Nicky Abela: Fast Track – AI and machine learning explained…

By | April 28, 2022

Artificial intelligence (AI) and machine learning (ML) are often referenced in discussions about the future of the igaming industry. We spoke to Nicky Abela, Producer at Fast Track, to get a better understanding of the applications for this technology in player engagement and how the company is helping operators prepare for a future of machine learning-powered CRM.

SBC: Can you explain the difference between AI and machine learning?

Nicky Abela: Artificial intelligence is all about making a computer capable of exhibiting human-like intelligent behaviour. This includes all attempts at making a computer think, feel, act, react, reason and make decisions like a human being. 

Machine learning can be seen as a special area of artificial intelligence that is focused around making a computer learn from data. The objective is to write software that is able to recognise patterns in data. In doing so, the software will be able to apply those patterns to data it has never encountered before and still be able to make decisions and predictions. This means that the software does not need to be explicitly programmed to handle every possible case that may be encountered, and still it will be able to handle all cases because it can apply the learnings from the historical data.

SBC: How does ML make it easier to get value from data?

NA: If we look specifically at player engagement, there are a huge number of variables that affect whether a player is going to respond positively (or how we would like them to) to a message. If we look only at player behaviour we already have many things to consider: 

  • Do they have a preference for sportsbooks or casinos?
  • What is their average bet size?
  • Do they have favourite sports bets/slots or do they try lots of different ones?
  • What is their average session length? 
  • How frequently do they make deposits?

The list goes on. Now imagine trying to find patterns in all of that data in order to predict which offer a player might be most responsive to. 

Machine learning has the power to create its own rules based on previous data. For example, you can use a ML model to look at a player’s past interactions with your offers and to select the offer they are most likely to respond positively to. It would take a huge amount of time if you had to manually figure out which factors play an important role in whether a player takes an offer, then analyse those data points for every player in your database! A well-trained machine learning model also gives you the power to know how to interact with a new player as soon as you start to receive data about them, rather than having to start your analysis all over again.

Nicky Abela – Fast Track

SBC: What steps should operators be taking to prepare for a future where they will be relying on AI / ML to make decisions about their player engagement?

NA: A strategy that involves machine learning isn’t something that you can just leap into. It’s important to make the necessary preparations and ensure that your entire team is ready for that step. At Fast Track, we guide our partners through this process using our “Pathway”. This can take an operator with absolutely no automation in their CRM to a place where they can confidently use machine learning to create 1:1 experiences for their players.

The pathway at a basic level: frees up time for CRM teams; enables them to scale with automation; encourages them to find bright spots in their data; allows them to scale with 1:1 experiences using the Singularity Model; and finally prepares them for a future of self-learning CRM.

The idea is that by the time you are able to use machine learning to create 1:1 experiences for your players, you have already started to shift your team’s time away from execution, developed habits of experimentation and analysis and understood the importance of automation in scaling personalised player engagement.

SBC: Tell us more about Fast Track’s Singularity Model, how it helps the operator and how it was developed?

NA: The Singularity Model is all about understanding as much as possible about each player so that we can maximise engagement through relevant communications and rewards. The model is able to analyse how players responded to previous engagements and use a scoring mechanism to make the optimal choice of the next piece(s) of content for each individual.

The great thing for operators is that their teams can focus on building collections of creative content. Meanwhile the Singularity model gathers data about how that content is performing across different player profiles and selects the right pieces to maximise engagement on an individual level. For example, the Singularity model might see that one player responds better to a playful subject line that teases the content of an email while another prefers more “straight to the point” copy. It is then able to select the best match for each individual player, from a bank of subject lines created by the CRM team.

We also put a lot of emphasis on making it easy for our partners to control their data and shape their models to best suit their needs. We have seen the limitations of using rigid predetermined models when brands and their player bases can vary considerably.

SBC: Why should operators prioritise getting ready to add ML to their CRM strategy?

NA: AI and ML will play a fundamental role in the future of player engagement. It’s important to ensure that your internal teams are prepared with the knowledge and tools they need, and that you carefully select the right external partners to support you in implementing new technologies. 

Machine learning models like Fast Track’s Singularity Model will lead the way for more data-led engagement strategies and free teams to focus on content creation and experimentation. The operators who do not adopt a new way of working, harnessing the power of technology available to them, are going to miss out on some really exciting opportunities.

___________________

 

Leave a Reply

Your email address will not be published.