Last Updated on March 5, 2020 by
Kount is an eCommerce fraud prevention solution with over 12 years of experience building and optimized machine learning and artificial intelligence models. Recently, they launched their new ‘Identity Trust Global Network’. The network connects the insights of order review across 6,500 different businesses and payments providers. It gives merchants within it a huge increase in the amount of data they have available to fight fraud within their own business.
We sat down with Rich Stuppy, Chief Customer Experience Officer at Kount for a wide-ranging discussion about the network.
1. Traditionally, Ecommerce fraud prevention has been order-data-centric. Increasingly, we’re seeing much more customer-data-centric approaches like the Identity Trust Global Network takeover. Why is a customer-data-centric approach better?
Kount’s Identity Trust Global Network looks at the identifiers in a transaction in order to create a risk or trust level in real time—but it does not score individual customer identities. This means that it is in between looking at order-data and customer-data, analyzing identifiers rather than identities. In each interaction, Kount’s AI real-time links payments data, location identifier data, and digital identifier data.
With this information, rather than order data alone, businesses can create personalized customer experiences. When there is low trust, interactions can be blocked. High trust means a VIP customer experience, and in between is step-up authentication or adaptive friction.
2. How does your customer-data-centric AI evaluate which customers require a higher-friction experience?
The combination of trust and fraud signals help businesses determine the level of risk in an interaction, and Kount’s platform is configurable to match our customer’s policies and tolerance for fraud.
The Identity Trust Global Network includes data from 32 billion annual transactions across more than 75 industries and over 6,500 customers, and it takes a network with the depth and breadth of data like Kount’s to understand the difference between customers that require a higher-friction experience and those who don’t.
3. What is your response to merchants who might argue that a customer-data-centric approach is not useful, since the drop-off rate caused by introducing friction makes it functionally equivalent to just declining these orders straight away?
Kount’s solution cuts false positives by up to 70%. Because Kount’s Identity Trust Global Network looks at identifiers rather than individual customer identities, friction is only introduced when trust levels are low enough to warrant it. Kount’s Customer Experience Engine allows business to control for a variety of situations, so low trust doesn’t have to mean a decline. Strategic, adaptive friction can mean only needing to verify one or two additional points, or it can be escalated depending on a business’s tolerance for risk.
4. What is an example profile of the kind of purchaser that can expect to have a higher-friction experience served to them by the Identity Trust Global Network?
Higher friction experiences are warranted when there is low identity trust, which generally occurs in the case of fraud. In any given interaction, Kount is analyzing 2.7B fraud and trust signals, so that no one or two or even a few factors would cause a good customer to experience friction.
From a website visit to login, checkout or account creation, Kount’s Identity Trust Global Network goes to work analyzing billions of identifiers to establish real-time links between identity elements in order to return identity trust decisions that provide the desired user experience ranging from low friction to blocking fraud.
5. What is an example of what a higher-friction experience looks like for different categories of high-risk purchasers?
Each Kount customer can choose to introduce step-up authentication or adaptive friction based on their individual appetite for risk. For example, consider an online gaming company that sells virtual goods. This company might not be as concerned with a chargeback but would want a frictionless customer experience. The gaming business may choose to let certain transactions go through with minimal friction, and others they may route through a longer in-game process to verify the interaction.
An online jewelry company, however, would want to avoid a chargeback and loss of goods, so higher friction may be verifying more order details directly with the customer. Kount’s solution allows a high level of customization based not only on the company but also individual products, so higher-friction is introduced only when necessary.
6. How do you ensure your system is only serving higher-friction experiences to purchasers that really need it? In other words, what is the feedback loop that allows the system to improve?
The Identity Trust Global Network is made up of 32B annual interactions, and each individual interaction links 2.7B trust and fraud signals. That level of data built over 13+ years, 75+ industries, and 50+ payment providers and partners and the decisions of thousands of fraud analysts feeds into Kount’s AI. Kount’s AI uses both supervised and unsupervised machine learning to learn from historical outcomes and detect emerging fraud, so that the entire system is continuously improving.
7. You wrote in your press release that Identity Trust Global Network creates a “65% increase in operational efficiencies”. What are specific examples of what that looks like?
Operational efficiencies include a number of business outcomes such as an improved customer experience, higher margins, higher revenue, fewer disputes, etc. Kount’s Identity Trust Global Network helps companies realize higher sales conversion rates and improved brand reputation by created seamless experiences.
8. Your press release also states that Identity Trust Global Network creates an “83% reduction in manual reviews”. What is the network doing that allows merchants to reduce their previous manual review rates?
With less time spent reviewing transactions, fraud analysts can spend more time in Datamart, Kount’s Business Intelligence reporting suite. Datamart provides merchants with the ability to create reports and dashboards on-the-fly from Kount’s enterprise data warehouse.
The tool provides deep intelligence into the types of fraud specific to their business and adds additional value by informing new and proactive initiatives, pinpointing areas for improvement to enhance operational efficiencies including customer experience, margins, and revenue.
9. How does the Identity Trust Global Network work for a new merchant, that has no data on what their specific fraud profile looks like?
When a business joins the Identity Trust Global Network, they are gaining access to Kount’s rich depth and breadth of data, which is unmatched. That is one of the key benefits to working with a network that can link across billions of transactions in real time, as consumer behavior differs from vendor to vendor.
If one company sells furniture, they may only see certain consumers at most a couple of times a year. A coffee shop may see the same consumer multiple times a day. That’s why the network effect is crucial—those fraud and trust signals with Kount’s adaptive AI make the solution effective for a wide variety of merchants, including new ones with little to no data of their own.
10. To conclude, we’ve now moved from static rules, to AI decisions, to adaptive friction. Is there anywhere left to go to decrease both false declines and customer friction?
Throughout the evolution of digital fraud prevention, we’ve seen a clear shift from simply blocking bad transactions to enabling personalized user experiences. That’s where fraud prevention will continue to move to decrease both false positives and customer friction.
Fraud and trust signals no longer only sit at the point of a payment transaction, but instead, at any point throughout the customer journey. From account creation, to login, to payments, and every step in between, there is opportunity to deliver a customized experience and ultimately address both of these business outcomes.
