< Back to CTAM Think: The AI Series
- November 2026
By Kaitlyn Lefeve
Director
KS&R
What if the experience that contributes to a customer’s decision to cancel happens weeks before they actually leave?
It’s Friday night, and I’m settling in to watch the latest episode of The Great British Bake Off. Halfway through, the stream starts buffering. I restart the app, check my Wi-Fi, and try again. After a few more interruptions, I give up watching for the night.
I’m not going to cancel my subscription right then. But a few weeks later, when I’m reviewing my monthly subscriptions, that streaming service might be one I consider dropping. My reason? “It’s not worth the price.”
The provider sees the cancellation, but probably doesn’t know about the frustrating Friday night that may have played a part in my decision.
And that’s where I think AI could help. What if providers could recognize these types of issues and help resolve them before customers even start thinking about canceling?
More Services, More Opportunities for Frustration
Consumers today are managing multiple entertainment services, along with broadband, home Wi-Fi, smart TVs, and streaming devices.
In KS&R’s nationwide survey of 1,100 consumers, 80% said their household uses streaming services, 46% have traditional pay TV, and a third have both.
With so many services and devices involved, there’s a lot that can go wrong. And when something does, it’s not always clear to customers what’s causing the issue or who they should contact for help.
For providers, it’s not just about getting customers to sign up. It’s also about making sure they’re having a good enough experience to want to stay.
What Really Drives Customers to Cancel?
Our research found that the most common reasons for canceling a streaming service were not using it enough, wanting to reduce spending, and feeling the service wasn’t worth the price.
Technical issues were mentioned much less frequently, accounting for only about 6% of cancellations.
But I think that’s worth looking at a little more closely.
Just because someone doesn’t say they’re canceling because of technical issues doesn’t necessarily mean those issues didn’t play a part in their decision.
Think about a customer who’s dealt with buffering several times, had trouble getting something to work, or spent too much time trying to troubleshoot an issue. They might not cancel because of any one of those experiences. But over time, those frustrations could start to change how they feel about the service.
And when they’re eventually deciding which subscriptions to keep, they might simply say, “It’s not worth the price.”
Our research doesn’t tell us whether technical frustrations are directly contributing to churn, but I think it’s an area worth exploring.
Especially when you consider that 40% of consumers in our research said they were spending less on discretionary items compared to a year ago, and 22% said TV or streaming services would be among the first expenses they would cut.
When customers are already looking for ways to save money, providers have even more reason to make sure they’re delivering a consistently good experience.
Where AI Could Make a Difference
Today, customer support typically starts when something goes wrong and the customer reaches out for help. They explain the problem, go through troubleshooting steps, and sometimes get transferred between teams before getting an answer.
But what if providers could get ahead of some of those issues?
AI could help providers connect information they may already have, such as playback failures, buffering issues, and changes in viewing behavior, to identify customers who might be experiencing problems.
For example, if a customer has several streaming issues over a short period and then suddenly stops using the service as much, that could be a sign that something isn’t working as expected.
AI could help identify those patterns and flag customers who might benefit from additional support.
It could also help make troubleshooting easier. Rather than asking customers to go through a long list of steps, providers could use available information to help narrow down whether the problem is related to the app, device, Wi-Fi, or broadband connection.
Of course, not every provider will have visibility into all of those areas, but even helping customers figure out where to start could make a difference.
And this doesn’t have to mean replacing customer service agents with AI. It could simply mean giving agents better information so they can help customers resolve problems faster.
The bigger opportunity, though, is figuring out whether providers can address some of these frustrations before customers reach the point of wanting to leave.
Not every buffering issue needs a proactive message, and providers would need to figure out when reaching out would be helpful. But I think there’s value in exploring how AI could make support less reactive.
Whether that ultimately improves retention is something providers would still need to test.
Consumers See Some Value in AI, but Trust Matters
Of course, just because providers can use AI to identify potential issues doesn’t mean every customer will want them to.
Our research suggests there’s some interest in AI when it’s helping solve a specific problem.
When we asked consumers about potential AI features in a hospitality setting, 23% expressed interest in AI helping troubleshoot Wi-Fi and TV issues. That increased to 30% among households with both traditional pay TV and streaming.
At the same time, 40% said they wouldn’t want AI assistance with any of the features we tested.
While these findings aren’t specific to using AI for customer support at home, they do suggest that providers need to be thoughtful about how they introduce it.
Customers should know what information is being used and why, and they should still have the option to speak with a person when they need help.
The last thing providers want is for AI to become another source of frustration.
Addressing Problems Before Customers Decide to Leave
A lot of retention efforts focus on customers who are already thinking about canceling. But by then, some of the experiences influencing their decision may have happened weeks or even months earlier.
I think there’s an opportunity for providers to start looking at those earlier experiences.
One way to test this would be to identify customers who are experiencing repeated technical issues, offer proactive support, and see what happens afterward.
Do they start using the service more often? Does their satisfaction improve? Are they more likely to stay subscribed compared to similar customers who didn’t receive that support?
Those are the types of questions providers should be asking to understand whether AI-enabled support can actually make a difference.
Going back to my Friday night watching The Great British Bake Off, what if the provider had recognized the buffering issue and helped me get it resolved before I gave up watching?
It wouldn’t necessarily mean I’d keep my subscription forever. But it probably would have made me feel better about the service.
And that’s really the opportunity I see with AI.
It’s not necessarily about predicting exactly who’s going to cancel. It’s about helping providers recognize when customers are having a frustrating experience and finding ways to make it better before those frustrations add up.
If providers can do that consistently, it’s worth finding out whether it leads to happier customers who are more likely to stick around.
Visit ksrinc.com
