On Thursday 27 August 2026, our second Ember Night turned Royal Star near Old Street into one shared room for meeting people in real life. Thirty adults joined. By the end of the night, they had sent 155 private waves and made 31 mutual matches.
That is nearly twice the room size of our first pilot at The Beehive Clapham, with more than three times as many waves and matches. It is an encouraging step, but it is still one event at one London pub. These results are a field report, not a promise that every Ember Night will look the same.
Here is what happened at our real-life social event in London, what the numbers can tell us and what we still need to learn.
Real-life social event results at Royal Star
All figures below come from Ember's anonymised room metrics for Royal Star on 27 August 2026. The event ran from 5pm until midnight, while the venue room remained open until its normal 3am close. Guest identities and message content are not included in these results.
- 30 unique guests joined, including three with a previous recorded Ember visit to Royal Star.
- Guests sent 155 unique wave attempts. 150 remained active and five were withdrawn.
- 26 of 30 guests sent at least one wave. Every guest received at least one.
- 31 mutual matches formed, involving 25 different guests.
- 23 in-app messages were sent across 11 matches.
- Ten match rewards were created and four were redeemed at the bar.
- No safety reports or blocks were recorded.
The clearest result is the denominator: 25 of 30 guests made at least one mutual match. That is 83% of the room. One person can match with several people, so 31 matches does not mean 62 different guests. Publishing the distinct count keeps the result honest and gives us something useful to compare with future nights.
A bigger room stayed active
At The Beehive, 16 guests sent 50 waves and made nine matches, involving 11 people. At Royal Star, 30 guests sent 155 waves and made 31 matches, involving 25 people.
The room did more than grow. Participation remained broad. Twenty-six guests sent a wave and all 30 received one. Five wave attempts were withdrawn, which matters too. Ember lets somebody change their mind quietly, and an individual wave stays private unless the other person waves back.
The first guest joined at 5.31pm. The first mutual match formed at 5.46pm, around 15 minutes later. Matches continued throughout the listed event and the final one formed just after midnight.
That timing is useful. Ember did not need a formal matching round or a single starting gun. People arrived, looked around and joined at their own pace. The product supported the evening instead of asking the whole pub to stop and follow a programme.
Most guests were open to friendship and conversation
The intention mix showed that people came for different kinds of connection:
- 19 guests selected making friends.
- Five selected hanging out.
- Three selected being up for a chat.
- Three selected flirting.
That means 27 of the 30 guests chose a main intention other than flirting.
People wanted a reason to meet somebody new without having to define the outcome in advance. A match might become a new friend, an introduction or simply a good conversation.
Ember should keep making room for that ambiguity. The private wave asks whether both people are open to saying hello. It does not decide what the connection has to become.
The phone remained a bridge, not the destination
Guests sent 23 in-app messages across 11 of the 31 matches. We cannot conclude from the event log that every matched pair spoke face to face. We can say that most matches did not turn into long conversations inside the app.
That is the intended shape of an Ember Night. Everyone shown in the room is already at the same venue. A short message can help two people find each other, but the useful next step is only a few tables away.
The presence data points in the same direction, with an important caveat. Recorded presence averaged 243 minutes and the median was 265 minutes. Twenty-seven guests recorded at least an hour in the room, and 24 recorded at least two hours.
This is based on join time and the last activity seen by the system. It is not an independently measured pub dwell time and it should not be treated as a sales figure. It does suggest that Royal Star was not a quick check-in followed by an immediate exit.
The busiest part of the night came later
Arrivals were spread across five hours. Four guests joined during the 5pm hour, six during the 6pm hour, nine during the 7pm hour, another nine during the 8pm hour and two during the 9pm hour.
In other words, 18 of 30 guests arrived between 7pm and 9pm. The final guest joined at 9.26pm, almost four hours after the first.
The practical lesson is simple: an Ember Night needs to be ready for late arrivals. Clear signage, staff awareness and an easy first-time setup cannot disappear after the opening hour. A welcome speech at 5pm would have missed most of the room.
Session recovery was quieter this time too. Royal Star recorded four rejoin events across three guests, compared with 38 rejoins across 10 guests at the Beehive. A rejoin means reopening the app or reconnecting a session, not another visit to the venue. The lower count is encouraging, although two events are not enough to attribute the change to any single product improvement.
Four pairs claimed their match reward
Ten matches created a match reward and four were redeemed at the bar. That is four pairs who completed the full flow together: join the room, exchange private waves, match and present the reward to staff.
The log cannot tell us why the other six rewards were not used. Some guests may have preferred to keep talking, left before claiming or missed the explanation. That is a question for direct participant feedback, not something we should guess from a database row.
For the next night, we want to keep the reward explanation short and visible at the moment a match happens. Venue staff also need a quick reminder of the redemption flow before guests arrive. Royal Star showed that the system works live; the next task is making sure every matched pair understands the choice in front of them.
What Royal Star taught us
The second Ember Night gave us five practical lessons:
- A room of 30 can still produce broad participation. Twenty-six people sent a wave and 25 made a match.
- A real-life social event can work without forcing every connection into a label. Friendship was the most common intention by far.
- The event has to work as a rolling evening. Most guests arrived after the first two hours.
- A small amount of in-app messaging is enough to support the room without replacing it.
- Product data can show joins, waves, matches and redemptions. It cannot prove whether two people enjoyed their conversation.
That fifth lesson remains the most important limit. Thirty guests and 31 matches are useful numbers, but the real outcome happened away from the screen. A respectful, optional end-of-night question will help us learn more without asking people to identify who they met or what they said.
No safety reports or blocks were recorded. That is good to see, but it does not validate a safety system on its own. Ember Nights remain for adults aged 18+, waves remain private, and blocking and reporting remain available throughout the night.
Royal Star event questions
Did guests have assigned conversations?
No. There were no rotations, assigned partners or forced conversations. Guests joined one venue room, chose their intention and decided who they wanted to wave at.
How private was the event?
An individual wave was only revealed when it became mutual. Guest identities and message content are not part of the anonymised figures in this report. The live room, including its waves and matches, was cleared within 24 hours of close.
Did everybody make a match?
Twenty-five of the 30 guests made at least one mutual match. Five did not. Ember can make the first signal quieter, but it cannot and should not guarantee a match.
What happens after Royal Star?
Royal Star gave us a larger and more active room than our first pilot, while keeping the experience open to friendship, conversation and new connections. Thank you to the venue team and everyone who joined, waved, matched or simply gave a new kind of night a try.
The next step is another room, another venue and another honest comparison. You can read how Ember works, check upcoming Ember nights, or learn about bringing Ember to your venue.