Skip to content

Decision guideDNS filtering alternatives

Pi-hole vs managed DNS filtering for a venue

Decide whether Pi-hole or a managed DNS filtering service such as Purple Shield fits your venue. You can weigh licence cost against patching, redundancy and on-call labour, and match per-SSID and per-VLAN policy needs to each option. You will also know how to model a multi-site estate and plan a no-downtime 30-day...

13 min read

Watch the explainer

Use Pi-hole when you run one site, need only ad and tracker blocking, and have someone to patch it. Use managed DNS filtering such as Purple Shield when you run several venues or need threat intelligence. Shield sets one policy per network and venue from one dashboard. On Purple's data, its ad blocking cuts DNS queries by up to 62%.

When should a venue use Pi-hole, and when managed DNS filtering?

Placement decides what either option can do. DNS (Domain Name System) filtering acts when a device asks for an address, before any connection opens. A guest's phone asks the resolver where a phishing domain lives. The filter refuses to answer, so no traffic reaches that host. A firewall acts one step later, on IP addresses and ports, after the lookup has succeeded. That is why protective DNS catches what a firewall misses. Examples include fast-changing phishing hosts, malware domains and trackers on shared delivery networks.

Pi-hole and managed services filter at exactly the same point. The difference is where the resolver lives and who keeps it running. Pi-hole is free, open-source software you install on a small server inside each venue. A managed service such as Purple Shield, DNSFilter or Cisco Umbrella runs its resolvers in the cloud. You change the DNS server your DHCP (Dynamic Host Configuration Protocol) scope hands out, then manage policy centrally.

The rule of thumb follows from that split. Pi-hole costs nothing to license, but every site adds boxes to patch, back up and keep alive. A managed service costs a subscription, yet the 30th venue takes no more effort than the first. For one site with a named owner, Pi-hole is a sound choice. For an estate, or any network with regulatory weight, managed filtering is the cheaper default once you count staff time. Our overview of DNS filtering alternatives covers the wider category.

Listed or riskyCleanGuest joins WiFiDHCP hands out thefiltering resolverDNS queryDevice asks for adomain's addressPolicy check atthe resolverPi-hole on site or amanaged cloud resolverBlockedNo address returned, noconnection openedAllowedAddress returned to thedeviceConnection andfirewallFirewall acts on IP andport after the lookup
Where DNS filtering sits in a guest connection

A guest device sends a DNS query to the filtering resolver, which blocks a listed domain or returns an address before the firewall sees any connection.

Connections

  1. Guest joins WiFi to DNS query
  2. DNS query to Policy check at the resolver
  3. Policy check at the resolver to Blocked (Listed or risky)
  4. Policy check at the resolver to Allowed (Clean)
  5. Allowed to Connection and firewall

How do you choose between Pi-hole and managed DNS filtering?

Four questions settle most venues. Two terms recur. An SSID is the network name a guest sees. A VLAN (virtual LAN) is a logically separate network on shared switches, which keeps guest, staff and tenant traffic apart. If different SSIDs or VLANs need different rules, the answer moves sharply towards a managed service.

Pi-hole or managed DNS filtering for your venue

See every path as a list
Will the filter cover more than one venue?
  • Yes: Do different networks need different policies?
    • Yes: Purple Shield with per-network policy Set a policy per SSID, VLAN and venue from one dashboard, standalone or alongside Purple Access. DNS filtering for multi-site estates
    • No: Do you have central IT staff to run resolvers at every site?
      • Yes: Model the fleet cost, then trial managed A Pi-hole fleet is possible. Count instances, patch cycles and on-call time before you commit. Run a 30-day trial
      • No: Managed DNS filtering Cloud resolvers, maintained threat categories and no box on site. Shortlist Purple Shield, DNSFilter or Cisco Umbrella. Cisco Umbrella pricing explained
  • No, one site: Do you need threat protection or content categories, not just ad blocking?
    • Yes: Managed DNS filtering Cloud resolvers, maintained threat categories and no box on site. Shortlist Purple Shield, DNSFilter or Cisco Umbrella. Cisco Umbrella pricing explained
    • No: Does a named person own patching, backups and blocklist fixes?
      • Yes: Pi-hole fits One site, ad and tracker blocking only, and a named owner. Run two instances so one failure does not take guest WiFi offline. What a venue gives up with free filtering
      • No: Managed DNS filtering Cloud resolvers, maintained threat categories and no box on site. Shortlist Purple Shield, DNSFilter or Cisco Umbrella. Cisco Umbrella pricing explained

