Most events need far more bandwidth than the venue offers and far less than a panicked organizer buys. The working formula: 0.5 to 1 Mbps per simultaneous device for attendees, with 2 to 3 devices per person, plus separate dedicated capacity for production systems. A 500-person conference lands around 300 to 500 Mbps for the audience, with registration, streaming and payments on isolated networks. The number matters less than the architecture: segmentation keeps events online.
Why “how much bandwidth” is the wrong first question
Bandwidth is the headline number everyone asks about, but connectivity failures at events are rarely caused by a thin pipe. They are caused by architecture: hundreds of devices fighting for the same access points, critical systems sharing a network with attendees, or a single uplink with no backup.
A quick definition before going further: bandwidth is the total data capacity of your connection, measured in megabits per second (Mbps). Concurrency is how many devices are actively using it at the same moment. Events fail on concurrency more often than on raw capacity, because 1,000 phones associating with three consumer-grade access points will collapse the network long before the internet line is saturated.
So the right first question is not “how many Mbps” but “which systems need connectivity, how critical is each one, and can they be separated?”
The four networks every professional event actually runs
Treat your event as four separate networks with different requirements. On large productions these are physically distinct; on smaller ones they are at least logically segmented with reserved capacity.
- Production and show control. Comms, show calling, lighting and audio networking, timing systems. Low bandwidth, zero tolerance for interruption. Always wired where possible.
- Critical operations. Registration, access control and accreditation, payments and point of sale. Moderate bandwidth, high criticality. Must survive an internet outage through offline validation.
- Streaming and content. The live broadcast, photo uploads, press filing. High sustained upload, needs guaranteed throughput and a backup path.
- Attendee WiFi. The public network. High concurrency, bursty demand, lowest criticality. This is the only network allowed to degrade under pressure.
The most expensive mistake we see international teams make is collapsing all four into one network because “the venue has good WiFi.” The venue’s WiFi was designed for a normal Tuesday, not for your load-in day.
A sizing table you can actually use
These are planning ranges, not quotes. The real number depends on your program, your venue’s existing infrastructure and how much of the demand you can push to wired connections.
| Event type | Attendees | Attendee WiFi | Dedicated ops/streaming | Notes |
|---|---|---|---|---|
| Board meeting / executive session | Up to 100 | 50 to 100 Mbps | 20 to 50 Mbps | Venue line often usable, tested in advance |
| Conference or congress | 300 to 1,500 | 300 to 800 Mbps | 100 to 200 Mbps | Session rooms need their own access points |
| Trade show / expo | 2,000 to 10,000 | 1 Gbps or more | 200 to 500 Mbps | Exhibitors buy dedicated drops separately |
| Festival or mass event | 10,000 and up | Often none for public | 200 to 500 Mbps ops only | Public relies on carrier networks, ops fully isolated |
Two patterns worth noting. First, at mass events, public WiFi is often deliberately not offered: the mobile carriers bring portable cell sites, and the production keeps its private networks for operations. Second, at trade shows, exhibitor connectivity is a separate commercial product; never let exhibitor demand ride on the attendee network.
What drives your number up or down
- Simultaneity, not headcount. 1,000 attendees spread across a venue all day is a different problem from 1,000 people in one keynote room told to “open the app now.”
- Program moments. Interactive polls, mass downloads, a QR on screen that everyone scans at once. These create spikes of 5 to 10 times baseline demand. Plan for the spike, not the average.
- Streaming direction. Download-heavy use (attendees watching content) is manageable. Upload-heavy use (attendees or press pushing video out) is where lines choke, because most connections are asymmetric.
- Venue construction. Concrete, steel and basements attenuate signal. A site survey tells you how many access points you need and where; guessing tells you nothing.
- Density per access point. Enterprise-grade access points handle 50 to 100 active devices well. Consumer hardware claims similar numbers and delivers a fraction of them under real load.
Design for failure, because failure will visit
Power and connectivity are the invisible infrastructure of any production, and both deserve a contingency plan of their own. We covered the wider discipline in our guide to power and connectivity management at mass events, but for WiFi specifically, three principles carry most of the weight:
Critical systems validate offline. Registration and access control must keep working with zero internet: local validation, local storage, sync when the line returns. If your accreditation vendor cannot demonstrate this, the venue’s router is now a single point of failure for your doors. We wrote a full breakdown of access control without internet because it is the difference between an invisible incident and a gate crisis.
Two paths for anything that streams. Primary wired line plus bonded cellular backup, with automatic failover tested before doors, not during the keynote.
Someone owns the network during the show. Not “the venue’s IT guy is around somewhere.” A named person on comms, watching dashboards, empowered to shed load from the attendee network the moment operations need headroom.
Questions to send your venue before you sign
A venue that answers these quickly and precisely is telling you something good about itself:
- What is the total dedicated bandwidth available, and is it shared with other halls or tenants?
- Can you provision a separate VLAN or physically separate line for production use?
- How many access points cover our specific rooms, and what hardware are they?
- When was the last load test, and at what device count?
- What is the escalation path if connectivity degrades mid-event, and what response time is guaranteed in writing?
If the answers are vague, budget for bringing your own connectivity. It is a known, controllable cost. A network collapse in front of your audience is neither.
Where a production partner changes the equation
Connectivity looks like an IT line item until the day it fails, and then it is suddenly the whole event. A production partner who has run accreditation and operations across venues of every size arrives knowing which questions to ask, which venues overstate their infrastructure, and how to design networks that degrade gracefully instead of collapsing. Having accredited more than 150,000 attendees, much of it in venues where connectivity could never be taken for granted, we design for the outage first and the demo second.
If you are planning an event and want the connectivity, registration and access architecture designed by people who assume the WiFi will fail, tell us about your event and we will map what your production actually needs, network by network.