Filtering only applies on-premises
Remote and hybrid staff spend most of the week unfiltered, on home Wi-Fi, airport hotspots, or cell tethering. There's no policy and no record of where they went.
A secure web gateway sits between your users and the internet and checks what they're reaching at the DNS layer, at the network layer, and inside the encrypted web request itself. AllTech deploys and runs Cloudflare Gateway as that layer, so the filtering follows the laptop home instead of stopping at the office firewall.
The content filter in most small-business networks lives inside the firewall at the office. It works well until the laptop leaves the building. Then the user is on home Wi-Fi, an airport hotspot, or a cell hotspot, with no filtering at all and no record of where they went. The gaps we find most often when we take over a network:
Remote and hybrid staff spend most of the week unfiltered, on home Wi-Fi, airport hotspots, or cell tethering. There's no policy and no record of where they went.
Blocking a domain at DNS is fast and cheap, but a user or a piece of malware that knows the IP address still connects. Cloudflare's own guidance is explicit: DNS policies need HTTP policies behind them to be a real control.
Almost all web traffic is HTTPS. Without decryption a filter sees only the hostname. It can't see the URL path, the file being downloaded, or the data being uploaded.
The appliance is generating records that no one has looked at since the day it was installed.
It has to be sized for peak load, patched, and replaced on a cycle you pay for no matter how the threats change.
Gateway is Cloudflare's secure web gateway, built into the Cloudflare One SASE platform. It inspects and filters traffic at three layers: DNS, network (Layer 4), and HTTP (Layer 7). That happens across Cloudflare's global network, at the edge location nearest the user, so you don't backhaul traffic to a single data center.
We deploy it, write the policy, tune it against real user complaints, and keep the configuration under version control alongside the rest of your Cloudflare tenant. Which of the layers below you can use depends on the Cloudflare plan you're on, and we scope that up front, before the contract.
Every DNS query your users make gets checked before a connection is ever opened. Block malware, phishing, and command-and-control domains by security category, restrict content categories, and force SafeSearch on Google, Bing, YouTube, and DuckDuckGo. It's the fastest layer to deploy and the broadest in scope.
Layer 4 filtering on TCP, UDP, and GRE traffic by IP, port, protocol, and the server name (SNI) sent at the start of an encrypted connection. This is where we control non-web traffic like SSH, RDP, and anything else trying to leave on an unusual port.
The full request, decrypted and inspected: URL paths, headers, uploads, and downloads. Block a specific page rather than an entire domain. Scan downloads for malware. Quarantine a suspicious file for sandbox analysis before it reaches the endpoint. Allow your company Microsoft 365 tenant while blocking personal accounts on the same service.
Gateway inspects what's leaving as well as what's coming in. DLP profiles detect sensitive data in uploads, such as account numbers, records, source code, or custom patterns, and block the transfer before it completes.
Send internal hostnames to your own DNS servers while everything else resolves through Gateway. Private resources keep working for users on the client, and you don't have to split DNS management in two.
Most legacy filters make decisions based on a source IP address. Gateway policies can factor in the user and the device as well, pulled from the identity provider you already run.
Email address, group membership, and authentication method from Microsoft Entra ID, Okta, Google Workspace, or another IdP.
OS version, disk encryption status, firewall state, and whether the device serial number matches your managed device list.
You can combine them. A policy can allow a sensitive SaaS application only when the user is in the right security group and the device is encrypted, and block it on the personal laptop that isn't. That's the difference between a content filter and a control.
How you connect determines which policy layers can be enforced. We usually pair a device client for people with a network tunnel for locations.
DNS Yes · Network Yes · HTTP Yes
Best for: Roaming users on managed laptops and phones
DNS Yes · Network No · HTTP No
Best for: Unmanaged devices, guest networks, first-week rollouts
DNS No · Network No · HTTP Browser only
Best for: HTTP filtering without installing an agent
DNS Yes · Network Yes · HTTP Yes
Best for: Branch offices, plant floors, data centers
In practice, the client covers the people, the tunnel covers the building, and DNS resolver configuration gets filtering live on day one while the client rollout is still in progress. See Zero Trust Access (ZTNA) for the client and connector side of that.
We look at what's filtering today, what isn't, and where your traffic actually leaves from. You get a written policy recommendation and a category baseline before anything changes. It's often the right first step.
We start with DNS-layer filtering for immediate coverage, then add the device client, then HTTP inspection and the root certificate, then DLP. Every layer we turn on is one more thing that can break a line-of-business app, so we stage them, and we'd rather find those breaks with a pilot group than with your whole company.
