La pile technologique du WiFi invité : Un guide d'achat pour les marques multisites

The Guest WiFi Tech Stack: A Buyer Guide for Multi-Site Brands. A Purple Enterprise Briefing. Introduction and Context. Welcome. If you're responsible for network infrastructure across multiple sites — whether that's a hotel group, a retail estate, a stadium, or a public-sector estate — this briefing is for you. Guest WiFi has quietly become one of the most strategically important pieces of technology a venue operator can deploy. Not because it keeps visitors connected, though it does that too, but because it sits at the intersection of network operations, data compliance, marketing intelligence, and customer experience. Get it right, and it becomes a competitive asset. Get it wrong, and you're managing a fragmented mess of vendors, data silos, and compliance risk across every site. In this briefing, we're going to walk through every layer of the modern guest WiFi tech stack — from access points at the edge all the way up to CRM integration and analytics. We'll talk about how to evaluate vendors at each layer, what integration really means in practice, and how to think about total cost of ownership when you're making a buying decision this quarter. Let's start with the architecture. Technical Deep-Dive. The guest WiFi tech stack has six distinct layers, and most IT buyers make the mistake of evaluating them in isolation. That's where the complexity and the cost creep in. Layer one is your radio frequency infrastructure — the access points themselves. This is where most procurement conversations start, and it's the layer where brand loyalty runs deepest. Cisco Meraki, Aruba, Ruckus, Ubiquiti, Extreme Networks — these are the names you'll hear most often in enterprise deployments. The key evaluation criteria here are not just throughput and coverage. For multi-site deployments, you need to think about centralised management, zero-touch provisioning, and how the AP vendor's controller integrates with the layers above it. Wi-Fi 6 and Wi-Fi 6E are now the baseline for any new deployment. If you're still speccing Wi-Fi 5 for a new venue, you're already behind. WPA3 support is non-negotiable for any deployment touching payment zones or sensitive data. Layer two is your network controller and, increasingly, your SD-WAN fabric. This is the orchestration layer — it's where you segment your guest network from your corporate network, manage QoS policies, and handle failover across sites. The shift to SD-WAN has been significant for multi-site operators. Rather than managing individual MPLS circuits and site-by-site configurations, SD-WAN gives you centralised policy management with local breakout. For guest WiFi specifically, this means you can enforce bandwidth caps, content filtering, and VLAN segmentation from a single pane of glass. Layer three is authentication — specifically RADIUS and the broader AAA framework. Authentication, Authorisation, and Accounting. This is the layer that most guest WiFi deployments get wrong, or more accurately, get lazy about. The default approach — a simple pre-shared key or an open network with a splash page — is not appropriate for any venue handling personal data or operating under PCI DSS scope. IEEE 802.1X with a proper RADIUS backend gives you per-user authentication, session accounting, and the ability to enforce role-based access policies. For guest environments, this often means a cloud-hosted RADIUS service that integrates with your captive portal. FreeRADIUS is the open-source option, but for multi-site deployments at scale, a managed RADIUS service removes a significant operational burden. Layer four is the captive portal and splash page — the guest-facing authentication experience. This is where your brand lives in the network journey. A well-designed captive portal does three things: it authenticates the user, it captures consent under GDPR or your applicable data protection regulation, and it collects first-party data — name, email, demographic information, marketing preferences. The technical implementation matters here. A poorly built captive portal that relies on DNS hijacking without HTTPS support will break on modern iOS and Android devices. You need a portal that handles Apple's Captive Network Assistant correctly, supports social login via OAuth 2.0, and generates a compliant consent record that you can produce in the event of a regulatory audit. Layer five is your analytics and data platform. This is where the strategic value of guest WiFi is realised. Presence analytics — dwell time, footfall patterns, repeat visit rates — give venue operators intelligence that was previously only available through expensive sensor deployments or manual counting. But the real value comes from identity resolution: connecting an anonymous device MAC address to a known customer profile at the point of authentication. Once you have that link, you can measure marketing attribution, segment your audience by visit behaviour, and feed that data into your broader customer data platform. The key technical requirement here is a data model that is both privacy-compliant and portable. You need to own your data, not have it locked in a vendor's proprietary analytics silo. Layer six is CRM and marketing integration — the layer that converts network intelligence into business outcomes. This means bi-directional API integration with platforms like Salesforce, HubSpot, Mailchimp, or your own CDP. When a guest connects to your WiFi, that event should trigger a workflow: welcome email, loyalty points update, personalised offer. When a guest visits for the fifth time in a month, your CRM should know. The technical requirement is a robust webhook and API layer