The Scoop: StubHub sympathizes, but defends its business after ticket problems

StubHub built its business with the idea that you buy a ticket through its platform, and if something goes wrong, the company will make it right. It’s a warm and fuzzy guarantee.

But a Wall Street Journal investigation found that promise didn’t hold up. Customers were denied entry to events, received worse replacement seats or spent weeks trying to get refunds they had already been promised. Complaints to the Better Business Bureau rose 47% from 2024 to 2025.

The recent World Cup added more scrutiny. Some fans said tickets they bought months earlier never showed up in FIFA’s ticketing app. Others said StubHub told them not to worry, only for the tickets to never arrive. StubHub then pointed to problems with FIFA’s mobile ticketing system.

Chief Business Officer Raj Beri also acknowledged: “We understand that missing an event isn’t just a bad user experience, it’s deeply disappointing on an emotional level. It’s also bad business, because our marketplace only works when fans attend an event.”

During the NBA playoffs, one customer paid more than $8,500 for two Knicks tickets but received tickets to a comedy show instead. StubHub promised a refund, but the customer spent three months calling the company trying to get his money back.

Meanwhile, his StubHub app was still blaming the refund delays on the “current COVID-19 pandemic.”

StubHub has since defended its overall performance. A spokesperson told WSJ that about 1% of orders result in refunds and that most are processed within a week.

Why it matters: StubHub’s major issue isn’t that ticket issues occasionally happen. It’s what happens after they do.

The company acknowledged the complaints broadly, but it didn’t offer any reassurance of what it was doing to mitigate these issues or improve response times.

Multiple delays, mixed messages and repeated calls can turn a customer service problem into a reputation problem really quickly.

And that old Covid message is a good reminder for communicators too: Automated emails, app notifications and help pages all count as brand communications. Six years later, COVID isn’t the cause of refund delays.

Audit what customers see when they have a problem. A bad or outdated message can make an already frustrating experience look like a sign that nobody is paying attention and trust will plummet.

How many users are still using Bluesky?

New data shows fewer people are regularly on Bluesky.

Its mobile app had 10.4 million monthly active users worldwide in June, down 27.2% from a year earlier, according to Similarweb data reported by TechCrunch.

Daily active users also fell about 26% in July to roughly 3 million. The drop is even more severe compared with people who joined after the 2024 election. Bluesky’s monthly active app users are down roughly 52% from the fourth quarter of 2024.

Meanwhile, Threads is moving in the opposite direction, with daily active users up 21.3% year over year in July.

Bluesky says it’s now adding group chats and other features for smaller communities. The platform’s bigger goal is to move social media away from individual platforms toward an open network of services.

Brands should use this data to examine user behavior. Bluesky has nearly 46 million registered users, but active-use numbers tell a much different story. Companies should look at where their specific audiences are actually spending time and engaging before investing heavily in any platform.

It also shows how social media is becoming less about building a following on one big app and more about reaching smaller communities across platforms.

Why did Radio Caroline announce King Charles’ death?

British radio station Radio Caroline wrongly announced that King Charles III had died after a staffer found prerecorded death announcement files and played them “out of curiosity.”

Vanity Fair reports the files went out live, followed by the national anthem and 16 minutes of silence. The station soon realized its mistake with the host saying, “I’ve just been informed that we’ve played in error some information a little earlier; I didn’t hear this myself, but it is incorrect, it’s a technical issue, and of course we apologize.”

But regulator Ofcom later said the mistake was not corrected quickly enough and found that Radio Caroline broke two broadcasting rules.

The employee has since been reprimanded and apologized, and Radio Caroline said the sensitive files will now be stored on an external hard drive.

The mistake is an unusual one, but the lesson for brands isn’t. Don’t rush to explain why something happened until you know what actually happened. It’s not a great look and it can quickly mar credibility.

Radio Caroline’s early “technical issue” explanation didn’t tell the full story, which gave the incident a second life later when Ofcom released its findings.

Brands should correct serious mistakes quickly, sticking to what they know for sure. When they mess up, or there’s confusion, clarify and own it. It’s the only way to win back trust.

Why did the New York Post develop its own chatbot?

The New York Post is rolling out its own AI assistant, called Hamilton, across its apps.

Axios reports readers can ask it questions about stories, get personalized news briefings and receive recommendations based on their interests.

Hamilton pulls from the Post’s current and archived stories. The Post also controls the editorial boundaries, while Google provides the tech behind it.

The move is partly about keeping audiences closer. Publishers are losing some search and social traffic as people increasingly get info through AI tools. Hamilton gives the Post a way to offer that same conversational experience without sending readers somewhere else.

The same idea can apply beyond newsrooms. If customers are asking AI for answers, brands should think about when those questions are about their products or expertise.

Companies may have an opportunity to build AI tools around their own content. The important part is more control and user engagement. Give the AI clear material, set boundaries for what it can say and make sure users can tell where its answers come from.