Policy changes, category tuning, block-page exceptions, new-hire provisioning, and log review. Billed monthly per seat. When someone calls because a site they need is blocked, they call us, not you.
Point DNS at Gateway, block the security-risk categories, and watch the logs. On its own this stops a large share of malware and phishing domains. There's no software to install on any device and nothing to decrypt, and it's easy to reverse.
Deploy the Cloudflare One Client through Datto RMM or Intune. Filtering now follows the device off-network, and Layer 4 policies can control SSH, RDP, and unusual outbound ports. Split-tunnel and posture rules get set per role here.
Install the Cloudflare root certificate so HTTPS can be decrypted and inspected, then enable download scanning, file sandboxing, and DLP profiles. This phase is where we find the pinned-certificate app nobody remembered was installed, which is why it goes last and starts with a pilot group.
Policy changes propagate across Cloudflare's network in under a minute, so tuning during rollout is fast. The pace is set by how carefully we test, since the platform isn't the bottleneck.
One policy set, enforced the same way in Logan, Boise, and the office you open next quarter, with no appliance to ship to each one.
The point where "we have a firewall" stops being an answer, because half the workforce isn't behind it.
Plant-floor networks and office networks with different filtering needs, plus vendors who need narrow, logged access instead of open outbound.
Organizations that need to answer "where did that data go" with a record instead of a guess, and that have DLP obligations attached to specific record types.
If the renewal quote just landed and the box is out of support, this is the conversation to have before you re-up for another three years.
We'd rather say this here than after you've signed something.
If you only deploy DNS policies, a user or a piece of malware that knows the destination IP address can still connect. DNS is the fast, broad first layer. HTTP policies are what make it a control.
Decrypting HTTPS means installing the Cloudflare root certificate on managed endpoints. That's a real constraint on unmanaged or BYOD devices: those users get DNS-layer protection but not full inspection.
Certificate-pinned applications, some banking and healthcare portals, and certain updaters do not tolerate inspection. The fix is a Do Not Inspect policy for that destination, which means that destination isn't inspected. That's a trade, and we'll document each one.
Gateway filters what is in transit. It does not stop a threat already resident on a device, and it is not a mail filter.
Gateway produces DNS, HTTP, and network session analytics. We review them as part of managed service; if you're self-managing, budget the time or pair it with Managed SOC.
Packet filtering, for example, is an Enterprise-tier feature. We'll tell you which tier your requirements actually need before you buy one.
For the two gaps Gateway leaves open, see Endpoint Security andEmail Security. For log review,Managed SOC.
DNS filtering is one layer of it. A secure web gateway also inspects Layer 4 network traffic and decrypted HTTP requests, which is what catches the threats that live inside otherwise-allowed traffic.
Yes. Blocked requests return a block page, and the client can display a notification. We'd recommend telling staff before rollout — every deployment we've run goes smoother when the announcement precedes the block page.
Yes. Policies can key off user identity and group membership from Entra ID, Okta, or Google Workspace, plus device posture signals like disk encryption.
Traffic is processed at the Cloudflare edge location nearest the user, not backhauled to a central appliance. For most users on most requests the effect is not perceptible. Full HTTPS inspection does add processing, which is one reason we scope which traffic gets inspected rather than inspecting everything.
Your firewall protects the office. This protects the user. Gateway's layers map to familiar firewall functions — DNS filtering, a Layer 4 stateful firewall, and a Layer 7 forward proxy — except they're enforced wherever the user is instead of only at one perimeter.
Yes. Resolver policies send queries for your internal domains to your own DNS servers while everything else resolves through Gateway.
DNS-layer filtering can be live in days. A full rollout including the device client, HTTP inspection, and DLP typically runs four to eight weeks depending on headcount and how many applications need testing.
The client and connector that carry traffic to Gateway, and the per-app authorization layer for internal resources.
EDR on the device itself — the threat Gateway cannot see because it is already resident.
The other half of the phishing problem, filtered before the message reaches the inbox.
Humans reviewing what the tooling flags, including Gateway DNS and HTTP analytics.
The full Cloudflare One platform this sits inside — Access, Tunnel, DNS, and DDoS protection.
The edge device, segmentation, and multi-WAN failover Gateway rides on top of rather than replaces.
Send us a short note about your environment. We'll book a 30-minute scoping call, tell you which policy layers you need, and say so plainly if DNS filtering alone is enough for where you are right now.
Trusted by dozens of businesses






















