By Emma Woodward-Church, Senior Strategy Director at Profusion. 

Read part 1 here.

Hello again! In my last blog we touched upon the new revolution of automating predictive modelling called AutoML – auto machine learning.  

We established that AutoML is the next step to making AI and machine learning accessible to everyone. It represents a fundamental shift in the way organisations of all sizes approach machine learning and data science in the future.   

AutoML covers ‘Plug in, automate and go’ algorithms that identify the critical insights hidden in your data including, for example, propensity to buy, at risk of churning/lapsing and customer lifetime value, to name but a few.  

The main benefit of AutoML is to streamline the data scientists’ project delivery process so they can do more with less of their time, enabling greater focus on higher value activities and ultimately enhancing the efficiency of your data team and wider organisation. 

In this edition of my blog, I would like to present to you Ai Marketer.  It combines the best of automation with expert practitioner insights to deliver a series of robust predictive models to super charge your marketing efforts.  

Ai Marketer has been especially developed for digital marketers. It minimises the complexity and enhances the accessibility of machine learning for organisations of all types. 

But first, a bit of background: Ai Marketer was designed by Profusion in collaboration with the University of Essex, home to the Institute for Analytics and Data Science. Leveraging state-of-the-art expertise in artificial intelligence, data science, statistics, natural language processing, UX and business knowledge we shaped the Ai Marketer suite of modules over time.   

We are proud to say that Ai Marketer won an award from the government’s Innovate UK programme for the best academic partnership into new product development. 

So how does it work? 

Ai Marketer is an agnostic tool that plugs straight into your preferred data selection environment –CRM, CDP, DMP or otherwise – in the most user-friendly way possible.   

The goal is to provide marketers with direct access to predictive models (the cool stuff) for direct output decisions and tests, across any channel. It adds another layer to your existing data analytics and modelling capacity and capability. 

As the background information states, Ai Marketer is powered by artificial intelligence, using the latest techniques in machine learning to identify each individual customer’s behaviour to enable highly personalised and targeted engagements & marketing campaigns. 

The mathematical modelling is built to enhance what you learn about your customers and enable you to make best-next-action predictions about whom to target, when to target them enabling you to build prediction modelling and make smart moves quickly.   

Ai Marketer has six predictive models built in, ready to be trained on your data, which means unique outputs based on your customers and the next best action for them.   

  • Customer Churn and Propensity to Buy models help to judge who is ready for sales-based messaging. Win back messaging for customers at risk of lapsing and ‘next best action’ messaging for customers ready to buy.  
  • Customer Lifetime Value will help you grow your ‘next best’ customer segment by spotting future VIPs from day one, enabling you to keep them thirsty.   
  • The RFM model is more descriptive of the current situation. However, we use machine learning to define customer recency frequency and monetary segmentation; a data driven, bottom up approach. Essentially the model creates a customer segmentation with 4 different categories enabling you to target incentives effectively: 

a) low value – low frequency spenders 

b) low value – high frequency spenders 

c) high value – low frequency spenders 

d) high value – high frequency spenders 

  • Our Inbox Accelerator model sends your ‘business as usual’ emails, using predictive intelligence to segment high value, highly engaged customers fast and first.  This includes the added benefit of enhancing your sender reputation making emails more likely to reach the inbox.  
  • The Send Time Optimiser model will know when they’re most receptive to receive each message achieving higher engagement rates and also great sender reputation benefits. 
  • Frequency Capper will take a predictive yet holistic view of contact touch points to ensure you ‘cap’ communications before a customer reaches a limit and starts to disengage with your brand.   

Next week we will look a lot deeper at the sales-based models of Churn, Propensity to Buy and Customer Lifetime Value to understand how they work and most importantly, how they benefit your marketing efforts with examples of how they will increase sales and ROI.  

Read part 3 here.

Until next week… Any questions: or contact us on

Sign up to our newsletter

Want to make life easier by staying on top of market trends?