Which option fits each type of venue?

Venue type is a useful proxy for the conditions in the chart. It signals how many networks you run, who is on call and how much regulatory weight the WiFi carries. Read the "why" column for the condition behind each call.

Pi-hole or managed DNS filtering by venue scenario Data as of 2026-06-10
ScenarioRecommendationWhy
Single café, owner-run, guests complain about adsPi-hole, with a second instanceAd and tracker blocking is the whole brief and the owner can replace a box
120-room hotel with guest and staff SSIDsManaged (Shield)Guest and staff need separate policies; hotel IT rarely has time to curate lists
Retail estate of 30+ storesManaged (Shield)One policy set for every store instead of 60 resolvers to patch
Stadium or events venue with match-day peaksManaged (Shield)Fan and attendee query peaks land on cloud resolvers, not one box in a comms room
Airport or rail hubManaged (Shield)Passenger, staff and tenant networks each need their own policy
Public library or council WiFiManaged (Shield)Category filtering and central reports support duty-of-care evidence
NHS trust patient and visitor WiFiManaged (Shield)Threat protection with central reporting; St George's Healthcare NHS Trust runs Shield
IT lab or test networkPi-holeFree, fully local and a practical way to learn DNS filtering
Recommendations assume no dedicated on-site IT. A named owner at a single site moves the answer towards Pi-hole.

Two patterns stand out. Pi-hole wins only where the brief is ad blocking and one person owns the box. Every scenario with separate audiences needs a policy per network. Pi-hole can approximate that with group management, but only per instance and only by client address.

Pi-hole for venue WiFi: what you get and what you run

What Pi-hole does well

Pi-hole is a DNS sinkhole built for network-wide ad blocking. It answers lookups for blocklisted domains with no usable address, so ads and trackers never load. It runs on Linux, in a container or on a Raspberry Pi. Its web interface shows every query in real time. It can validate DNSSEC, forward to an upstream resolver you choose, and act as the DHCP server.

For a single venue, that covers a lot. Query logs stay on your own hardware, which some privacy teams prefer. Group management applies different blocklists to different clients. Teleporter, the built-in backup tool, exports settings so you can rebuild a failed box quickly.

Where Pi-hole strains in a venue

The limits appear when Pi-hole moves from a home network to a guest network that hundreds of strangers rely on.

  • A single point of failure. If the only resolver stops, guests lose name resolution and report the WiFi as broken.
  • Lists, not categories. Pi-hole blocks what its lists contain. A fresh phishing host passes until a list maintainer adds it.
  • No block page. A blocked domain returns no usable address, so the guest sees a browser error. Front-of-house staff field the question.
  • One instance per site. There is no built-in console for many instances. Community sync tools exist, but you maintain those too.
  • Community support only. Help comes from forums and documentation, not a support desk with a response time.

Managed DNS filtering: Purple Shield, DNSFilter and Cisco Umbrella

How managed filtering works

A managed protective DNS service runs the resolvers for you. You point each network's DNS at the service. Every lookup is then checked against maintained category and threat feeds. The provider classifies new domains and patches the infrastructure. You set policy in one console and read reports across every site.

