How to detect if your affiliate program is shaving

If an affiliate program deleted some players in your affiliate accounts, how would you know?

Shaving in affiliate marketing is like a taboo topic in affiliate marketing. It gets talked about amongst affiliates at conferences but doesn’t always end up in a forum post or as written content. 

What is shaving in affiliate marketing?

Quick shoutout to StatsDrone and FTDx for being our awesome sponsors!

Shaving in affiliate marketing is when a program deletes or edits data that benefits the operator at the expense of the affiliate. In iGaming, this could be deleting a single player from your reports. 

I know when running affiliate sites and talking to fellow affiliates, there would always be the dreaded end of the month player wins that would wipe out commissions. 

At the end of the day, affiliates are trusting operators that the data shown is not only accurate, but not manipulated. Affiliates typically didn’t have any means of detecting shaving but this might be changing. 

What is revenue leak?

The core idea behind StatsDrone was all about helping affiliates with revenue leak. 

The idea was to help understand potential shaving but now that our app is more evolved, I didn’t realize how much more useful it is as a tool to discover shaving and manipulation of affiliate data.

Previously, I only had a few ideas on how an affiliate stats app could detect shaving.

Monitoring EPC, earnings per click

First it would be simply highlighting the EPC values. Now EPC alone won’t tell you if a program is shaving or not but it will tell you if a program is giving you a better ROI than others.

What affiliates have less control over when sending players to operators is the conversion rate of getting FTDs and the retention rate. Essentially, a great product could technically attract better valued players as well.

The 2nd way of highlighting this was in my experience of previously using Stats Remote and feeling more data features were missing.

In particular, if affiliates would say they felt some programs had consistent magical winning players in the last few days of the month, they just didn’t trust it.

What I wanted to do was to make the data a little bit more visible in your daily analysis. We don’t have this feature yet to tell you that maybe the 29th, 30th and 31st days have lower revenue amounts or more winning players. Perhaps the chart I shown would show revenue by days of the month rather than traffic itself.

Stacking the data in a daily format, would make this easier and at some point we can do some data analysis to automate these insights. 

Data viz to the rescue

Below is a Tableau dashboard I created 2 years ago from taking a data viz course by Kevin Hartman. 

What you’re looking at is the daily commission of the program stacked over 2 years for 2020 and 2021 at an affiliate site I was managing for a single affiliate program. 

Just a quick shoutout to Lynda Salem who's designing Tableau dashboards for StatsDrone users. Reach out to her if you want to discuss a customized dashboard.

Here you can see winning players would take the commissions negative but the negative commissions were very short lived. 

It did make me wonder, if I had all the other affiliate programs stacked together in this chart format, would I discover other accounts that endured sustained wins wiping out commissions?

Would it be here that I might discover a monthly blip towards the end of the month of winning players? I think it is possible we might be able to help affiliates detect this. 

One time, an affiliate program deleted commissions on me

About 12 years ago, I had an affiliate program do something that I wouldn’t wish upon any affiliate. A program which I won’t name and isn’t very relevant today, did something that to this day still makes me weary of any affiliate program.

The program had a payment commission threshold of $500 which isn’t anything special. It took about 8 months for a small site to generate that. Once I reached the threshold and wasn’t paid by the program, I reached out asking when the payment would happen.

The response I got back the next day was appalling. The affiliate manager said you didn’t meet the payment threshold and had earned less and the commission was in Euro. Partners is a word that some of these affiliate programs shouldn’t be calling themselves. I had screenshots of the revenue which was in my email. The next day, my commission totals were slashed down to a fraction of what I had made. 

The program literally deleted commissions dating back the past 6 months!

To add insult to injury, I informed them I had before and after screenshots and that I was considering posting this on GPWA. Within an hour of that email, I got a new Skype contact trying to send me an .exe file which was suspicious as hell. 

Needless to say, that program never got another click again. To add insult to injury, the program won numerous prestigious awards. 

All of this experience made me realize, detecting this stuff and documenting it was very difficult. I knew this could be done with a stats app but I didn’t fully know how.

The old problem with most stats apps is when you ask to refresh the data, it is sometimes overriding the data and deleting it. With Stats Remote, you at least were building data files that were stored locally but they required a little bit of technical expertise in being able to view these files and analyze it. 

On top of this, all of this data analysis takes time especially when you have to manually work with data files. 

How affiliates are using StatsDrone to notice changes in affiliate data

I do think it is cool that some of our affiliates have said our app has caught odd data discrepancies. 

I mean I’m not happy this happened at all but I’m glad they said our tool helped them find this. 

One customer shared that they were able to recover $150,000 across two accounts by using a second StatsDrone account to resync data from the past 12 months. I won’t quote who the customer is because I would need permission from the affiliate owner themselves but here is what the account manager told me about their process. 

If any affiliate was manually adding data into a giant spreadsheet or database, human errors alone would become a big problem. 

When they reported it to the affiliate managers, one affiliate manager said the player went onto dynamic revenue share. The other said the terms of the insertion order had expired but was incorrectly retroactively applied to all of their players. In both cases after a bit of discussion and providing evidence, both the programs paid out the commissions. 

At the time, our app does this feature where when collecting your affiliate stats, it is requesting the previous 10 days of data and it does this on a daily basis. The StatsDrone app does give you the ability to resync data going back to any date you wish. The problem is when you do a full resync, it overwrites all of your data.

In theory this shouldn’t be a problem because your old data shouldn’t be changing. However, when it does change is becoming important to know. 

Getting accurate data from your iGaming affiliate programs

I might need to write a whole other section about the problem of getting accurate data from affiliate programs. This is something we pride ourselves on at StatsDrone and I can’t recall a situation where we haven’t been able to fix data accuracy issues for any customer.

Unfortunately, affiliate marketing isn’t standardized in many ways and this is one of the reasons why many apps aren’t flooding the marketing beyond what I consider the big 3 of StatsDrone, Routy and VOONIX.NET. Our job is difficult.

In saying that, if you test out StatsDrone and find anything not accurate, let us know and we’ll fix it. We are library based for our project so there is a good chance if we fix an accuracy issue, we’ll have fixed it for our next future customer. 

Need more help?

If you want help in using StatsDrone to detect if any shaving might be happening with your accounts, contact us or send a DM to Darrell Helyar, Joe Hatch or myself John Wright and watch out for the other 2 John Wright guys in iGaming!

So what is next on our product roadmap to help with shaving?

• Improved data visualization

• Automated notifications of data points worth reviewing

• Automation of syncs to store in a separate database for comparison

The last point of data comparison can be achieved by exporting all of your data and doing another sync. If you’re good with data, this should be somewhat easy to do. If you need help, we can help you with a few ways of achieving the same result.