then our data & AI analytics result bring the real value. Our main goal is to be able to transform the information on Customer Data Platform to engagement via mobile web, social media, or even bots – another piece of artificial intelligence at the very end of the process.
The structure of Customer Insights
Now, let’s take a more detailed look at the elements, which build the Customer Insights.
If we’re a bank, this will be mainly transactions, product usage, demographic data. But we can count on many different types of data, especially when it comes to external sources. If we are working with a partner, why not get this data to enrich our customer database?
We can also think about social media, e.g. interactions with our company profiles, as a huge source of behavioral analysis data. If we use a mobile app, why don’t we get the geographic data to show the true location of our customers and put them on the map? Customers Insights has the ability to automatically connect to many kinds of data we might need.
This is where all the magic happens: ingesting, matching conflating, and enriching. At this point, the Customer 360 profile is being created, so we can use it later.
There are probably two main approaches to the data we are getting. We can either create a more traditional analysis in the form of dashboards or a general overview. But AI and machine learning is a way to go here.
We have so many possibilities to use the computing power, e.g. with Microsoft Azure, and to create an action dedicated to each individual customer (even if we have millions of them!). It would be a shame to waste that.
Finally, we have the actions which we can take via different sample Microsoft’s services. Our main goal is to be able to transform the information on Customer Data Platform to engagement via mobile web, social media, or even bots – another piece of artificial intelligence at the very end of the process.
They are endless! Let’s start with marketing. We can create the segments, but not so-called “traditional”, a priori segments. We don’t just divide our population into “young” or “well-earning” customers. We are able to ask the algorithm to do that and to look for some previously unseen connections between our customers.
We can also use this data to score leads. If we introduce new products, we want to find these customers who are most likely to answer for the new offering.
Product recommendation. If we know that our customer is interested in traveling and has, say, two houses and drives three cars, this is an excellent knowledge for us to decide what kinds of products they might need. We are able to treat them completely differently to the general group of customers.
Another application may be providing service and proactive support via different channels. We don’t really want to wait for our customers to ask for the service. We want to be the first! The goal is to introduce them to some new products or service and provide an excellent care.
The picture below shows us what we are essentially aiming for. A general overview with dashboards of each single person with different characteristics, showed graphically.
We can use this information to engage and to develop an individual approach for every single customer, but what is very important and what should be treated as a final remark here is that we need to use this data wisely. Our customers give us the data and they trust us, so we have to engage with them carefully.
We cannot really approach a customer saying “I know that you are going to buy a house in two days”. This is not the way to go. We have to be very careful how we approach customers, but this is probably the topic for a completely different discussion.
If you are more interested in this topic specifically or anything related to data and AI, do not hesitate to contact me. I also encourage you to read my other data-related articles. They come from real-life implementations and include the most modern and advanced techniques.
Read similar articles