How does fraud analytics work in the cloud?

Fraud analytics involves the use of data and statistical techniques to identify suspicious activity that warrants further investigation. There are a number of different approaches and analytical methods that can be used, but the premise remains the same across all fraud analytics solutions: Accurately find more suspicious activity faster than the old-fashioned way. 

Using a cloud-hosted, or SaaS, approach carries three key benefits:

1. Speed. Installing a traditional on-premise solution can take months. It involves procuring hardware, doing installation and configuration work, and allocating resources from already over-burdened IT departments. Because SaaS deployments eliminate most of this work, they can usually be deployed in less than half the time of traditional on-premise projects.
2. Affordability. With the economies of scale gained in cloud operations, most SaaS applications can be offered at a lower price point than their traditional counterparts. SaaS can often be billed on a pay-per-use basis and recognized as an operational expense. For some organizations, this is preferable to the large upfront investment of a capital expense.
3. Security. Cloud computing involves sending data over the Internet, but that does not mean the data is unprotected. When properly configured, SaaS applications are very secure. Most reputable cloud hosting providers likely spend far more on security technology than your company's IT department. 

Next page: Three things to keep in mind when looking for a cloud fraud analytics provider

Shopping for fraud analytics in the cloud

If this approach sounds like a good fit for your organization, consider three things when looking for a cloud fraud analytics solution provider:

• Another asset that provides speed and efficiency in SaaS projects is the use of consistent, repeatable pre-built components. Evaluate which components are available "off-the-shelf" and which ones require extensive configuration or customization. Look for these common pre-built elements: data model, data quality and entity resolution routines, fraud detection scenarios and models, intuitive user interface, and administrative and management reports.  If these things are not available, they may add time and cost to your project.

• Don't assume that a vendor provides adequate security. Look for security certifications. A SOC 3 report will provide information about how a provider complies with industry standards for security, availability, processing integrity, confidentiality, and privacy. A TRUSTe EU Safe Harbor certification indicates that a provider complies with the stringent European data protection regulations. Talk to your own information security department who can help evaluate whether a provider meets your security requirements. Some organizations choose to do their own audit, either remotely via a detailed questionnaire or in-person via an on-site inspection. As part of your security review, don't forget to check on the Service Level Agreements around guaranteed uptime, disaster recovery, and support.

• Finally, be sure that any provider you choose understands your business. Look for domain expertise and consultants that have worked in fraud investigation roles in your industry. Fraud detection in insurance is complicated. Jurisdictions have varying regulations that can impact fraud types. Each insurer has a different book of business that may influence the type of fraud exposure observed. Make sure that the partner you choose has the knowledge to tune your solution in a way that best fits your data and your business. This will help insure that you get the best results.

Cloud hosting options make fraud analytics solutions more accessible to insurers of all sizes. More and more companies are choosing this approach. For those companies researching fraud analytics technology, a cloud hosting option may be the best way to get started.

James Ruotolo is an insurance fraud technologist, thought leader and the principal for insurance fraud solutions at SAS. Connect with him on Twitter.