Because the change happens in DHCP, managed filtering is hardware-agnostic. It works on Cisco Meraki, HPE Aruba, Ruckus, Juniper Mist, Ubiquiti UniFi, Cambium, Extreme and Fortinet networks without new equipment. Follow your vendor's documentation for the exact DHCP setting.

How the three services differ

DNSFilter and Cisco Umbrella are general-purpose protective DNS services. Both sell to corporate IT teams and managed service providers, and both document customisable block pages. Umbrella sits inside Cisco's wider security portfolio, so licensing follows Cisco's packaging. Our guide to Cisco Umbrella pricing explains what to model before you request a quote.

Purple Shield is built for venue WiFi. You can run it standalone on any WiFi network, or alongside Purple Access. Either way, you set a policy per SSID, VLAN and venue from one dashboard. That matches how venues are segmented: guest, staff, tenant and back-of-house networks rarely want the same rules. Purple runs across 80,000+ live venues in 90 countries, and Shield shares that platform.

Shield also blocks ads, which matters on capped or shared backhaul. On Purple's data, ad blocking cuts data by up to 44% and DNS queries by up to 62%. That headroom goes back to the content your guests came for.

What does a self-hosted DNS filter cost across sites?

The licence line is where Pi-hole looks unbeatable, and it is the only line where it is. The total cost of a self-hosted DNS filter is hardware plus labour, multiplied by instances.

Cost lines for a self-hosted vs managed DNS filter Data as of 2026-06-10
Cost linePi-hole (self-hosted)Purple Shield (managed)
Software licence£0, open sourceSubscription, see Shield pricing
Hardware per siteTwo small servers, VMs or Raspberry PisNone; resolvers run in the cloud
Setup per siteInstall, static address, DHCP change, blocklistsDHCP DNS change, policy set in the dashboard
PatchingOS and Pi-hole updates on every instanceRun by Purple
BlocklistsYou choose, curate and fix false positivesCategories and threat feeds maintained by Purple
Multi-site policySeparate admin interface per instanceOne dashboard, policy per SSID, VLAN and venue
Failure handlingYour team restores the box on siteNo on-site resolver to fail
ReportingPer-instance query log, retention set by youCentral reporting across venues
SupportCommunity forumPurple support team
At more than a handful of sites, labour rather than licence becomes the largest Pi-hole cost line.

A simple model for your estate

Build the model from instances, not sites. Two instances per site is the minimum that avoids the secondary DNS trap. So 30 sites means 60 resolvers, each with its own operating system and blocklist settings.

Then put a time figure on each recurring task. The figures below are planning assumptions to show the arithmetic; substitute your own.

  • Updates and health checks: assume 15 minutes per instance per month. At 60 instances, that is 15 hours a month.
  • Blocklist fixes: each allowlist change lands once per instance, unless you run a sync tool.
  • Incidents: a failed disk at a remote site needs someone on site.
  • Audit evidence: each instance produces its own report, and someone combines them.

Multiply the monthly hours by your loaded staff rate and add hardware refresh. Set that against a subscription from the Shield pricing page. At one site, the self-hosted total is small. At 30, labour is the line to watch, before you value threat categories Pi-hole never offered.

Worked scenario: a 30-store retail estate

A fashion retailer with 30 stores piloted Pi-hole on shopper WiFi in three branches. The pilot did its job: ads stopped loading on the shopper network. Scaling it meant 60 instances, a sync tool and an unbudgeted on-call rota. The IT team modelled routine upkeep at 15 hours a month using the assumptions above. It moved all 30 stores to managed DNS instead, with one policy for shoppers and one for staff. The outcome: 60 planned resolvers replaced by zero on-site boxes, and each policy change made once instead of 60 times.

What changes the answer?

The rule of thumb bends under five conditions. Each one can move a venue from one option to the other.

Number of sites and networks

One site with one guest network keeps Pi-hole viable. Every extra site adds two instances. Every extra SSID or VLAN with its own rules adds group configuration on each instance. Our guide to DNS filtering for MSPs and multi-site estates covers estates in depth.

