Using Data Science to Change Subscriber Strategies and Enhance Customer Experience (Part 4 of 4)
What service providers can do to detect and stop unwanted credential sharing
Any business that provides subscription services online must have mechanisms in place to ensure account security, manage subscription sharing, and prevent fraud. Because of new subscriber strategies, streaming video content and service providers experience high levels of illegitimate account sharing. As seen in our survey, nearly 30% of consumers admit to using shared account credentials – and perhaps the reality is more than consumers are willing to admit. This activity can result in harming businesses by adding costs and losing revenue.
Trade-off between security and consumer expectations
It may seem that the easy answer to prevent credential sharing is heavily restricting access. Indeed, many service providers have recently tightened up on account sharing with policies such as restricting the number of devices, geography, content, and concurrent streaming.
To capture nuance, companies need data science that leverages the vast amount of information captured by their video services and that combines machine learning with behavioral analysis.
However, consumer expectations, fueled by the concept of “TV Everywhere”, are that they should be able to use any device, watch anywhere they want, access all the content, and have multiple users – users that include close friends and family they don’t live with. As one respondent commented, “If I can make my friends and family happier by sharing accounts, by all means I will share them!”
Some users might also misunderstand service agreements and see multiple devices as permission to share subscription services, “I pay for two screens, so therefore I feel it’s ok to share my extra screen with a close family member”.
Content providers appear to be faced with a trade-off between platform security and building an experience that meets increasing consumer expectations.
Source: How Consumers Access Streaming Video: The Risks of Credential Sharing
Applying blanket restrictive policies may negatively impact the user experience which would not only lead to customer churn but viewers to turn to piracy. For example, restricting location and content could see a rise in geo-filtering via VPNs.
However, it’s important to note that most account sharing occurs only within households (as allowed), and a significant number of subscribers do not share, nor do they intend to. In our survey, 72% of subscribers do not share account subscriptions at all or only share within the household.
Instead, service providers and content owners need solutions that seek to protect account access and the user experience. Is there a way to achieve both?
Using data science for targeted action
Making general platform restrictions is not the answer to uncontrolled credential sharing if companies want to meet and exceed growing consumer expectations. Like the solutions to pirate activity, taking targeted action is.
Unlike many forms of piracy, service providers can track and evaluate sharing behavior using the information they already have.
|Read: Case Study: Credential Sharing Analytics and Strategies for a Service Provider|
The kind of analysis needed goes beyond simple account-level metrics such as looking at the number of active devices or activity to identify sharing accounts. Like general platform policies, a simple approach will fail to capture the nuance involved in many cases.
For example, take a household of a family with two working adults and two young children. Their viewing footprint could look like:
- Watch at home on a TV screen
- Use two tablets for travelling (e.g. for kids on a long journey)
- Use individual laptops to view content, both in and out of the home
- Occasionally cast to a TV at someone else’s house
- Occasionally watch content on mobile devices
All this usage is legitimate. But a simple analysis of this data would register a high number of devices and so activity may be classified incorrectly as illegitimate.
To capture the nuance, companies need data science that leverages the vast amount of information captured by their video services and that combines machine learning with behavioral analysis.
Potential fraud indicators include out-of-home devices, geographically dispersed activity, out-of-home streaming volume, and repeat device types.
Identifying sharing activity
Video services collect vast data that contains valuable signals to help determine which activity is legitimate or likely abuse or fraud.
Source: How Consumers Access Streaming Video: The Risks of Credential Sharing
In our survey, of those who share account details with others, one out of three (28%) share with someone outside the household – not legitimate use for most subscription contracts. When analyzing viewing information, additional signals could indicate sharing with a non-household viewer. In our family example, we can use these potential red flags:
- Watching from a TV screen at another location becomes regular
- Viewing live programs concurrently in multiple locations
- Repeated content streams
Using combinations of multiple signals, service providers can better track and evaluate usage patterns and discrepancies to create a holistic view of sharing on an account. From here, they can select accounts and match them to the appropriate action.
All this information is legitimately gathered for a service provider to effectively deliver a video service. In the same way, they use it to improve services and the user experience, they can leverage the data to deter credential sharing abuse and identify account theft.
The fight to protect content distribution has a new challenge in the form of credential sharing. What is clear is that credential sharing is no longer the safe subscriber strategy for new and old players alike.
Consumers share their video streaming subscriptions a lot more freely than companies might believe – and it certainly has consequences.
Ignoring credential sharing activity puts businesses at risk for increased costs, service quality issues, litigation, security breaches, and revenue opportunity loss.
Alarmingly, subscription fatigue has also come where many consumers do not want to pay more for content, “There are so many different ones to pay for that it gets ridiculous to pay for them all yourself when you want to watch say one show on Amazon and one show on Netflix and then the other show you like is on Hulu. Makes sense to share with people you know to save money for you both.”
With more online video services launching in the US and elsewhere, will this drive competition among legitimate services or push consumers to piracy or to share accounts?
Yet there are sharers who are willing to pay and either lack the incentive to do so or find that borrowing credentials is an easy alternative. There is hope yet for companies to convert these viewers.
Applying blanket usage controls is a solution but not a satisfying answer. They seem out of place in the world of “TV Everywhere” with the push for ubiquitous connectivity and numerous connected devices.
|Read: Streaming Video Credential Sharing: How Service Providers Can Identify and Stop Unwanted Password Sharing|
Like fighting piracy, detective work is needed instead, aided by effectively using data science and machine learning. Video service providers can identify issues with the viewing information they already have and take appropriate action to reduce unwanted sharing activity.
Credential sharing is a problem that has gone on unchecked for a long time. The solution to curbing account sharing abuse and fraud lies within viewing data, and it is up to video service providers to take the lead. <>
> Cartesian report: How Consumers Access Streaming Video: The Risks of Credential Sharing
Referencing survey results and drawing from our work in analytics and content security, this report:
- Reveals consumer viewing habits and attitudes towards video services access;
- Explains credential sharing in the context of piracy and content protection;
- Defines unwanted credential sharing and its three main types;
- Examines the risk to content owners and service providers, and;
- Outlines an approach to track, stop, and prevent unwanted sharing activity while minimizing the negative impact on customer experience
Cartesian’s Streaming Video Credential Sharing Detection & Prevention Solution uses data science, machine learning and behavioral analysis to detect, prevent and stop unwanted credential sharing, while allowing operators to continue to provide a rich customer experience. Is this something your business needs? Get in touch to see how we can help.