in your WiFi platform that can push events in near real-time and handle the data mapping between your network schema and your CRM schema. Implementation Recommendations and Common Pitfalls. Now let's talk about how to actually deploy this in practice, and where things typically go wrong. The first decision you need to make is build versus buy versus integrate. Building your own stack — stitching together an AP vendor, a RADIUS server, a custom captive portal, and a homegrown analytics pipeline — is technically feasible but operationally expensive. You're looking at a minimum of six months to first deployment, significant ongoing engineering resource, and a compliance posture that you're entirely responsible for maintaining. Best-of-breed integration — picking the best vendor at each layer and integrating them via APIs — is a common approach for large enterprises with mature IT teams. The integration complexity is real, though. Every vendor upgrade is a potential integration break. Data models diverge. Support tickets bounce between vendors. The third option is a unified platform that covers multiple layers in a single solution. The trade-off is flexibility versus simplicity. For most multi-site operators with lean IT teams, the unified platform approach delivers faster time to value and lower total cost of ownership over a three-year horizon. The second major pitfall is compliance architecture. GDPR, and in the US CCPA, place specific obligations on how you collect, store, and process personal data captured through guest WiFi. The consent record generated at the captive portal must be granular — separate consent for network access, marketing communications, and data sharing with third parties. Your data retention policies must be enforced at the platform level, not just documented in a policy. And your data processing agreements with every vendor in your stack must be current. This is an area where a fragmented stack creates real risk — each vendor is a separate data processor, each with their own DPA, each with their own breach notification timeline. The third pitfall is AP vendor lock-in at the captive portal layer. Many AP vendors offer their own captive portal solution, and it's tempting to use it because it's already integrated. The problem is that these native portals are typically limited in their data capture capabilities, their GDPR tooling, and their integration options. Separating your captive portal from your AP vendor — using a platform that integrates with multiple AP vendors via standard protocols — gives you the flexibility to change your radio infrastructure without losing your guest data history or your portal configuration. Rapid-Fire Questions and Answers. Let's run through some of the questions I hear most often from IT buyers. Question: Do we need Wi-Fi 6E, or is Wi-Fi 6 sufficient? For most venue deployments today, Wi-Fi 6 is sufficient. Wi-Fi 6E adds the 6 GHz band, which is valuable in very high-density environments like stadiums or large conference centres where spectrum congestion is a real constraint. If you're deploying in a venue with more than 500 concurrent users in a confined space, spec Wi-Fi 6E. Otherwise, Wi-Fi 6 gives you the throughput and latency improvements you need. Question: How do we handle MAC address randomisation and its impact on analytics? This is a real challenge. iOS 14 and Android 10 onwards randomise MAC addresses by default, which breaks device-based analytics. The solution is to shift your identity anchor from MAC address to authenticated user identity. When a guest authenticates through your captive portal, you bind their device session to their profile. From that point, analytics are identity-based, not device-based. This is actually a better data model — it's more accurate and more compliant. Question: What is the right SSID architecture for a multi-site deployment? The standard recommendation is three SSIDs per site: one for corporate devices on 802.1X, one for guest devices on the captive portal flow, and one for IoT devices on an isolated VLAN. Keep your guest SSID on a separate VLAN with no route to your corporate network. Use a firewall policy to restrict guest traffic to internet-only. This is the baseline. For venues with PCI DSS scope — hotels with in-room payment systems, for example — you need additional segmentation and a formal network diagram that your QSA can review. Summary and Next Steps. To bring this together: the guest WiFi tech stack is a six-layer architecture, and the buying decision is fundamentally about how much integration complexity you want to own. For most multi-site operators, a unified platform that covers the captive portal, analytics, and CRM integration layers — sitting on top of your existing AP infrastructure — is the fastest path to value and the lowest operational risk. The three things to prioritise in your evaluation are: first, data ownership — make sure you can export your guest data in a portable format at any time. Second, compliance architecture — your platform should generate audit-ready consent records and enforce data retention automatically. Third, AP vendor compatibility — your portal and analytics platform should be hardware-agnostic, supporting the major AP vendors via standard integration protocols. If you're at the evaluation stage, Purple's platform covers layers four through six of the stack — captive portal, analytics, and CRM integration — and integrates with over 90 access point vendors. It's worth a technical demo to see how it maps to your specific estate. Thanks for listening. The full written guide, including architecture diagrams, vendor comparison tables, and worked deployment examples, is available at purple dot ai. End of briefing.