Events · Social media analytics
Knowing what the room thinks while the event is still running
Re-Work, an international conference organiser, needed to follow audience sentiment across social media during live events. We built a real-time analytics tool that handles opinion monitoring, influencer identification and trend detection, and replaced the manual scrolling entirely.
Key results
Client: Re-Work
An international events company organising conferences and exhibitions focused on technology and innovation.
- Industry
- Events and conferences
- Use case
- Social media analytics for events
- AI approach
- NLP with sentiment analysis
- Data source
- Social media platforms
- Output
- A dashboard with real-time insight
- Product
- Twitter Board (theBlue.ai)
In short
By the time someone noticed, the moment had passed
- At a Re-Work conference, hundreds of attendees, speakers and exhibitors post continuously: reactions to sessions, feedback on logistics, complaints about everything from wifi to catering.
- Before this, someone from the team scrolled through Twitter, guessed at the sentiment and flagged anything urgent. Slow, incomplete and reactive.
- The tool now ingests the posts by hashtag, mention and keyword, scores each one for sentiment and shows the team at a glance whether the mood is positive, negative or mixed.
- It also surfaces who is generating the most reach, where the audience is, and which topic is starting to trend, while there is still time to do something about it.
The starting point
The challenge
Thousands of social media posts were generated during an event, but getting useful, timely insight out of them was manual, slow and unreliable. By the time feedback was noticed, it was too late to act on it.
Re-Work organises conferences and exhibitions where hundreds of attendees, speakers and exhibitors generate a constant stream of social media activity: reactions to sessions, feedback on logistics, shout-outs to speakers, and complaints about everything from wifi to catering.
Understanding that in real time matters twice over: for improving the event while it is running, and for planning better ones afterwards.
Before the project it was done by hand. Someone scrolled through Twitter, tried to gauge the sentiment and flagged anything urgent. By the time negative feedback was noticed, the moment to respond had usually passed.
The build
What we built
We deployed Twitter Board, our social media analytics tool, to give Re-Work a live view of what attendees were saying as they said it.
Sentiment scored as it arrives
The tool continuously ingests posts related to the event through hashtags, mentions and keywords and scores each one. The team sees at a glance whether the overall mood is positive, negative or mixed, and can drill into the specific posts driving that score.
Who is actually being heard
The system identifies which voices are generating the most reach and engagement, speakers, attendees, journalists and industry figures, so the team spends its attention where a reply has the most effect.
Where the audience is
Dashboards show where the activity is coming from geographically, which tells Re-Work how their audience is distributed and where regional engagement is strongest.
Trends before they are obvious
The tool surfaces emerging topics and conversation threads in real time: a session generating buzz, a logistics problem escalating, or an unexpected topic taking hold among attendees.
What changed
The results
Before
Manual social media monitoring and reactive responses. No structured sentiment data, and feedback reviewed after the event, if at all.
After
A real-time dashboard with sentiment scoring, influencer tracking, geographical insight and trend alerts, all while the event is live.
Re-Work moved from reactive to proactive event management. Negative feedback can be addressed while the event is running, and a positive moment can be amplified while it is still happening.
Post-event analysis now rests on structured data rather than anecdotal impressions.
The value also outlasts the individual event: accumulated sentiment and engagement patterns feed decisions about future programming, speaker selection and audience targeting.
Questions about this project
Social media analytics for events. Thousands of social media posts were generated during an event, but getting useful, timely insight out of them was manual, slow and unreliable. By the time feedback was noticed, it was too late to act on it.
Live monitoring of the conversation during an event: Real time. Sentiment scoring and trend detection without manual review: Auto. Geographical view of where the engagement comes from: Geo. The voices with the most reach identified automatically: Influencers.
NLP with sentiment analysis. Technology used: Natural Language Processing, Sentiment analysis, Social media API integration, Real-time data processing, Influencer detection, Dashboards and visualisation.
Technology used
Talk to the people who built this
theBlue.ai comes back within one business day.
More case studies
Telecommunication · Arvato Bertelsmann
Call summaries written by the system, not by the agent
Read more
Pharmaceutical · under NDA
Millions of medical texts read for drug safety signals
Read more
HealthTech · Xpomet Medicinale 2024