Philo + Enterpret: The $1M Question No One Thought to Ask

One question about late-night support contacts saved Philo $1M. Then the rest of the company started asking questions too.

The Foundation

The first thing Jenny McCullough did with Enterpret wasn't ask a question. It was to define categories.

Every support leader I've talked to in the last year would have done the same thing: start asking the big questions immediately. What are customers angry about? Where are we losing subscribers? Which features need fixing? When Philo deployed Enterpret across their feedback ecosystem, Salesforce cases, Amazon Connect calls, app reviews, social media, CSAT surveys, they took a different approach.

Jenny's strategy was a bit more tedious, and certainly less glamorous: she spent the first weeks pruning the taxonomy. Recategorizing. Making sure the system's classification reflected how Philo's customers actually talk about problems, and how internal teams would label them.

With the taxonomy set, Jenny started by asking a question that had been burning for a long time: why do customers contact Philo in the middle of the night? She knew overnight volume was very low, sometimes just a couple of contacts per hour, which greatly increased Philo's cost per contact. Since customers watch TV 24/7, Philo was hesitant to reduce support hours in case there were time-sensitive issues.

With Enterpret, Jenny was able to understand what topics customers were contacting about by interval. It became clear that after midnight, not only did volume decrease greatly, but the main topics were not urgent, and often had steps clearly defined in the Help Center. So Jenny made the decision to reduce Philo's hours of operation by just a few hours each night, saving Philo $1 million annually.

With the $1M win under their belt, Support began leveraging dashboards with other departments to help drive strategy. In late 2025, when Philo made significant changes to their packages, Jenny was able to leverage Enterpret to share feedback with the relevant teams, ensuring messaging was tweaked, and customers were clear about what their package contained and what they were receiving for the price point.

"We had wrestled with the thought to reduce our HOOP for a while, but it wasn’t until we understood contact reasons by interval that we felt comfortable pulling the trigger, saving us $1 million per year."

Early Indicators

Like many streaming services, Philo depends on multiple billing methods to make paying for subscriptions more seamless for customers. Enterpret's quality monitor agent has flagged multiple billing issues, and the incredible thing isn't that Enterpret is raising them. It's the speed: Enterpret alerts Philo of issues hours, sometimes days, faster than they would have learned about a problem under their old weekly reporting model.

When Enterpret flagged increased chatter around a billing topic, Jenny was able to raise the issue and resolve the error quickly. The Support team also played offense, proactively educating agents and customers about what was happening before it escalated.

In early 2026, Philo launched a highly requested feature: the ability to upgrade and downgrade between plans. The company looked to the Support team to share sentiment, and Jenny was able to gather that story using Wisdom in minutes. Philo quickly understood that only a very small percentage of feedback was negative, and identified exactly what was driving the frustration. Jenny used that insight to update Help Center articles and the agent knowledge base, so agents could better explain the process. Within a week, sentiment had improved significantly.

The Compound Effect

Enterpret is improving the life of 'Philons' beyond the Support Team:

• Marketing has leveraged to understand churn patterns
• Business teams have used Enterpret to better understand how changes to Philo’s packages impact key performance metrics
• Product is chatting with Enterpret weekly to ensure their priorities are indicative of customer feedback

Add it up: the hours reclaimed from manual triage, the churn caught before it compounded, anticipating customer concerns, the product decisions that are able to be made in real-time. Jenny's team estimates the operational impact at over a million dollars in the first year, not from one dramatic moment, but from the compound effect of making customer voice accessible to everyone who needed it.

The use cases keep growing, too: In a single week in February, twelve new users started querying feedback for the first time. By March, twenty-four people across six teams. No rollout memo. No mandatory training.

$1M+ estimated Year 1 impact • 6 months from taxonomy to AI agents • 82% positive feature launch sentiment.