Affiliate marketing is a massive industry that’s showing no signs of slowing after over a decade. U.S. retailers spent roughly $4.7 billion on affiliate marketing in 2016, and that number is predicted to climb to $6.8 billion by 2020. The popularity of this channel has attracted widespread attention, but unfortunately, not all of it is wanted: advertising fraud is a massive and costly issue, with advertisers shelling out about $19 billion on advertising that was fraudulent in 2018.
In case you’re wondering if partner marketing is worth the risk, where your money is going, and how to tell for sure if your campaigns are wasting or making money, there are several concrete steps advertisers can take to reduce any risk associated with partner marketing. In this blog post, we detail key metrics to watch for that indicate suspicious traffic, and how to create a foolproof fraud protection system in your partner marketing program.
5 Metrics to Help Spot Fraud
Spotting advertising fraud comes down to knowing your usual data patterns and partners well. What’s unusual for one company could be entirely reasonable for another, depending on company size, industry, traffic volume, average order size, seasonality, partnership types, and more.
Various devices are leveraged differently for fraud, as well. For example, desktop-based ad fraud campaigns can involve opening a browser on the victim’s computer and clicking on an advertisement, or stuffing cookies onto a browser to take credit for future purchases. In contrast, mobile app fraud is one of the fastest-growing forms of advertising fraud, mainly because ads in mobile apps are rarely blocked in ad-blocking software, and due to the nature of Android’s open-ended architecture. Mobile app fraud can involve click bots performing fraudulent in-app actions such as clicking on ads and click farms where low-paid workers are trained to efficiently install apps or navigate through them while clicking on ads.
Regardless of whether you’re advertising on desktop, mobile apps, or both, digging into patterns, trends, and anything out of the ordinary helps you spot fraud before it costs you too much money and energy to fight it. Use the five metrics below to get started.
1. Time to action
Time to action is the amount of time it takes between an initial click and a conversion. It is also called mean time to conversion, time to install, and a few other variations. This metric varies from shopper to shopper, so the key here is in the outliers. If you see publishers with a large amount of volume paired with a short conversion window, fraudsters could be stuffing cookies or injecting clicks to take credit for future sales. (TUNE’s platform comes with a time-to-action report built in — email email@example.com to learn more.)
2. Partner conversion rate
For every person who makes a purchase or conversion, there are two or more who clicked but didn’t end up converting. That’s your conversion rate, and each partner has a different one based on the quality and engagement of their traffic. If you see partners with a high volume of clicks but low conversion, it could be a sign of click fraud, especially if the publisher is measured on a cost-per-click basis. Ideally, you want to set up performance automation features that allow you to automatically detect and respond to conversion rates above or below a custom threshold.
3. Traffic by URL rank
Your partners with the highest-ranking URLs likely drive the bulk of your traffic. There can be exceptions, but if you see a partner with a low-ranking URL suddenly generating a ton of traffic, clicks, or sales, it indicates something fraudulent could be to blame.
4. Quantity of redirects
This is something to keep an eye on when reviewing the URLs that are referring traffic. If you notice suspicious URLs or several redirects, it could mean an affiliate is trying to hide the sources of the traffic from you. Similarly, if you notice multiple transactions come from the same IP address, one person could be placing multiple orders with stolen credit cards.
5. Average order size
Most businesses have predictable order sizes that rarely fluctuate except possibly with holiday, season, or special promotions. If you notice something abnormal on an otherwise regular day, it may be worth ensuring that it’s not related to advertising fraud. By setting up profit rules, you can detect and respond to profit margins below a custom threshold.
How to Identify and Prevent Ad Fraud
Get familiar with your flavor of normal
One of the keys to spotting fraud is to know what you’re looking for in the first place. That means having a sense of what normal looks like for you. Get familiar with what typical conversion ratios, traffic volume, and profit volumes look like for your company and partners. Knowing the trends, including how they vary based on holiday, time of year, and promotions, helps you spot any outliers quickly.
Get to know your partners
That goes for the good guys and the fraudsters. If possible, aim to manually approve new partners or vet them with the help of reputable sources. After a while, you should have a sense of which partners drive massive volume and which don’t. If the latter suddenly start driving a lot of traffic or clicks, you have reason to be suspicious.
You may want to keep a list of partners who have been declined from your program or blocked altogether. If they try to sneak in and re-apply, you’ll be able to spot them right away instead of wasting your time going through the vetting process all over again.
Combat fraud with powerful technology
If a manual review doesn’t sound fun to you, a fraud prevention solution can help automate the process, make it easier, and alert you of things you have missed. Check out our recent article on the performance automation features we recommend, including things like conversion rate optimisation to keep an eye on suspicious conversion rates, profit optimisation to ensure campaigns only run profitable offers, ways to automatically warn or block offers and affiliates based on pre-selected settings, and event logs that provide real-time visibility into every automated action taken.
By making a routine out of getting to know your data — and leveraging the right technology to help — you can ensure fraud is a rare occurrence for your partner marketing program, and leverage it instead for all the good it can bring your business: new audiences, more revenue, and a predictable profit stream.
For more information on performance automation, including use cases and feature details, see ”The TUNE Guide to Performance Marketing Automation.” If you have questions or would like to learn more about the TUNE platform and automation features, please email firstname.lastname@example.org.