How the on-demand economy is impacted by fraud

This article was first published on Payments Cards and Mobile here.

On-demand: at any time that someone wants or needs something

The term on-demand refers to the way in which a good or service is sold as opposed to what is sold. The primary characteristic of on-demand is that the customer request is satisfied instantly even if the actual service is delivered a little later. According to Business Insider, “the number of companies, categories, and the growth of the industry is expanding at an accelerating pace. The businesses in this economy represent the manifestation of years of technological innovation and an evolution in consumer behavior.”

And the rapid growth of the on-demand economy is spreading across a variety of industries. Take the taxi industry for example: when a customer in Dublin uses MyTaxi to order a cab, the request for a taxi is satisfied instantly, even if the taxi actually arrives a few minutes later.

The taxi or cab hailing industry perhaps drove the market but the model was swiftly adopted in a number of industries, such as:

Food and grocery delivery: customer in Berlin uses Delivery Hero to choose and select their pizza for the evening. The order is accepted instantly, delivered a little later. Services like Quiqup in London will literally deliver almost anything to anyone.

Marketplaces: While food delivery and cabs are marketplaces a more classic model is a services aggregator. For instance Handy or Bizzby will put a customer in touch with a host of cleaners or other service providers. As well as the service themselves, these businesses really sell convenience and certainty of delivery.

On-demand is in demand

The numbers for the on-demand economy are significantly increasing. According to Harvard Business Review, the on-demand economy is attracting “more than 22.4 million consumers annually and $57.6 billion in spending.

“The largest category of on-demand spending is online marketplaces, with 16.3 million consumers each month spending almost $36 billion annually. Transportation comes in second with 7.3 million monthly consumers and $5.6 billion in annual spending, followed by food/grocery delivery at 5.5 million monthly consumers and $4.6 billion annual spending.”

Why on-demand businesses especially dislike fraud

The marketplace business model typically means very low margins. The majority of any payment goes to the supplier (e.g. the restaurant) while the market maker has to cover its operational expenses from what’s left. This model only works when scale is achieved, but a significant retardant on reaching that scale is fraud.

A single chargeback, for which the marketplace take responsibility, can wipe out the margin of multiple deals and have a vastly disproportionate impact on a marketplace business. This can only be ignored for so long before action is taken to stop it.

What kinds of fraud do businesses like this encounter?

In terms of what the day to day online operations, payments and customer service teams can expect to deal with, there are typically four big categories of fraud, most of which have payment fraud at their core:

Chargebacks: Typically, a user using stolen payment details to purchase goods. Easily the most damaging form of fraud from a financial point of view. Can grow to greater than 2% of all transactions if left unchecked or using a PSP-level fraud tool.

Account takeover fraud: A fraudster using stolen login details to purchase from a legitimate user’s account – or brute forcing their way into the account. The most damaging form of fraud in terms of reputation damage for a business. The social sharing of a hacked account can be expensive to recover from.

Terms of Service abuse: Customers, often legitimate, attempting to re-use vouchers, sharing vouchers, or seeking to exploit a generous returns or disputes policy. Most expensive fraud in terms of operations expense as there are phone calls and investigations to conduct.

Supplier and collusion fraud: A supplier working with a willing accomplice to accept stolen payment details for large orders knowing they will be charged back. Most expensive in erosion of trust in the marketplace community. Also expensive to investigate.

Machine learning and fraud

On-demand has changed how we travel, how we shop, and how we eat. It’s changed the way fraudsters operate too. It’s bought a whole new set of conveniences and services into the reach of unscrupulous actors.

The good news for any scale-up on-demand business is that the data is available to them across their own customer systems and their PSPs. Once that data is ingested by the models, they start to generate fraud scores which are fed back to the merchant. The right shape, size and skillset of a team is determined ultimately by the requirements of the business. There is no single correct answer. However, the key to success is to leave the repetitive, data-driven, rote work to the machines and keep the the human element focused on making customers happy.

To learn more about online payment fraud, visit our insights page.