Who owns it at 2am

Pi-hole depends on a named owner with time in their week. If ownership means "whoever set it up", often a contractor who has moved on, choose managed.

Ad blocking or threat protection

If the brief is fewer ads and lighter pages, Pi-hole's lists cover it. If it includes phishing, malware and newly registered domains, you need maintained threat feeds. That is what a protective DNS service provides.

Duty of care and regulation

Public WiFi in libraries, schools and healthcare carries a duty to filter harmful content. Schools in England, for example, follow the Department for Education's filtering and monitoring standards. Central category reports make that evidence easier to produce than per-site lists. For PCI DSS, DNS filtering does not replace segmentation of the cardholder data environment. It does reduce malware risk on the networks around it. Our enterprise WiFi security guide covers segmentation further.

Privacy and data protection

DNS query logs linked to a device can count as personal data under GDPR. With Pi-hole, you set retention and access on every box. With a managed service, you need a data processing agreement. Purple is ISO 27001 certified and GDPR compliant.

Worked scenario: a 120-room hotel

A 120-room city hotel ran one Pi-hole on a mini PC for guest WiFi. DHCP handed it out as the only DNS server. When its storage failed overnight, guests could join the WiFi but load nothing. The duty manager logged a run of complaints before the IT contractor arrived. The hotel then moved guest and staff SSIDs to managed DNS with separate policies. The outcome: zero resolvers on site, two policies set from one dashboard, and no local box left to fail.

For any estate of two or more venues, the default is managed protective DNS, and for venue WiFi we recommend Purple Shield. It filters at the resolver before a connection opens, and it sets a policy per SSID, VLAN and venue. It also removes every on-site resolver from your patch list. Pi-hole stays the right answer for a single site with a named owner and an ad-blocking brief, and for labs.

Pi-hole or managed DNS filtering: where each fits

Pi-hole (self-hosted)

Fits when

  • One site with a named owner
  • Ad and tracker blocking is the whole requirement
  • You want query logs kept on your own hardware
  • Lab, test or staff-only networks

Not when

  • Two or more venues
  • You need threat categories or a block page
  • Nobody is on call if the box fails

Managed DNS filtering (Purple Shield)

Fits when

  • Two or more venues
  • Guest, staff and tenant networks need separate policies
  • You must show auditors a policy per SSID, VLAN and venue
  • Bandwidth is capped and ad blocking pays back

Not when

  • A single site where free ad blocking is enough and someone owns the box

Our recommendation: Default to Purple Shield for any venue estate. Keep Pi-hole for a single site with a named owner, or for a lab.

Your first action is to pick one guest SSID at one venue and run a 30-day trial. Point its DNS at Shield and leave every other network untouched. Then compare blocked threats, data volume and support tickets. The 30-day DNS filtering trial plan sets out what to measure each week.

