When you tell a room of event organizers that you want a social wall, the first question is always the same: ‘Which software will we use?’ That is precisely the wrong question to start with. Working with a 295-employee hotel group in Antalya that runs four large international congresses per year, I watched this play out in real time. The technical team had the Twitter stream feeding onto the main conference hall screen in three days. Forty minutes into the opening session of the first congress, management called me: two politically charged tweets, three comments praising a competitor hotel, and an inappropriate hashtag containing attendee names were displaying in full view. The system worked. The process did not exist.The core tension in a social wall project is structural, not technical. Attendees want to share the event in the moment — that is the whole point. The organization wants to protect brand integrity and control the narrative. These two desires are in natural conflict. Technology does not resolve that conflict; it makes it visible to everyone in the room simultaneously. The software choice — whether you use Tweetwall, Visible Tweets, or a custom-built display solution — is the smallest decision in the project. The significant decisions are: which content passes automatically, which waits for moderator approval, who holds authority during a crisis, and how that authority chain operates not on paper but in a live, high-pressure environment. A social wall deployed without answering these questions hands the organization a brand risk it did not budget for.In Turkey at the start of 2011, integrating social media into live events is still an emerging discipline. Facebook’s user base is growing rapidly, and Twitter has become an influential channel, especially in media and business circles. Yet the majority of event organizers still treat a social wall as an IT installation task. They hand the project to the IT department or an agency, define no editorial rules, and expect a professional outcome. The result is a technically functional but editorially unsupervised screen. This article addresses what a social wall project actually requires — and which operational questions must be answered before any technical configuration begins.On the technical side, the options available in early 2011 are reasonably well-defined. Twitter’s public API provides a streaming feed of tweets filtered by hashtag, with rate limits that are manageable for single-event deployments. A dedicated event hashtag is essential: it creates a bounded input stream and gives the moderation layer a clear target. Display hardware is straightforward — a laptop connected via HDMI to the venue’s main screen is sufficient. The one genuinely consequential technical decision is between live streaming and a buffered feed. Live streaming puts a tweet on screen within seconds of publication; it creates energy but leaves zero time for intervention. A buffer holds content for 60 to 90 seconds before displaying it, giving a moderator a real intervention window. The Antalya group had started with a live feed. From the second congress onward, we moved to a 75-second buffer. Every piece of content was reviewed by a human before appearing on screen. The energy loss was marginal; the brand exposure reduction was substantial.The moderation protocol is the actual design work of a social wall project, and it has three layers. The first is the automated filter: a blocklist of profanity, competitor brand names, and pre-identified sensitive terms that the system rejects before content reaches a human reviewer. This list is prepared before the event and updated in real time as new edge cases emerge. The second layer is moderator approval: content that clears the automated filter is reviewed by a person physically present in the event hall, in the same room as the screen, with a direct communication line to senior management. The third layer is the crisis authority chain: when the moderator encounters content they cannot confidently approve or reject alone, who do they call, what is the expected response time, and what is the default decision if no response arrives within two minutes. At the Antalya hotel group, this chain was undefined before the third congress. In two earlier events, the same moderator made two different judgment calls on comparable content — inconsistency that itself created a minor brand problem. Writing the chain down in a single page of protocol resolved it.Capacity planning is a dimension that event teams consistently underestimate. A social wall feed during quiet moments of a congress may generate three to five tweets per minute. During session breaks or following a high-energy keynote, that figure can reach 30 to 40 per minute. A single moderator working alone during a peak period will fall behind: the moderation queue builds faster than it is cleared, the screen stalls, and attendees notice the feed has frozen. Some will conclude the channel is broken; others will simply stop contributing. At the second congress in Antalya, the moderator — overwhelmed during a break — disabled the approval queue and switched to a live feed to keep the screen active. Forty minutes later, a problematic post appeared. The fix was straightforward: for events above 400 attendees, two moderators working in shifts with an explicit handover protocol. A small procedural adjustment; a significant operational safeguard.The management expectation conversation deserves its own section, because skipping it creates problems no amount of technical refinement can solve. Senior management typically wants two things from a social wall: visible audience engagement and positive brand association. These two goals are not always compatible simultaneously. Tight moderation produces a sparse, clean feed; attendees whose posts don’t appear disengage. Loose moderation produces a dense, active feed with elevated brand risk. This trade-off should be presented explicitly to decision-makers before the event, not explained apologetically during it. At the Antalya group, we had this conversation before the third congress. Management stated a clear preference: a clean feed over a busy one. The moderation protocol was configured accordingly — a slightly higher threshold for approval, fewer posts displayed per hour. After the congress, the responsible executive offered this assessment: ‘We saw fewer tweets on screen than expected, but every one we saw was exactly right.’ That is a realistic success criterion, and it is only achievable when the expectation is set in advance.If you are starting a social wall project for an upcoming event, the work to do before any software configuration is this. Define your three moderation layers: the automated blocklist, the moderator approval window, and the crisis authority chain. Document all three on a single page before the event day. Match moderator headcount to expected volume: one person can manage a feed for an event under 250 attendees; above 400, two moderators with a formal shift handover are not optional. Set and communicate the event hashtag at least two weeks in advance, ideally embedded in pre-event emails and printed materials. Attendees who arrive knowing the hashtag produce a more manageable, better-labeled stream than those who invent their own tags in the room. The technical setup takes a few days. The process design takes a few hours. The organizations that invest those hours are the ones whose social walls work in the full sense of the word — not just on the screen, but for the brand.
This article was originally published in Turkish by Gökhan MERCANOĞLU on January 20, 2011. The English edition has been reviewed and edited by the author.