The Challenge

“It’s never quite good enough.”

That’s how it can feel for organizations trying to implement a fraud detection and prevention strategies these days. As settlement speed increases and payments travel across what seems like a new rail every week, the pressure on those charged with protecting payment transactions and the technology behind them can be overwhelming. In recent years, the response to this reality has been to implement the latest in enterprise fraud tools in the hopes that the promise of advanced analytics and scalable architecture can allow those responsible for fraud prevention in your organization to rest easy.

Unfortunately, those fraud systems are far too often put in place and then trusted to consistently deliver results without any monitoring or adjustment. This may be from too much faith in the “learning” of machine learning or due to an overly onerous process for monitoring the health of the analytics. Regardless, when this happens the initial gains in fraud detection become overshadowed by a flood of false positives resulting in an overbearing workload for your fraud investigation team.

The simple truth is that for every transaction that a fraud detection solution flags, an investigation needs to take place and that investigation consumes resources.

So, when your fraud detection solution is not able to perform to expectations and evolve with new fraud trends, your organization is faced with a choice of minimizing risk as much as possible by devoting higher levels of resources toward fraud investigations or failing to prevent fraudulent activity. This is a particularly sensitive inflection point when dealing with commercial payments where the frequency of fraud is lower, but the impact of the fraud that does occur is much greater. It becomes a question of cost vs. risk, a constant balancing act that is becoming ever more challenging as more payments are processed through digital channels.

The Solution

In order to maximize your fraud detection investments, your organization must find a solution which not only understands the transactions and channels you are trying to protect, but is also able to provide the tools and expertise to evolve the solution as the fraud landscape and your organization’s specific needs change. Bottomline’s Secure Payments solution offers the ability to do just that.

The solution is built leveraging Bottomline’s unique combination of commercial payments and fraud detection expertise.

This unparalleled understanding of commercial payments fraud allows Bottomline to deliver fraud analytics specifically designed to meet the needs of the commercial payment space with its lower volume of higher value payments.

With this in-depth knowledge, the Secure Payments analytics deliver immediate value out-of-the-box with a vast library of fraud scenarios and dynamic profiling to stop fraud without inundating your organization with false positives. These proven analytics are supplemented with intelligent machine learning that both identifies new patterns and suspicious anomalies with unsupervised learning and also tunes the base analytics with supervised learning based on investigator decisions. This unique fraud detection capability has not only been chosen to protect the payments of hundreds of thousands of organizations globally, but is also trusted by Bottomline to protect the trillions of dollars of commercial payments that we process annually.

While the standard Secure Payments analytic capabilities provide superior protection, Bottomline puts a premium on helping organizations adapt with the payments landscape. This involves strategic roadmap investments aimed at identifying new ways to protect transactions and improving investigator efficiency while also providing the tools to support individual needs of each of our customers.

As part of this commitment, Bottomline has developed these additional capabilities:

fast feature mediatext


Lane assist in your car is a great safety feature, but it’s not an excuse to take your hands off the wheel. The same can be said for your fraud analytics. While many fraud detection solutions take a “set it and forget it” approach to analytics, Bottomline provides expert fraud advisors to work directly with customers to tune their Secure Payments solution to match their risk appetite.

This approach allows customers to know that the solution has been maximized based on their unique needs and industry best practices so that they only see the most meaningful alerts while feeling confident they are not missing any suspicious activity.


For even greater impact, organizations can combine BTune with F.A.S.T., Bottomline’s one-of-a-kind tool for previewing how changes in risk thresholds will impact investigator workload.

The solution allows organizations to see how different analytics tuning approaches will impact their alert volumes based on their past transactions.

Fraud advisors work with you to run multiple scenarios against historical data and compare how changes to the analytics would have impacted both alert volume and quality over the selected timeframe. This allows you to compare risk models and then choose the best one for your business to move forward with rather than having to wait for the results of a champion / challenger risk model evaluation approach. When used alongside BTune, F.A.S.T. helps customers to reduce costs and risk without having to compromise one for the other.

Real Time Analysis Illustration 1260 706 2
Machine learning mediatext


Every organization has made a foray into machine learning. Whether it is through investment in data scientists or machine learning tools, you want that investment to benefit your business as much as possible. You can leverage that investment and use it to enhance your fraud detection capabilities by using Bottomline’s advanced machine learning engine. Your organization can opt to utilize Bottomline’s patented algorithms as a starting point or take advantage of Bottomline’s Machine Learning Studio for a more hands-on approach. The Machine Learning Studio capability comes in a Do-It-Yourself (DIY) package where your data scientists use Bottomline’s tools to create their own model or in a Bring-Your-Own (BYO) package enabling you to import models from other systems.

In both cases, you gain access to the Machine Learning Studio and are able to add another layer of fraud detection by incorporating unsupervised machine learning models of your design into the Secure Payments analytics engine.

Why Bottomline

You can choose to take advantage of any or all of these capabilities to enhance your Secure Payments experience. In all cases you can be assured that the steps you take toward improving and monitoring your fraud detection capabilities are industry proven. Bottomline customers of all sizes (and Bottomline itself) have seen the advantages of leveraging commercial payment expertise in enhancing fraud analytics first hand.

Examples of Bottomline’s successful approach are all over the globe. From the private bank in Europe that utilized BTune to refine their analytics approach and saw alert volume improve to the point where most alerts they now receive require a call to the customer to verify payment legitimacy. To the large US bank who was trying to budget for new fraud team headcount in anticipation of a projected rise in ACH volume, but avoided it thanks to the F.A.S.T. offering that allowed them to see exactly what the impact of that increased volume would be and how they could tune their fraud detection to manage that increase without impacting alert quality before the volume increase actually happened. In these cases, and more, Bottomline’s Secure Payments analytics services have delivered market leading protection without impacting operational costs.

As you evaluate fraud detection vendors and look for ways to maximize your fraud analytics, remember that the technology is only part of the equation. For the best, ongoing protection, you need to have the right tools to constantly evaluate and adjust your analytics as the payments landscape changes and you need a partner who understands the intricacy of the payment rails you want to protect.

footer curve