What You Should Consider When Evaluating a Fraud Prevention Solution

What You Should Consider When Evaluating a Fraud Prevention Solution

This is the fourth blog post of a five part series.

Part one: The Fraud Prevention Space’s Complex Optimization Problem
Part two: What You Can Gain by Partnering With a Fraud Prevention Vendor
Part three: An Overview of Fraud Management Solution Types: Pros & Cons

In this blog post, we are going to focus on the questions you should ask yourself when choosing a new fraud prevention solution. We know that this is not an easy task and that it can be confusing since there are many players in this space and some of them might look more or less the same. We’ve listed the main factors you should take into consideration as part of an evaluation process below. Remember, this partnership will generally last for 3-5 years on average, and can have a huge impact on your business in a positive or negative manner. Be sure to ask all the questions and conduct a thorough evaluation process. The following criteria should help:

Chargeback guarantee vs. scoring

Probably the first decision you need to make is whether you’d like to partner with a scoring solution or a chargeback guarantee one. For more information, see An Overview of Fraud Management Solution Types: Pros & Cons. In essence, the main question you should ask is: Are you interested in getting only a recommendation and making the decision yourself? Remember that while this allows for more control over the process, it also requires greater expertise and leaves you with the liability for chargebacks. The other option is to receive a definitive decision that comes with a liability shift along with greater certainty and predictability, but also less control over the process. 

Vertical expertise

Different verticals have different fraud challenges and fraud patterns. The level of expertise a fraud prevention vendor has in your specific vertical is key to its ability to perform well. A fraud vendor that is capable of providing great performance in one vertical might not be really accurate and good in others. Here are a few examples to help explain: 

  • Physical goods vs. digital goods: Digital goods don’t have a shipping address and the order confirmation is usually instant and gets sent to the email or phone. This turns the email address and phone number into crucial data points. It also requires faster response time and relies on different machine learning models to analyze those kinds of transactions.
  • Travel industry: flights vs. experiences: For flights, the time before departure is a key consideration to determine the risk level of the transaction because, by nature, people would usually book their flight in advance. On the other hand, in the experiences space, which is also part of the travel industry, people tend to be much more casual and spontaneous, and, oftentimes, they will book tickets for an attraction at the destination itself, only hours or even minutes before attraction/event starts.
  • Gift cards: Valuable gift cards like, Amazon gift cards, are more or less equivalent to cash. This imposes a higher risk on a merchant and makes them considerably more valuable for fraudsters. 

Market expertise

The characteristics of markets frequently vary from one another. For example, in Asia, consumers tend to use mobile phones and applications more than consumers in Europe. Payment methods also tend to change. In Asia, several markets have their own preferred payment method from Wechat to Grab to Line. Expectations for delivery speed also differ greatly from market to market with a higher risk factor as delivery speeds increase.

As a result, a merchant that can provide great performance for US merchants, especially with US-based customers, might not be effective at all in Asia. 

The power of network 

There is a famous saying: “No one is as smart as an experienced person.” In fraud, this idea relates to solutions that use the network effect to be as accurate as possible. An experienced fraud solution will cross-reference every transaction you send them to their past transactions database to check if they’ve already seen this customer before in other transactions with other merchants. Then, the vendor will check for a positive indication (good transactions) as well as negative indications (transactions that returned as chargebacks). The larger and more global and diversified the network is, the stronger the network effect is. 

Large and well-known merchants also affect the power of network significantly. At the end of the day, all of us tend to buy more from the largest eCommerce merchants in every vertical and industry. This increases the chances that the fraud prevention vendor already encountered your customers in the past with another large merchant. It is important to mention that the number of transactions processed by a fraud vendor is important but the geo expertise and globality of its network is as important. For example, if a fraud vendor has a strong presence in the US and works with several US merchants it doesn’t necessarily mean its network will be very powerful and effective for Asia or Europe. 

