If there is one thing that connects everyone in the iGaming industry, it is a skillset and experience we all share but don’t realize it.
We all work in data.
If there is something I do believe it is that iGaming itself has been slow to take data analytics to the next level. My belief is the industry has been so successful, it’s not felt the pressure to go in the trenches to collect and inspect data to build products and optimize revenue for a business.
When it comes to affiliate data, I’ve felt that affiliates for a long time have been lacking the tools they need to do their job to the best.
I believe not enough affiliates are focusing on the KPIs that matter the most. KPIs like EPC and instead we are chasing flat fees, listing fees, CPA and high revenue share deals.
On the other side of the fence, affiliate managers are no different chasing FTDs as if it is the only metric that matters most to a brand.
Before I dive into some practical advice for affiliates, let me share where I think we are today.
- All of the top 25 EGR Power Affiliates use StatsDrone, Routy, Voonix or an in-house affiliate stats tracking system
- Most iGaming affiliates don’t use postbacks or dynamic variables
- Many affiliates still send traffic to closed brands or have broken tracking links
Here is my overview of what an iGaming affiliate needs to have on their radar and part of their business strategy to be successful today.
- Demand for owning their own data
- PAM for affiliates
- Page level tracking
- Shaving and revenue optimization
- AI & ML
Owning Your Data
I’m going to write something here that first had me concerned about my own company. Would affiliates want to stop using hosted data tools like StatsDrone, Routy, Voonix and Affcollect and build their own in-house stats app?
We know some of the biggest affiliates do manage this in-house and their top reason is for data privacy.
When I first met Mark Flores Martin, founder of XGENIA, he told me he believed more companies were going to invest more time to own their data stacks. He said this before AI got really good with programming and it is playing out as we speak.
I still believe this is going to happen and oddly enough, it is something we want to help affiliates with. In experimenting with more open source data tools, I’ve come to realize this is going to become more of the future.
PAM for Affiliates
I’ve been writing about PAM for affiliates for well over a year. It is the idea that an affiliate could own the player when they have first party data, consent and can use a postback and/or a dynamic variable to track their players.
AI is making this easier to build.
I’ve built multiple models of this using Claude Code over the past 10 months and at the time of writing, I worked on another version of this using Grok Bot.
I think PAM is going to be very important as more affiliates work towards truly owning their own audience. This is now more than just capturing emails, phone numbers, Telegram connections and sites with SSO. I still believe affiliates have an unfair advantage over operators and that will change when more affiliates start building their own PAM.
Page Level Tracking
I already know that most affiliates don’t use postbacks or dynamic variables so by default, they are not tracking revenue by page.
We are just starting to help affiliates with this with StatsDrone but I gotta give credit where credit is due, this is something that Routy.app have been doing for quite some time. And I have to say they’ve done a good job of it too.
I believe affiliates should know which pages are generating their FTDs. When you put your data into a single bucket, you can’t decide which pages are your most valuable. You can’t even tell if your player values are different amongst content.
We already know this to be true but I think more affiliates need to see this to believe it.
I do think this is something operators could learn from too if they were to have some access to this data.
Shaving and Revenue Optimization
Affiliate shaving is as old as affiliate marketing. Simply put, affiliates have to blindly trust the data that operators send them is true.
How can affiliates detect shaving? Here’s my short list of things you could look out for. Keep in mind it can take a big of effort to analyze the data.
- Track short term removal of players (new FTD appears today, is gone tomorrow)
- Track long term removal of players
- Players being moved from higher rev share campaigns to lower (or even 0% campaigns)
- Tracking links given where no commission has been assigned
- Winning players wiping out commissions towards the end of the month
I won’t go into depth on all 5 points but I’ll highlight 2 interesting ones.
On the comment about having tracking links where there is no commission assigned, there are a few affiliate software where this is a feature of the software. That is, you can join a program, think you have a default campaign and grab a tracking link and send traffic. Only to find out later no commission was assigned meaning you’ve sent free traffic. This has happened to me numerous times in the past and I hear this from affiliates all the time.
Regarding winning players appearing out of nowhere, this is something I know some affiliate programs are a little too consistent with. I know because I’ve analyzed the data on behalf of many groups of affiliates.
In theory, player clicks, signups, deposits and withdrawals should have an even distribution of data from the 1st day to the last day of the month. The only exceptions being the 31st happens 7 months of out 12 and of course February being shorter.
I’ve setup haveyoubeenshaved.com which was a site I setup a year ago. I’m surprised to see the interest in this.
AI & Machine Learning
My Ryzen Strix Halo machine with 128gb of RAM
So where does AI and ML fit into our affiliate space?
It is everywhere.
Machine learning can be applied to any affiliate data set, any operator’s data and this could be combined with multiple participants. There is a lot to unlock.
Big data is where I think the space will grow and where affiliates will be able to unlock more revenue. But in order for big data to be at its best, affiliates need to ensure their data is cleaned up. That is that the data they are getting isn’t broken or skewed and that they themselves aren’t leaking revenue.
Were are at a weird cross road where big data requires data sharing but on-prem solutions means owning your data and in theory not sharing it. There will be solutions to make both of these things true for people that they could share some data without compromising their privacy and get the value back in return.
I do believe more people need to look at how AI and ML fit into our affiliate space and this needs a collaborative effort to make anything happen.
3 key points about John Wright
- Host of the Revenue Optimization podcast
- AI & ML research in iGaming
- Affiliate shaving & revenue predictions
Short description
John Wright is an iGaming veteran with over 25 years in iGaming and affiliate marketing. He is currently working on machine learning and AI in affiliate marketing.
