Key frameworks igaming needs to avoid stunted AI builds

AI in igaming has been an igaming talking point for years, mostly around customer experience personalisation, content writing and proofreading. Those projects are about bringing value to the business but they have very limited scope of what’s possible, nor are they very innovative in the era of GenAI and Agentic AI. I’d say, they’re even eight years late, compared to other industries.

Igaming has been facing hyper-growth for years, and the poor innovative AI solutions on the market are probably linked to this fast-paced growth. No one has time to explore and build innovations when the whole business is caught up in the grind of reaching new markets and launching new games and products.

On top of that, the latest market conditions from 2018, with the increasing speed in the US until 2023 and LatAm’s growth, were bringing enough revenues and profits to the whole value chain not to have a strong interest in cost efficiency. I am not saying here that industry leaders did not pay attention to their cost structure, but that they did not have either the time or the motivation to focus on it until the market contracted in 2024.

On top of those industry-related reasons, we have also to admit that those technologies were (and still are) a bit blurry. Month after month, we can see a bit better through how it works and what it can bring to any business. And it’s hard to run exploratory projects on blurry topics without any top management sponsorship.

Having worked on AI & data-based initiatives, I’ve identified some key success factors to help move toward such shifts and a framework for leaders to get the most out of AI.

Structuring your ideas through a framework

Following a framework offers numerous advantages for innovative projects. Start-up studios & Corporate Venture Builders have all built their own in order to industrialise and mitigate the risks of launching new innovations.

The framework listed below helps to give structure and pre-built components that guide our thoughts and it significantly reduces the project's time for faster results. And when it’s about innovation, testing concepts, and ideas and failing as fast as possible is very important.

1. Define the role you truly want (and can achieve)

Taking the time to question yourself on the role of your business is critical. Humility must be part of those reflections because aiming too high will only result in failures. Not only the role you want your business to play but also how to get there, setting up the right milestones.

Ask yourself :

• How can the business reinvent itself to lead the industry? 
• Why & how this shift will bring value to the business?
• What should be impacted?
• What wouldn’t be impacted?
• How comfortable are you not having results before 6 months? A year?

2. Start with pain, not by desire

The entry point of any innovation is the problem it solves, not an idea of what solution should be delivered. There is this very well-known quote from Henry Ford about the first car, saying “If I had asked people what they wanted, they would have said faster horses”.

AI should be about making work easier. Hence, It’s all about pain points. It can be pain points impacting your customer (such as payment methods, after-sales services, VIP, and technology issues when registering) as well as internal pain points, such as invoicing, getting paid and IT infrastructure slowness.

Pain points must be audited across the whole company. Identifying them is sometimes as hard as solving them. A brainstorming session could help, but this is where ‘lean management’ and lean Six Sigma expertise will be game-changing. Business pain points are not detected by thinking in a meeting room, but by measuring what’s exactly happening at each step of any of your processes. This will give you a great overview of processes and what could be optimised.

3. Document your existing assets 

Before starting to deliver, collecting the key assets that your AI could use is a major step. You’ll need to know what data (and how it is stored), the tech stack, and the APIs that you’re using (and its source).

Assessing your data quality at this point is critical. The way it is collected, stored, updated & secured, makes your use cases safer for the whole company.

Knowing how your employees are already using AI tools will also translate into opportunities for your company. Pay attention not to fall into the trap of thinking that if your teams are using GPT to write content, you should build your own GPT. (Bloomberg spent $10M+ for its own LLM, finding out that it did not compete with ChatGPT)

4. Seek after enablers

Enablers will make it easier (or at least won’t enhance the difficulty) to reach AI goals. Here is a list of enablers to pay attention to:

• Leadership: is someone on the board or ExCom a sponsor of this?
• Talent: Does the company have someone particularly talented, experienced or trained on such topics?
• Culture: what is the true culture of your teams? Not the one decided with post-it notes in the last off-site retreat, but the one that each of your core team actually works by.

5. Prioritise your actions

From now on, you have a good view on what are the pain points of your teams, per department, and what assets and enablers you have at your disposal to solve those issues.

That’s the best way to kick off the first initiative. To make it concrete and tangible, it’s critical to build OKRs (objectives & key results) for such matters. OKRs would be at two levels:

• The project itself: what are its objectives (do not build more than two objectives for a project) and would you measure your achievements?
• The project’s delivery: measuring progress quarter after quarter and making sure the delivery remains on track.

Without any OKR, it will be very hard to define if your AI Project brings value to your business. It’s also an efficient way of keeping our ambitions driven by results and metrics, without being too frustrating.

Running such innovative projects requires another methodology, and I’d recommend having a look at Lean Start-up. In short, Lean Start-up relies strongly on data, experimentation and learning; it closely resembles the scientific method: 

Build - Measure - Learn

The goal is to rapidly test a product or service, validate it in the market, and measure progress regularly to gather feedback in short cycles while minimising financial investment.

Direction to follow

Having a framework is never sufficient alone to reach your goals, nevertheless, it brings a direction to follow in order not to forget critical points.

To guarantee to get the most out of your ambitions and of what AI can bring to your business it’s also important to keep in mind that having an external point of view on your daily matters, pain points and innovative projects can be very differentiating.