As the appeal of eCommerce grows in the eyes and wallets of new populations of shoppers, selling online is becoming more lucrative. But for merchants, it’s more demanding than ever to scale, expand, and keep up with the dynamics of both shoppers and fraudsters. Future-proofing your business raises many new challenges like embracing omnichannel flows, supporting cross-border sales, and rolling out new products. In this blog post, we’ll unpack the pros and cons of building vs buying.
A lot to consider…
A first point to consider when thinking about an in-house fraud prevention platform is scale. New platforms are chosen with existing workloads in mind, but will your fraud management system align with future growth strategy? Will it support holiday spikes, flash sales, and unforeseen events, like Covid-19, that can overwhelm and inundate your digital operations? Next – will it outsmart new and evolving fraud MOs and keep up with the neverending arms race between retail and fraud? These questions have direct ramifications on the time, effort, manpower, and expertise it will take to develop and maintain such a robust system.
Analyzing, developing, and implementing a proprietary fraud prevention platform also requires significant investment. This includes upfront costs as well as recurring operational expenses to maintain the system and ensure it’s not stagnant. Our data science team recently analyzed which machine learning algorithms we’ve tested and found that in an ever-changing world, algorithms that worked well two years ago, are often outdated for fraud purposes today. High costs, underperformance, and obsolescence play a crucial role when considering building a fraud prevention solution.
Fraud management has a direct impact on chargeback rates and approval rates, and these don’t only impact revenue, they also directly affect merchants’ reputation with banks, credit card companies, and customers. When the stakes are high, entrusting a 3rd party solution to manage fraud could seem to be a challenge. Building a proprietary system seemingly grants merchants a sense of control. And, theoretically, merchants can tweak their system, embed their expertise into it, and accumulate knowledge in-house. But resources and expertise always seem to be scarce. Some merchants may have chosen to build on fraud operations that were initiated before 3rd party systems were even available and they may believe they are doing well enough. All too often, when they dig in, they will find that they are declining legitimate orders at a high rate and their platforms are underperforming – negatively impacting their sales.
Fraud has become a lucrative industry. In the past, merchants needed to seed out orders made by lone wolves who stole credit card details and went on to use them to purchase goods online. The first fraud management platforms were built to stop this kind of bad actor. Today’s fraud teams are tasked with identifying advanced fraud MOs that are the result of tremendous, ongoing technological investment. The use of bots has brought to the table both sophistication and scale. Fraud management today requires the latest technologies, data enrichment, and analysis tools, and mandates hiring hundreds of the best data scientists, engineers and operations analysts. To keep an edge, you need to constantly research what are the newest data sources to leverage, engineer new features as fraud changes, etc. For example, account credentials have become easy to obtain for fraudsters. A good account history is no longer as indicative as it used to be of a legitimate order. New models, new data sources, and approaches are required to catch account takeovers.
For many merchants, this investment is just too big. Staying ahead of new fraud trends is too resource-intensive. And, keeping up with fraud, is not part of their core business.
3rd party solutions also provide a network effect. Merchants who build their own platforms only benefit from their own order database. Partnering with a fraud provider lets you benefit from the identification of fraud at other merchants. And with fraud migrating from merchant to merchant and geography to geography, the network effect enables 3rd party systems to clamp down on fraud that a single merchant may never have seen previously.
Buying a machine learning fraud management platform with a chargeback guarantee lets merchants focus on their core business – selling goods online. A chargeback guarantee means that the vendor’s interest is aligned with the merchant’s. The vendor is only paid when the vendor approves your orders and merchants don’t pay for chargebacks on orders the vendor approves. Moreover, there’s no need to invest in fraud research – instead, merchants can focus on their business, providing a great customer experience, and expanding globally to new markets and geographies.
Fraud is more complex than ever
Machine learning is the standard for fraud prevention platforms nowadays. Unlike rules-based systems that are reactive, machine learning is proactive. It is more adept at handling scale and better at evolving with the eCommerce landscape. It can also determine risk based on the most common patterns, to identify attacks before they occur.
The move to machine learning has made in-house fraud detection even more out of reach for any merchant without a significant R&D team. As merchants roll out new omnichannel flows, and as fraudsters try to find vulnerabilities in new shopping and payment options, remaining on top mandates tremendous investment.
To build or buy? That is the question!
Ultimately, building and maintaining a machine learning fraud prevention platform is a massive and complex undertaking that requires world-class expertise to maintain an advantage against fraudsters. Our advice: find a solution that is transparent, provides merchants with the insight they need, has a proven track record, and can support current strategy and future goals.
If you want to hear what went into building our own fraud prevention solution, we’ll be lifting the curtain on that for the first time during our upcoming merchant summit next month, stay tuned! You can also read our Fraud Management Solution Buyer’s Guide to ensure you have access to all the questions you need to be asking when procuring a fraud management platform.