3rd party data enrichers

As we mentioned above, data is what makes the machine learning world go around. The more relevant data a fraud solution is taking into consideration in reaching a decision the higher accuracy it will get, if used correctly. Some fraud prevention vendors achieve this by using external data sources to gather more data such as the email, the address, phone number, and social networks connections of a potential buyer in order to increase the accuracy of the model. 

R&D team size 

AI and machine learning may be the hottest technology buzzwords in the past few years. But developing and building accurate machine learning models is a complex task that requires not only talent but also serious investment and work. A data science team is required to think of new features, to test them, and see how they affect a model’s accuracy. Then the new features must be implemented into the model and the model must be retrained on a regular basis. 

As we mentioned earlier, a fraud prevention vendor should also have a vertical-specific model in order to be as accurate as possible. But it also takes time and a team of talented data science engineers and fraud experts to develop expertise and continue to refine it. A merchant should look for a vendor with a larger R&D team that has the capacity to focus on improving their models for fraud prevention and, in parallel, work on developing new products that will assist merchants in increasing their revenue, provide better customer service, increase conversion rates and create operational efficiency. Fraud is not the only challenge in the payments ecosystem — products that minimize friction with account protection, help with authorization optimization, and limit chargeback disputes can often be found under one roof.  

Remember, you are likely going to work with a new vendor for a 3-5 year period. You want to have a partner that is proven and invested in innovation to support your evolving needs and stay ahead of the game. 

End-to-end solutions 

In recent years, vendors have evolved to offer more than just fraud prevention at checkout. This evolution is intended to catch all potential fall-out points throughout the customer journey. A broader range of products can translate into higher revenues, better conversion rates, improved customer experience, and greater operational efficiency. When you evaluate the vendor’s offerings don’t just look at your current needs but also at your long-term ones. Your needs will change as you grow or change course and you will want to partner with a vendor who will be able to serve your evolving needs and priorities. 

Flexibility 

Most large merchants generally don’t look to switch from one fraud prevention vendor to another in one night. Some have their own internal fraud prevention solution they have built over time and they are only seeking help for certain segments that are more challenging for them such as 3DS. 

Despite these varying needs, some fraud prevention vendors will take only the full volume of a merchant and others are only open to taking a portion of the order volume and to assist with the most challenging and riskier segments. Prioritizing flexibility with a partner enables you to do things your way. To build trust and grow the partnership, adjusting and finding ways to optimize your business as you go. 

Professionalism 

This is something that is hard to quantify, but we can’t emphasize its importance enough. At the end of the day, you are forming a partnership and not just buying a service. The professionalism and responsiveness of the vendor’s team during the evaluation process is usually a good indication of what the future partnership will look like. Make sure you are going to have a dedicated account manager and 24/7 customer service support. Consider the value the account manager can bring to your team. How often are you going to be in touch with them? During the evaluation process, pay attention to the amount of time the company rep is listening to your specific needs. Those details usually reflect on the company’s culture, which is an important part in assessing any partnership. Be sure to ask to speak with a reference client and focus not only on the performance, but also on the partnership itself. 

Sustainable offers

Some chargeback guarantee vendors will commit to approval rates in the agreement. A commitment is a great perk for a merchant because it gives you predictability both on your approval rate (revenues) and on your cost of fraud (the fee the vendor is charging). 

The recommendation we have for you here is this: If it looks too good to be true it probably is. Beware. In a chargeback guarantee model, the fraud prevention vendor can lose money on the deal. If the commitment for the approval rate is too high and the fee is lower than the chargeback rate, the vendor has a problem because it will bleed money every month. You can tell yourself that this is not your problem — you might actually enjoy it. But this may not be the right approach. Remember, no one stays committed to a business relationship where they lose money. If the parameters of an agreement are not sustainable on the fraud prevention vendor’s side, the vendor will find ways to break the deal or just not respect the commitment. The problem for a merchant is that, at this point, you’re already integrated and it will take a lot of time and work to switch to another vendor.   

We really hope this was helpful and that you now have all the tools and insights you need to make an informed decision. For your convenience, we’ve added a table below that summarizes the criteria mentioned in this post.