First Published on 29th March, 2017
Ai Editorial: False positives have never been easy to identify. But with machine learning, 3D Secure 2.0 and data intelligence, airlines can be in better control of the situation, writes Ai’s Ritesh Gupta
The denial of a digital service that is routine or legitimate is annoying.
For instance, as a credit card user, if I intend to access my statement on bank’s website, and if even after filling of details and password (which most users tend to detest), the access to the same is denied then it disappoint us in a big way. Similar is the experience of a genuine digital transaction that either gets denied or takes longer than expected duration to finish due to seemingly stringent security measures.
Airlines need to invest in apt user experience strategy and an integral part of the same is to work out right acceptance gap for payments so that revenue generation doesn’t get impacted in a negative away. Conversion rate in commerce isn’t just about getting traffic to digital platforms, but what also matters is not turning away authorized orders and how companies deal with false positives.
Also, from a business perspective, any travel company can’t afford to have a poor profile i. e. being bracketed as a high-risk merchant. Any business with low transaction volumes needs to be wary, as even a single chargeback will be termed as a significant development to the issuing bank than it would for an airline with relatively higher transaction volumes.
So airlines need to be vigilant of the new developments, and make incisive moves to deal with this problem.
· Finding ways to distinguish real buyers from fraudsters: It is time digital enterprises evaluate the limitation of automated responses from rule-based filters. KPIs for authentication for CNP transactions include historic chargeback data, card acceptance rates, cart abandonment, issuer declines, merchant reversal declines, interchange cost etc. But areas like how online anti-fraud technology impacts false positives need to be assessed. Airlines can dig deeper into the efficacy of rules-based authentication, seeking control over their transactions, and also sort out the manual review problem as human analysis can’t be done away with.
Today negative attributes are being assessed in a dynamic manner and this is where machine learning can contribute when it comes to dealing with constantly evolving patterns. At the end of the day, the list of potential offenders can’t be too restrictive, and at the same time, airlines can be lenient with fraud prevention as well. Machine learning can cut down on unauthorized transactions while also reducing the risk of denying a genuine payment. Key here is the fact that the model keeps training through regular feedback. How? With information about traffic (based on behavioral, identity, and network patterns), and more of the same being garnered, more precise predictions can be. This results into fewer false positives. Also, this analysis can also pave way for customized checkout experience with lesser number of fields or even the fastest possible processing for authentic travellers.
· Counting on data intelligence: Rigidity due to pre-constructed rules can now be combated with data sharing and data intelligence. And the release of 3D Secure 2.0 specifications, too, needs to be followed for the same. One way to ensure the decline rate is relatively lower could be via availability of quality data. Giving issuers a chance to interject themselves into the checkout can improve upon the risk assessment. So what was being done sporadically can be done in a widespread manner i. e. enabling issuers to amend their authorization risk settings and tie the authorization to the authentication.
Enriched data flow provides stakeholders with a better ability to approve “good” transactions.
As indicated by CardinalCommerce, in the past, merchants and issuers relied on “relatively simple data points in making authentication decisions”. But by relying on emerging sources of data and data intelligence, there is a chance to cut down on fraud and false positives. Legitimacy can be verified via purchase history, device and behavioral information, relevant social media details etc. So airlines better lookout for extra authentication data elements that are available at the time of the transaction so that both the merchant and the issuer can make a more informed and precise decision as whether or not to complete or deny a card-not-present transaction.
(Related article: How Amtrak worked out 99.85% acceptance rate, significantly better than the airline industry 96.3% acceptance rate).
· Impact of 3D Secure 2.0: The new specification when officially launched is being tipped to contribute in a big way. 3D Secure 2.0, with plans for early adoption in mid-2017, supports new transaction attributes, and this is expected to curb the level of false positives.
The objective of new specifications is to aid verification on the basis of data elements pooled through the protocol with focus on a frictionless shopping experience for card users. Also to make the message interface and authentication flows agreeable to mobile platforms.
3-D Secure 2.0 will enable increased use of risk-based and dynamic authentication.
The risk-oriented outcome will be shaped up by uniform and extended set of data elements.
Owing to risk-based authentication method, issuers relying on static or partial passwords have to mull the improbability that consumers will memorize or easily use their passwords if these are only used for 3D Secure and are occasionally called for.
So be prepared for abandonment and failure rate. Accordingly issuers need to find ways for authentication.
Are you bold enough to survive in the brave new world? Assess your preparedness at 11th Airline & Travel Payments Summit (ATPS).
Date: 03 May 2017 - 05 May 2017
Location: Berlin, Germany
For information, click here
Follow Ai on Twitter: @Ai_Connects_Us