Every subscription video on demand (VoD) service we’ve worked with has wanted more subscribers. More subscribers equals more revenue, adding to the bottom line.

To achieve this, VoD services understandably exert a huge amount of effort on customer acquisition — devising ways to bring people to the product, eventually to convert them into subscribers. A service will secure rights to exclusive content, design beautiful products, and raise awareness through integrated marketing campaigns, all with this goal in mind.

When ever-increasing subscription targets aren’t being met, the temptation is to double down on the following approach: add more desirable functionality, in the hope of attracting more users.

If successful, this can achieve results. Killer new features can give acquisition metrics a boost. Yet killer features are few and far between. Often, creating more features leads to what Jared Spool, the renowned usability researcher, describes as ‘Experience Rot’:

“As you add features, you’re adding complexity to the design, and decreasing the quality of the experience.” – Jared Spool

In pirate metrics — a marketing phrase coined by Dave McClure — acquisition has a less glamorous but equally powerful sibling: retention.

Acquiring new subscribers may boost metrics in the short term, but if you have problems keeping people you’re fighting a losing battle.

To get better at retaining subscribers for longer, we first need to understand why customers are cancelling their subscriptions. Reassuringly, the causes for churn are usually quite predictable.

Churn tends to happen for the same reasons

Having worked across very many subscription VoD services over the last decade, we’ve observed that the reasons users cite for leaving are surprisingly common:

• Technical issues: I can’t login or subscribe; it doesn’t work on my device; streaming quality is poor
• Lack of value: I only watch one show, so once I’ve watched the latest season there’s nothing else for me; there’s not enough content; there’s not enough good content; there’s never anything new

VoD service providers find it difficult to work out which problems their audiences have, which are the most important, which are avoidable, and which the best solutions are.

Your users expect technical excellence

With a high bar set by the most popular streaming services, users need little encouragement to post technical issues on forums or social channels, or to contact support teams. For issues that aren’t reported, analytics and performance monitoring tools can highlight problems: for example, failed payments, playback errors, outages.

By no means are the solutions for technical issues trivial — an eagle-eyed client of ours identified that payments were failing more at certain times of day, and has been exploring solutions to mitigate this — but at least it’s clear what the issues are.

Whether or not these problems are avoidable usually comes down to internal backlog prioritisation — what the business is focused on now — or trying to influence the roadmaps of third party service providers so that they can provide the required functionality.

Value is nuanced and non-rational

The value (or lack thereof) of a streaming service is inherently more subjective; each user makes their own value judgement, based on their own criteria and perceptions.

This makes value more difficult to interpret, but also increases our ability to influence it; if we can change people’s perceptions of what their subscription gives them, we can affect the value they place on the service.

Understood well enough, seemingly unavoidable churn can become avoidable.

A few years ago, one of Europe’s biggest media companies commissioned us to find ways to reduce churn across their product suite. People who left complained that there wasn’t enough good content, therefore the product wasn’t worth the money.

Our client showed lists of content to those subscribers who’d left, and asked them if they thought that package would be worth the money. The majority of respondents replied positively: that content would definitely be worth it.

The kicker? They’d shown users content which had been available through their subscription. Content they already had access to.

The problem wasn’t a lack of good content, as users had reported. In reality, they weren’t aware of how much good content they had access to, meaning the package was less valuable in their eyes.

Taking this feedback literally would suggest putting more content on the platform, or lowering the price of subscription, neither of which was feasible. However, changing users’ perceptions of how much content they had was very achievable. We needed to find the right features and functionality to improve content discovery throughout the product.

How to fight churn

In more than 12 years of designing and building VoD products, these same reasons for churn arise time and time again.

Because each client’s content catalogue, editorial capability, and audiences are slightly (or completely!) different, however, there is no ‘best practice’ feature set which can magically reduce churn. What works for one VoD service won’t necessarily work for another.

The best approach is always to talk to users, find out what their issues are, and which you can realistically tackle. With this knowledge, you can then design and test a range of solutions, before shipping them to the live product.

Talking to users is the best way to find out why they’re leaving

We’ve consistently found that by far the most efficient way to learn what’s driving user behaviour — including churn — is to talk to users.

Semi-structured interviews will help you hone in on the issues that you can easily overcome. Complementing this with user testing on your live product is even better. Often, users aren’t the most reliable source in speaking about their behaviours, but seeing them using the product can really help you see where the problems are.

With just 2 to 3 days of interviews, you can get a huge steer as to where you should be investing time and effort. Once you understand the concerns about value that users have, you can assess whether they’re avoidable or not:

• If a large percentage of users only watch one show, can you design flows which encourage them to watch more?
• If users feel there’s not enough good content, can you help them easily discover more?
• If users feel the catalogue is static, can you do more to highlight the new or upcoming content?

Prototyping is the quickest and cheapest way to test solutions

Even armed with knowledge about what the problems are, finding the right solution is tricky. Users are familiar with the way the product is now — for better or worse — and so introducing any changes or new features is risky.

Mistakes in the live product can be costly in more ways than one: you can spend a lot of time and effort on something that has no effect. Or worse, has a negative effect. Before adding features to the backlog, you need to have high confidence that they’re going to do what you need them to, and move the right metrics.

Prototyping is an effective way to test and tweak a range of solutions quickly and cheaply. You can test lots of ideas, iterate rapidly, and fine-tune promising options.

Again, you can learn a lot in 2 or 3 days, and gain some confidence as to which ideas are most likely to succeed. Testing sessions usually consist of:

• 5–10 minutes for an initial interview, to settle the user and their perceptions of the service
• 20–25 minutes of testing solutions with the user, asking them to think aloud as they browse the prototype. (We tend to prefer free-to-explore testing, rather than asking users to complete specific tasks, as this is how they actually use products.)
• Setting aside a separate observation room, where the team will be capturing insights in real time. We encourage clients to participate in this as much as possible. There is nothing quite like watching testing sessions

How to put this into practice

If you are looking to increase your VoD subscription numbers, it’s crucial to keep the audiences you’ve worked so hard to acquire.

It’s worth spending some time understanding in detail why users are leaving. More than likely it’s because of technical issues, or a low value perception.

Learning exactly what the churn issues are, prototyping, testing and then scoping preferred solutions, takes as little as 4–6 weeks end-to-end:

• Week 1: Explore analytics, understand existing research, and recruit users for research
• Week 2: User interviews to pinpoint the key issues
• Weeks 3–4: Design potential solutions and create prototypes for testing
• Week 5: Rapid, iterative testing to test and tweak solutions, and validate which are most likely to succeed
• Week 6: A final week to scope the agreed solutions for build

As an important side benefit, you’ll also learn more about your users in general, which can help you with your other objectives.

If you’d like some expert help to address your churn rates, we offer consultancy to walk your team through this process. We’re also happy to boost your team with experienced designers and researchers to do this on your behalf.

This article was written by Chris Belmore