Why Anti-Fraud Technology Is The Secret Sauce to Winning Ride-hailing Platforms

‘Grab Lawan Opik!’ (Grab Fights Fake Orders)

Fraud is hardly a new phenomenon, but with the rise of the digital economy, it has taken different and aggressive forms. Our mobile phones have increasingly become ‘remotes for our lives’, enabling us to handle many of our everyday needs, such as messaging, transport and shopping. Over the last decade, novel ways to exploit technology have appeared, from click farms to fake credit card transactions.

As a result, millions of people have been impacted and millions of dollars in revenue have gone lost.

This post will share how fraud has impacted the ride-hailing industry overall and how Grab has succeeded in protecting both drivers’ earnings and overall revenue as the leading platform in Southeast Asia.

Up to 20% of Rides in Southeast Asia Impacted

Since its entrance into Southeast Asia over five years ago, ride-hailing as a young industry has been heavily impacted by fraud. The big influx of capital into the industry has led to fraudsters, sometimes individuals, sometimes organised in gangs, trying to game incentive and sign-up schemes. As a result, a stolen ride-hailing driver profile today is worth up to US$30 on the black market, even more than stolen credit card information.

Some examples of the methods scammers use include the following:

• Setting up multiple phones and multiple accounts and pretending to execute many rides in cash, ultimately benefiting from bonus payouts upon hitting incentive marks.
• Use fake GPS tools and modded phones to simulate driving behaviour and completed rides to game the system.

In the early years of the ride-hailing industry, fraudsters have had a lot of success in gaming the system. In 2015, Uber noted that fraud accounted for less than 10 percent of bookings in China, but local media estimated the fraud rate to be multiple times higher at 30% to 40%. At the height of its competition with Didi at the time, a quick search for Uber on Chinese e-commerce website Taobao, netted 200,000 search results for fake accounts being sold.

Unsurprisingly, fraudsters have also targeted the highly competitive and growing ride-hailing market in Southeast Asia. Internal research by Grab indicates that fraud in the ride-hailing industry in Southeast Asia, if left unchecked, can represent up to 20% of all rides. In other words, without stern commitments from ride-hailing companies, for every dollar invested in ride-hailing, as much as 20 cents could be misspent on fraudulent activities.

The main victims of this phenomenon? Drivers. When accidentally matched with real drivers, fake passengers cancel their rides, leading to lower ‘completion rates’ for drivers and thereby impact driver bonus payouts and earnings. Many drivers have rightfully expressed their anger at unfair drivers being left unchecked.

Protecting Driver Earnings: Prevention, Detection and Fairplay

As a regional champion and market leader, Grab has dedicated the resources necessary to make combating fraud a top priority. Long-term success for the Grab platform means an obligation to protecting driver earnings and overall platform profitability. Grab’s investment in dedicated anti-fraud technology and programmes culminated in the launch of its ‘Lawan Opik’ programme or ‘Fight Fraud’ programme in Indonesia beginning 2018.

In the first six months since the launch of the campaign, the programme has seen remarkable results. Internal research metrics point to a dramatic decrease of 80% in the estimated number of fraud cases on the Grab platform in the last 12 months. Grab’s market research point to the platform being two times more resilient against fraud than any competitor in Southeast Asia. These figures benchmark Grab’s technology as being more than 12 months ahead of the competition.

Beyond internal benchmarks however, the programme has also led to some highly-visible results.

In the last six months, Grab was a driver in helping local police bust over 10 syndicates in Indonesia. From Jakarta to Makassar, Grab’s technology and innovative programmes brought up concrete evidence on several groups of criminals who were perpetrating fraud, enabling the police to round up entire operations.

Grab uses a mix of the following tools and programmes to combat fraud:

1. Preventive tools: Over the years, Grab has implemented many proprietary machine learning algorithms on the platform that identify circumstances in which fraud might happen. We have built robust models to predict which rides are ‘high-risk’ and prevent most of these risky or fraudulent rides from taking place pre-emptively. We also use proprietary graph and network models to uncover and take down syndicates.
2. Detection tools: Grab has also built out a wide array of fraud-detection tools in the Grab app. These range from the detection of fake GPS-apps to much more sophisticated forms of fraud.
3. Fairplay programme: Rolled out in late 2017, Grab’s whistleblower programme is an industry-first for the ride-hailing industry in Southeast Asia. It allows drivers to participate directly in the eradication of driver fraud by reporting suspicious fraud activities as they happen. Whistle-blowers are remunerated for successfully verified tips, creating a win-win situation. The Grab Fairplay programme has already generated more than 9,000 tips in the first quarter of its existence and was the direct source of one of Grab’s high-profile gang busts in 2018.

Finally, Grab has also rolled out a number of new measures to better protect driver incomes from the efforts of fraudsters. In just the last 2 months, Grab rolled out three small, but impactful new features for drivers:

• Cancellations within 30 seconds not counting towards ‘Completion Rate’: Fraudulent ‘passengers’ launching fake orders, end up canceling their rides within 30 seconds from being allocated to an honest driver. This ends up impacting driver’s ‘Completion Rate’ and their bonus payouts at the end of the week. This new feature will ensure these rides do not impact driver’s ride tallies any longer.
• Cancellations when passengers don’t show up: This new feature accounts for another scenario in which fraudsters forget to cancel their rides overall. Drivers who are forced to cancel ride-hailing jobs in case passengers don’t show, no longer see these rides affecting their ‘Cancellation Rate’ and ‘Completion Rate’, both metrics which help to determine the level of bonus payout a driver might receive.
• The option to automatically accept requests: While not directly related to fraud, this feature, when turned on, enables drivers to accept jobs automatically and maximise their time on the road. Drivers can ensure they don’t miss out on jobs when they are distracted or if they step away from their cars or bikes. Another feature to give drivers the opportunity for secure earnings.

Everyone Can Play Their Part

To keep ride-hailing safe and profitable for everyone, we all have a role to play in fighting fraud and helping to protect honest drivers’ incomes in Southeast Asia. We encourage everyone, from consumers to driver partners, to participate and contribute the Grab Fairplay programme. Spot any suspicious behaviour? Report it to fairplay@grab.com

Originally published on grab.com