Betrugserkennung für Online Shops

The detection and targeted combating of fraud in an online shop

Important prerequisites for detecting fraud

In order to ensure effective fraud detection in an online shop, it’s necessary to fulfil the right prerequisites – because, although it may be possible to manually check all incoming orders and purchases in very small online shops with correspondingly low order volumes, this is nothing less than an impossibility in the case of large shops with correspondingly high turnovers. Certain basic requirements must therefore be met in order for a fraud detection solution to work effectively.

Use Case: Chargeback rates at adidas

Read how anlyx helped adidas to reduce chargeback rates by up to 75%.

Download Now

A risk management solution must be deployed

If your online shop is to be effectively protected from fraudsters, it needs a professional risk management solution. On the basis of pre-set risk rules, these types of systems check whether an order will be accepted or refused.

The important thing here is for this solution to be adapted to the needs of the online shop using it.

A good, suitable risk management solution simplifies the recording of individual incidents and their assessment, can monitor and control measures to rectify these incidents and considerably reduces the manual workload. Also, this solution can either be active only in selected countries or, equally, be extended to cover all online activities. It should integrate itself into the existing order processes of the online shop seamlessly and unobtrusively, thus being able to do its job unnoticed by the customer.

The payment data must be accessible

The customer’s payment details are of course particularly sensitive data, over which no company likes to relinquish control – and which are subject to legal restrictions. However, as a rule it’s precisely this data which makes fraud detection in an online shop possible in the first place – whether based on false statements (keyword: identity theft), on stolen data & credit cards or on other crimes of fraud, which can only be identified and uncovered through analysing the data.

Thanks to different systems, many of the payment details can be checked for plausibility, so that this data alone can already make a considerable difference. Thus, for example, many online shops have for some time stopped offering payment methods that pose a high risk to the company. However, this also reduces the potential number of customers who are ready to pay.

Thus, targeted, modern anti-fraud measures are generally a better choice if the online shop is to remain accessible and attractive.

Other factors also play an important role in fraud detection. We’ve defined these in more detail for you below.

Different categories of characteristic signs of fraud

There are different characteristic signs of fraud that are particularly important for effective fraud detection. As a first step, every online shop should watch out for these signs when it’s a matter of exposing potential fraudsters at an early stage.

Seasonal effects in online trading

Fraud crimes in trade are especially frequent around the time of the different public holidays. This is not just because, for example, the fraudsters stand to particularly gain from a simple resale at these times, but is above all also due to the greater workload of companies.

Without an efficient fraud detection solution, a shop can hardly check every single customer and customer contact during periods of strong demand.

The risk of the fraudster’s being caught is therefore lower. This is why shops without their own fraud detection systems very often fall victim to fraud at around the time of major holidays and sales days like Christmas, Black Friday or Easter.

A professional fraud detection system knows that fraud patterns change at Christmas or during Cyber Week and acts to prevent this with alternative risk profiles. In order to find these profiles, it’s necessary to analyse the data gathered during the same season in past years. Because every online shop must be optimally safeguarded.

Geographical effects as a characteristic sign of fraud

At first hearing, “geographical effects” sound insignificant and improbable to most users. However, precisely these effects are not only verifiable but also serve as an excellent tool for fraud detection.

Thus, the fraud detection system in an online shop can, for example, compare the invoicing address to the delivery address in order to check plausibility. Many people do of course have their goods sent to their workplace, but it’s usually in geographical proximity to their own home. If there are excessive discrepancies regarding proximity, that can already be a sign of fraud.

Again, if the customer’s IP data don’t correlate in the least with their place of residence, or if the customer is even using an anonymisation service, the fraud detection system can be triggered and activated.

No successful online shop should fail to take advantage of these options, as they offer a very easy method of effectively detecting and excluding many potential fraudsters.

Sequential effects: purchases without a customer relationship

So-called “sequential effects”, too, are important in fraud detection – because, as a rule, most customers act according to a fixed pattern. Only a few customers will, for example, immediately buy a particularly expensive product on their first purchase in the online shop –since a customer, too, first needs to check whether the dealer is trustworthy.

Although it may be different in the case of really big brands, in a normal online shop a customer frequently has a purchasing history before buying particularly expensive products. Any missing information, such as the date of birth or telephone number for potential queries, can also already be the first sign.

A good fraud detection system covers all these characteristic signs and effectively puts them together to create a risk assessment.

In sum, it can be said that fraud detection in an online shop depends on many different factors. In addition, it isn’t possible to rely completely on the automated systems – as organised crime, too, learns new skills and is continually adapting its behaviour. The specialisation and experience of the decision-makers play a particularly important role here – because, although it isn’t possible to rule out fraudsters 100 per cent using a professional fraud detection system, it is possible to significantly reduce the number of fraud crimes in an online shop.

Use Case: Chargeback rates at adidas

Read how anlyx helped adidas to reduce chargeback rates by up to 75%.

Download Now