FAQ for Business Users

Answers to most frequently asked questions by Business Users

What’s the difference between Xpanse AI and Xpanse Insights?

Xpanse AI is used to rapidly build Predictive Models from databases. Using this tool you can quickly understand and predict key customer lifecycle events such as conversions, cross-sell, up-sell or customer churn. 

Xpanse Insights is an interactive insight discovery tool, which you can use to explore your customer data and generate meaningful segments or groupings. The easy to use point-and-click interface means anyone can explore the data in real-time and produce actionable insights at the click of a button. 

Essentially, Xpanse Insights gives you the understanding you need about your customers without generating the Predictive Models you’d get with Xpanse AI. 

Who should use Xpanse Insights and Xpanse AI?

Both Xpanse AI and Xpanse Insights are designed and built in such a way that any member of an organisation can be up-skilled to get the most out of the features and functionality. However, we have found the below examples of typical users to be helpful when trying to decide what solution might be right for you. 

 

Xpanse Insights is for anyone who relies on data to deliver their KPIs. For example, if you rely on your data to answer questions relating to revenue generation or churn reduction, then the chances are you will find value in Xpanse Insights. 

Xpanse AI is razor-focused on Predictive Analytics so, ideally, it should be used by someone with a good understanding of your data and an ability to read the accuracy of Predictive Models. However, we’ve made sure that the tool is built in such a way that this knowledge is not essential and we can train up any member of your team in these areas.

I’m not a Data Scientist. Can Xpanse Insights help with my business?

Yes! The only thing you need as a user of Xpanse Insights is a burning desire to answer your most enigmatic data questions. Our aim is for you to get the most meaning from your data with as little effort as possible. 

For example, Xpanse Insights can help you answer complex questions like does customer tenure have an impact on churn? Or, what are the shared characteristics of my high value customers? These are the questions being asked every day by business users and Xpanse Insights helps them find the answers...without needing a PhD in data analytics. 

What makes Xpanse Insights different from other software?

Xpanse Insights is a bespoke tool that is implemented and calibrated to respond to the unique characteristics of your data. It is built to generate insights and profiles that are most meaningful to your industry and business. 

Unlike other tools, the focus of Xpanse Insights is on Customer Lifecycle analytics. It's not a traditional BI in that it is not limited by calendar timelines that may have no bearing on how your business interacts with your customers. You choose how to profile your customers based on key behaviors e.g. conversion, revenue, churn etc.

Are there hidden costs for on-going support with the software?

We pride complete transparency with pricing and our excellent after-sales service. We provide full support during the hours you need it most, so our team is available to respond to your queries 9 to 5.30pm, Monday to Friday at no additional cost.

There is a lot on your website about Automated Feature Engineering. What is it?

You might be relieved to learn that there is actually no consensus about what “Feature Engineering” includes. Linguistically it’s somewhat of a mess. But when we use that term about it we are essentially talking about the following:

  1. Generally, it relates to the essential step of preparing a dataset for Machine Learning. 
  2. We refer to Feature Engineering as the process of transforming data from a relational database to one specific table (called Modelling Dataset) digestible by Machine Learning. 
  3. The “automated” description means that the process is executed totally autonomously by a machine instead of a human data scientist.

So really the term Automated Feature Engineering relates to the process of intelligent automated data transformation for use by Machine Learning. If you want more information on this you can find more in the FAQ section for data scientists. 

Churn Models, Cross-sell, Up-sell, Conversion, Risk- what are all those models for?

These are all examples of Predictive Models (also: "Propensity" or "Affinity" Models) Models. The models point to a specific type of customer behavior and the data will be analysed at great depth to build a propensity (or likelihood) for the desired behaviors. This means you will be able to tell how likely an individual customer (or a group of customers - called a segment) is to behave in a certain way. 

While they may be used to predict different things, all of these propensity models have some things in common:

  • They are built on historical individual customer data.
  • They combine the historical and current data to generate a “propensity score”.
  • The scores are assigned between 0 and 1 and they describe the likelihood of the individual customer to carry out a defined behavior. 

The process of building and deploying the models is the same, regardless of the application or industry, which is why we can transfer Xpanse across all manner of businesses and organisations. 

But let's break them down in more detail...

Part 1: What are Conversion and Up-sell Models?

Conversion and up-sell are the same in that a customer is moving from one definition to another. So a customer might move from being a cold lead to a user on a free-trial version of the product. A non-paying user may convert to a paying customer, or a bronze value customer may move to gold value. In all those cases cases, we use historical data to build an understanding of what drives the change in behavior.

The Models assign a score to each customer reflecting their "propensity" to move towards a better version of our product or a different definition of behavior. 

This information allows you to increase your conversion or up-sell performance by focusing only on valuable, high-propensity segments.

Part 2: What are Cross-sell Models?

Cross-sell Models are not that much different to Conversion & Up-sell Models in that you are trying to understand the drivers behind changes in customer behaviors and assigning associated scoring.

In the case of Cross-sell Models, the scores will reflect a customer’s interest in other products or services and their likelihood to carry out specific actions. This scoring can be used to define high-propensity segments for targeted, intelligent marketing campaigns. 

Part 3: What are Churn Models?

A Customer Churn Model examines how likely it is for each customer to stop using your product or service. When a score is assigned to these customers you will be able to ascertain how likely they are to stay with you (be retained) or leave (churn). 

When you have identified “who” is churning, you are then in a better position to understand “why”. For many companies, discovering churn triggers can lead to massive improvements in processes or offerings by reducing customer pain points and maintaining relevance. 

Once you deploy the Churn Models all you need to worry about is generating a compelling retention strategy for these high-risk segments so you can finally say goodbye to saying goodbye to your customers! 

More questions?

Don’t take our word for it...

Book a no-strings-attached consultation today and find out how Xpanse AI and Xpanse Insights can make an impact.