The 2026 World Cup gave Bluesky the clearest live-event test it has had since launch: a month-long global tournament, 104 matches, and a platform that spent its 2026 roadmap explicitly promising to be better at exactly this.

Here is the July picture, and what the tournament did and did not do to it.

July 2026 Snapshot

Metric July 2026 vs June
Registered accounts ~46 million +~1.5 million
Monthly active users (estimates) 12–27 million Flat
Daily active users (estimates) ~1.5–3 million Flat, with event spikes
Posting volume during matches Sharply elevated Up

Spikes Are Not Retention

This is the distinction that matters, and it applies to every platform, not just Bluesky.

A live event reliably produces a volume spike: existing users post more, post faster, and stay in the app longer during the event window. That shows up clearly in firehose throughput and it is genuinely useful – it is why our trending hashtags tool looks completely different during a match than at 3am.

What a live event does not reliably produce is new daily actives who stay. Someone who opens Bluesky to follow a match and closes it at full time has not been retained; they have been visited. The retention question is whether they come back on a Tuesday in September when nothing is happening.

Historically, across the platform's entire history – election nights, awards shows, previous tournaments – Bluesky's event spikes have not converted well. The pattern is a sharp peak followed by a return to baseline within days.

Why That Is Not Actually Damning

Two pieces of context before anyone extrapolates too far.

First, the roadmap work is not finished. Bluesky's live-moments investment – curated event surfaces, better real-time feeds, improved Discover during breaking news – was announced in January 2026 and is a multi-quarter effort. Judging it on a tournament in June and July is premature. Feed and retention work shows up in the data on a six-to-twelve-month lag, not a six-week one.

Second, the conversion problem is upstream of the event. If someone arrives during a match, opens the app, and sees an empty Following feed and a mediocre Discover feed, no amount of live-event tooling saves that. The onboarding fix and the live-event fix are the same fix, which is presumably why both are on the same roadmap.

What Actually Worked

Being fair to the platform, some things went well during the tournament:

  • Chronological feeds during live matches remain genuinely superior to algorithmic timelines. A feed that does not reorder itself while you are watching a game is a real advantage and people noticed.
  • Multilingual coverage was a standout. Bluesky's large Brazilian and Japanese communities meant the tournament conversation was meaningfully international rather than an English-language monoculture. That is a differentiator few platforms can claim.
  • The tactics and data community delivered. Bluesky's technical skew produced analysis quality well above the platform's size.
  • Group chats found a use case. Anecdotally, the group chats launched in June saw heavy match-watching use – which is good for retention and bad for public-timeline metrics, simultaneously.

The Metric Nobody Is Measuring

Worth flagging: the shift toward private conversation makes Bluesky harder to measure in a way that will distort third-party estimates going forward.

Network-side MAU estimates count accounts writing public records. Group chat messages are not public AT Protocol records – they run on separate infrastructure and never touch the firehose. A user who moves from public replying to group chatting will look less active to every third-party tracker while being more retained in reality.

Expect this to widen the gap between what estimates say and what is actually happening. It is another reason to treat published MAU figures with the skepticism we laid out in May's methodology breakdown.

Where This Leaves Mid-2026

  • Registrations: steady at ~1.5 million a month. Approaching 46 million. Unremarkable and healthy.
  • Daily engagement: flat, below the June 2025 peak of ~4.1 million, with event-driven spikes that do not persist.
  • Direction of travel: the platform is investing in retention rather than acquisition, which is the correct read of its own problem.
  • Verdict on the World Cup test: incomplete. Good spikes, unclear conversion, and the tooling it was meant to showcase is not fully shipped.

Frequently Asked Questions

Did the World Cup increase Bluesky's user count?

Registration growth remained near its steady ~1.5 million per month. There was no dramatic tournament-driven signup surge of the kind the 2024 migration events produced.

What is Bluesky's latest MAU figure?

There is no official figure. Third-party estimates for mid-2026 range from 12 to 27 million depending on methodology.

Is Bluesky good for live events?

Good for quality of conversation and chronological reliability; weaker on raw volume and highlight video. Improving, per the platform's stated roadmap.

Watch the Spikes Yourself

Download Skyscraper for iOS →

Related Reading