68 Million Customers Told You What's Wrong. Your CRM Never Saw It.

Most voice-of-the-customer (VoC) programs are built on owned channels: support tickets, NPS surveys, CSAT scores, post-interaction emails. Valuable, yes. But they represent a sliver of what customers are actually saying.

Every day, millions of users leave detailed, unsolicited feedback in app store reviews—describing exactly what’s broken, what’s frustrating, and what’s eroding their trust.

This feedback is public, it’s continuous, and it’s specific enough to route to engineering tomorrow morning. Almost no CRM or customer intelligence team is using it.

We analyzed 67.7 million app store reviews across 8,000 iOS and Android apps from 2025—the largest public study of app quality ever conducted. The findings challenge several core assumptions on which CRM teams operate daily.

Your Customers Aren’t Asking for What You Think They’re Asking for

The dominant assumption in most product and CX organizations is that users want more: more capabilities, more innovation. They don’t—or at least, not first.

Users filed complaints about broken user experience: confusing navigation, login failures, layout problems, broken workflows—at six times the rate of feature requests. Six to one.

This ratio held across every vertical we studied: finance, gaming, health and fitness, shopping, social networking, travel. The signal is consistent and overwhelming. Users want what exists to work before they want anything new.

Most VoC programs miss this entirely because app reviews sit outside the traditional CRM stack. Support tickets capture users who are frustrated enough to contact you. Surveys capture users who are engaged enough to respond. App reviews capture everyone in between: the silent majority who won’t open a ticket but will tell the App Store exactly what’s wrong.

Monetization Is Now a Customer Relationship Problem

“Too many ads” was the No. 1 quality complaint across every vertical in our study. Ad-related complaints nearly tripled year over year, topping 1 million across 8,000 apps. In gaming, users now complain more about ads than crashes, meaning the monetization model generates more frustration than the app literally breaking.

For CRM teams, this should be a red flag because the monetization model is actively undermining the customer relationships your programs are designed to protect. Your retention campaigns and loyalty programs are working against a headwind created by the product itself.

CRM teams tracking churn drivers would benefit from connecting this signal. The user who churns after their third session isn’t showing up in your NPS survey. But there’s a strong chance they left a one-star review that says exactly why they left.

Trust Is Eroding in Ways CSAT and NPS Don’t Surface

More than 555,000 complaints in our dataset involved login failures and account access issues. Another 220,000-plus high-severity reports cited account compromise, unauthorized payments, or fraud.

Users are no longer evaluating apps according to just “Does this feature work?” They’re asking a more fundamental question: “Do I feel safe here?”

This is a critical blind spot for traditional CRM metrics. CSAT measures satisfaction with a specific interaction. NPS measures likelihood to recommend. Neither captures what happens to a customer’s relationship with your brand when they can’t log in for three days, see an unauthorized charge, or start wondering whether their data is safe.

These trust signals show up clearly in app reviews—often in granular, specific detail that no survey would capture. A user who writes “I’ve been locked out for three days and nobody has responded” is telling you something your CSAT score won’t.

The Next Generation of Customer Intelligence Needs Public Signals

The CRM stack was built for owned channels. That made sense when those were the primary places customers talked to companies. It doesn’t account for the fact that customers now talk about you far more than they talk to you.

App reviews, social posts, forum threads, and community discussions contain feedback specific enough to act on that most organizations never see—or see too late, in a quarterly digest that’s already stale by the time it lands.

The opportunity for CRM and customer intelligence leaders is to expand the aperture. Not to replace owned-channel VoC programs, but to augment them with the public signals that capture what customers say when they’re not talking directly to you.

When we looked at the gap between four-star and five-star reviews across every vertical, the difference was almost never a missing feature. It was fixable friction: ad frequency, login issues, navigation confusion. These are users who already like the product; they’re one update away from becoming advocates.

That's the kind of insight that should be flowing into every CRM system, every customer health score, and every retention model. Right now, for most organizations, it isn't.

The feedback already exists—public, specific, and continuous. The only question is whether your customer intelligence program is built to hear it.

Christian Wiklund is CEO and c-founder of unitQ, an AI-powered customer feedback intelligence platform used by companies like PayPal, Pinterest, DraftKings, and Bumble to turn public and private feedback into actionable insights. The 2026 unitQ Benchmark Report is available at https://get.unitq.com/2026-benchmark-report.