Key terms

DNS filtering (protective DNS)
Policy enforcement at the Domain Name System resolver, the lookup service defined in RFC 1034 and RFC 1035. The resolver refuses to return an address for listed or risky domains, so no connection opens to that host. Both Pi-hole and managed services filter at this same point. The decision is about where the resolver runs and who maintains it, not about filtering capability at the lookup.
DNS sinkhole
A resolver that answers queries for blocklisted domains with no usable address instead of the real record. Pi-hole works this way, matching each lookup against the lists it has loaded. A sinkhole blocks only what its lists contain, so a fresh phishing host passes until a maintainer adds it. Guests also see a browser error rather than a block page.
DHCP
Dynamic Host Configuration Protocol, specified in RFC 2131. It assigns each device an IP address and network settings, including the DNS servers the device should query. Changing the DNS server in a DHCP scope is how you deploy either option. This is why managed filtering needs no new hardware on Cisco Meraki, HPE Aruba, Ruckus, Juniper Mist, Ubiquiti UniFi, Cambium, Extreme or Fortinet networks.
SSID
Service Set Identifier, the network name a wireless access point advertises under IEEE 802.11. One access point can broadcast several SSIDs, each mapped to its own network segment. Venues run guest, staff and tenant SSIDs that rarely want the same rules. Purple Shield sets a policy per SSID, while Pi-hole approximates this only per instance and by client address.
VLAN
Virtual LAN, a logically separate network on shared switches, tagged under IEEE 802.1Q. Each VLAN carries its own traffic and can be assigned its own DHCP scope and DNS settings. VLANs keep guest, staff and back-of-house traffic apart. If different VLANs need different filtering rules, the decision moves sharply towards a managed service with per-VLAN policy.
DNSSEC
Domain Name System Security Extensions, defined in RFC 4033 to RFC 4035. It adds cryptographic signatures to DNS records so a validating resolver can confirm answers have not been forged. Pi-hole can validate DNSSEC on site. It is one of the capabilities that makes Pi-hole a sound choice for a single venue with a named owner.
DNS over HTTPS (DoH)
A protocol specified in RFC 8484 that sends DNS queries inside encrypted HTTPS connections to a chosen resolver, rather than to the resolver DHCP assigns. A device using DoH or a VPN skips your filtering resolver, whichever option you choose. Block known public DoH endpoints to keep filtering in place for most guests.
Canary domain
use-application-dns.net, a domain Mozilla documents for Firefox. If the local network's resolver blocks it, Firefox keeps using local DNS instead of enabling DoH by default. Blocking the canary domain is a simple control on guest WiFi. It protects the majority of guests who never change a browser setting.
Secondary DNS trap
The failure mode where DHCP hands out a filtering resolver plus a public resolver as backup. Devices query both, so blocked domains resolve through the unfiltered server. It is why a Pi-hole venue needs two Pi-hole instances per site. Count the second instance in hardware, patching and labour costs.
Personal data under GDPR
Article 4(1) of the General Data Protection Regulation defines personal data as information relating to an identifiable person, including online identifiers such as device addresses. DNS query logs linked to a device can count as personal data. With Pi-hole you set retention and access on every box; with a managed service you need a data processing agreement.
Cardholder data environment (PCI DSS)
The systems that store, process or transmit payment card data, as defined by the Payment Card Industry Data Security Standard. Segmentation isolates it from other networks to reduce scope. DNS filtering does not replace segmentation of the cardholder data environment. It does lower malware risk on the guest and staff networks around that boundary.

Worked examples

A fashion retailer with 30 stores piloted Pi-hole on shopper WiFi in three branches. Ads stopped loading, so the team considered rolling it out to every store.

Scaling the pilot meant 60 Pi-hole instances, two per store to avoid the secondary DNS trap, plus a sync tool and an unbudgeted on-call rota. The IT team modelled routine upkeep at 15 hours a month, using a planning assumption of 15 minutes per instance per month. Every allowlist change would also land 60 times. The retailer moved all 30 stores to managed DNS instead, with one policy for shoppers and one for staff. The outcome: 60 planned resolvers replaced by zero on-site boxes, and each policy change made once instead of 60 times.

A 120-room city hotel ran a single Pi-hole on a mini PC as the only DNS server for guest WiFi. Its storage failed overnight.

Guests could join the WiFi but load nothing, because no resolver answered their lookups. The duty manager logged a run of complaints before the IT contractor arrived on site. The failure exposed two gaps: no second instance and no owner on call at 2am. The hotel also needed separate rules for guests and staff, which Pi-hole handles only per instance and by client address. It moved guest and staff SSIDs to managed DNS with separate policies. The outcome: zero resolvers on site, two policies set from one dashboard, and no local box left to fail.

A venue running Pi-hole wants to move to Purple Shield without guests losing name resolution during the change.

Move one network at a time. First, export settings with Pi-hole's Teleporter tool and note the allowlist entries you still need. Recreate those exceptions in the new Shield policy so known false positives stay fixed. Then change the DNS server on one SSID's DHCP scope and watch it for a few days. Move the next network only once the first is stable. Keep Pi-hole running until the last network has moved, so any network not yet migrated keeps filtering. This sequence avoids downtime and gives you a clean point to roll back.

Frequently asked questions

Is Pi-hole free to use at a commercial venue?

Yes, the software is free and open source, and you can run it on guest WiFi. You still pay for hardware, two instances per site for redundancy, and staff time for patching, blocklist fixes and incidents. At one site, those costs are small. Across an estate, labour becomes the largest line. Total cost, not licence cost, should drive the decision.

Will Purple Shield work with the access points I already own?

Yes. Shield works at the DNS layer, so it needs a DHCP change rather than new hardware. It runs on networks built with Cisco Meraki, HPE Aruba, Ruckus, Juniper Mist, Ubiquiti UniFi, Cambium, Extreme and Fortinet. You can run it standalone on any WiFi network or alongside Purple Access, with a policy per SSID, VLAN and venue.

How does Shield pricing compare with running Pi-hole?

Shield is a subscription and Pi-hole has no licence fee, so compare total cost rather than list price. For Pi-hole, add hardware for two instances per site, monthly patching hours and on-call time. Set that against the subscription on the [Shield pricing page](product:shield/pricing). The comparison shifts towards managed with every site you add.

Can we migrate from Pi-hole without downtime?

Yes, if you move one network at a time. Export your settings with Pi-hole's Teleporter tool and note the allowlist entries you still need. Recreate those exceptions in the new policy, then change the DNS server on one SSID's DHCP scope. Watch it for a few days before moving the next network. Keep Pi-hole running until the last network moves.

Does DNS filtering help with GDPR and PCI DSS?

It helps, but it does not make you compliant on its own. For PCI DSS, segmentation keeps guest and staff WiFi out of the cardholder data environment. DNS filtering lowers malware risk around that boundary. For GDPR, DNS logs tied to a device can be personal data, so set retention and access rules. Purple is ISO 27001 certified and GDPR compliant.

How does Shield compare with Cisco Umbrella and DNSFilter?

All three filter at DNS resolution in the cloud, so the placement benefit is the same. DNSFilter and Cisco Umbrella are general-purpose services sold to corporate IT teams and managed service providers. Shield is built for venue WiFi, with a policy per SSID, VLAN and venue. It can run standalone or alongside Purple Access. Whitbread, AGS Airports and St George's Healthcare NHS Trust run Shield.

Can guests bypass DNS filtering?

Some can, whichever option you choose. A device using DNS over HTTPS ([RFC 8484](https://www.rfc-editor.org/rfc/rfc8484)) or a VPN sends lookups past your resolver. Block well-known public DoH endpoints. Also block Firefox's [canary domain](https://support.mozilla.org/en-US/kb/canary-domain-use-application-dnsnet), which tells the browser to keep using local DNS. On guest WiFi, the aim is to protect the majority who never change a setting.

Sources

Decide what your network reaches, free for 30 days

Tell us which networks and sites you run and who is on them. We map a policy to each one: staff, guests, students, residents and IoT. Then point one network at Shield and see what it stops for 30 days, before you pay anything. Running change control? We plan a pilot on one site with your team instead. No hardware to install.

  1. We map your networksA Purple expert maps your networks, your venues and the people on each, and answers your questions on bypass, outages, data handling and integrations.
  2. You get a policy for each oneStaff, guests, students, residents, IoT and paid tiers, each with the policy it needs.
  3. You prove it, then get your tailored quote30 days free on one network, or a pilot on one site planned with your team, then Shield priced standalone or as an add-on to your Access plan.

Your policy map

Tell us your networks. We map a policy to each one.

A Purple expert comes back with a policy for every network and group you run, from the team that has run content filtering since 2012 and runs WiFi for 80,000+ venues in 90 countries.