Hi everyone! 👋 I'm pretty new to the whole SASE/SSE world and we're looking at options for our office setup. We have about 100 users, mostly remote but a central office.
Has anyone here gone through a Cato Networks demo or POC for a similar size? I'd love to hear what that process was like. Were they able to show a realistic simulation for a network our size? Also, any gotchas or things a beginner like me should specifically ask about during the demo?
I'm still wrapping my head around how the policy engine works in practice. If you have a simple example of a firewall rule or how you set up a zero-trust application policy, that would be super helpful for me to understand.
Thanks in advance for any guidance! This community has been a huge help already.
Their demo is slick, almost too slick. They'll show you the policy engine with maybe five users in a test tenant. The real test is getting them to model your full traffic load and show you the invoice that comes with it.
For a 100-user office, the gotcha is the per-site commitment. Ask them point blank: what's the minimum term and what happens if you need to drop an office location? The zero-trust policy example they'll show you is the same one every vendor uses. Ask to see a more complex one, like segmenting contractor access from internal employees.
Be ready for sticker shock. Their pricing model loves to hide the bandwidth and support add-ons.
Your stack is too complicated.
Great point about the five-user test tenant. It's a classic move, isn't it? Everything looks effortless when you're only simulating a handful of connections. The real friction shows when you try to map their policy engine logic to your actual 100-user directory groups and specific app IDs.
You mentioned invoice modeling - that's crucial. I'd add, ask them to run that model *after* you've built a few of your own complex policies. The difference in quoted bandwidth between a basic web rule and a full zero-trust app policy can be massive, and that's where the hidden costs often live. Did your quote end up including their managed threat detection, or was that a separate line item?
Yeah, the per-site commitment worry is real. I'm trying to understand how that works if we later switch a bunch of remote users to a hub model instead.
When they model the full traffic load, did they give you a breakdown of what *kind* of traffic they were assuming? Like, was it mostly video calls and file sync? That feels like it would change the cost a lot.
Hey, welcome! I went through a demo with them last quarter for a team our size.
You're right to focus on how the policy engine works. A simple example that clicked for me: you can create a rule that says "Allow Sales group to access Salesforce app-ID 67890 from managed devices only." That's your zero-trust building block. They'll likely show you this.
But definitely ask them to replicate one of your actual, messy use cases live. Like, "Show me how to give temporary access for a contractor in Asana but not our financial software." That's when you'll see if the logic holds up or gets clunky.
The demo was impressive, but echo what others said about pushing for a real traffic model. Our simulated "light" traffic looked nothing like real-world Zoom and cloud backup streams.
edge cases matter
The demo for a 100-user setup will show you clean policy logic. The disconnect comes when you try to apply that to your actual, messy Active Directory structure and application mix.
Push them hard on the SLA details for your specific size. Ask for the exact remediation steps and service credits if their cloud points of presence underperform for your region. That's more telling than any policy walkthrough.
Forget the simple "Sales to Salesforce" rule. Make them build a policy during the demo that shows conditional access based on device posture *and* location for one of your real teams. If they can't map that logic clearly on the spot, the tool won't get easier post-sale.
SLA is not a suggestion.
Spot on about the five-user demo. It's a rehearsed performance. When they model your traffic, demand they use a packet capture from your actual edge device for a typical week. Their default "assumed mix" is always undersized.
The invoice shock is real, but it's not just bandwidth. The real sleeper is the policy count. Those complex rules for contractors everyone wants? Each one can be a separate charge once you're out of the starter bundle. Ask them to price the exact scenario you build in the demo.
And on dropping a location - get the clawback clause in writing before you sign. It's often brutal.
You nailed the sleeper cost with policy count. The "advanced policy modules" or whatever they brand them are almost always licensed separately. A policy that uses geo-IP, device posture, *and* time-based rules can easily be three line items.
The "rehearsed performance" part is key. The moment you ask them to deviate from the demo script to model a clawback, the tone shifts. The sales engineer will almost always have to "loop in their finance contact" which tells you everything about how punitive those terms are. Get that separate meeting booked before you invest any more time.
Been there, migrated that
Everyone's piling on about the five-user demo, which is fair, but the bigger issue for your size is the user-based vs. site-based licensing trap. For 100 mostly remote users, ask them to price it two ways: all users licensed individually versus that central office as a "site". The math gets weird fast, and the support tiers change.
The policy engine example they'll give you is useless. Tell them to skip the "Sales to Salesforce" spiel. Instead, give them a real, messy source like "All Azure AD synced users in the 'Contractors-US' group" and make them build the zero-trust rule to block access to your finance app but allow project management. If they can't map your actual directory groups to app IDs in three clicks, the tool will fight you daily.
And for the love of god, get the traffic model in writing *after* you define what "zero-trust" means for you. Their default assumption is basically web browsing, not the 8-hour Zoom calls and massive cloud backups your users actually run. That delta is the real cost.
been there, migrated that
The licensing trap is a critical observation. When we evaluated for a similar size, the site-based quote came with a much higher mandatory support tier than the user-based one, which they didn't initially disclose.
Your point about defining "zero-trust" for the traffic model is spot on. We found their baseline assumed a 70/30 split of web vs. non-web traffic. If your team uses heavy SaaS apps like Figma or continuous data pipelines, you need to explicitly state that. The cost multiplier for 'optimized' vs. 'standard' traffic for those streams was over 2x.
Ask for the price book entries for those advanced policy modules by name. Geo-IP and device posture are often separate SKUs, and combining them creates a stack.
CloudCostHawk
Good catch on the support tier being bundled and hidden. That's a common margin protector.
The 70/30 default traffic split is a fantasy for any modern office. Demand they run the model on a Friday afternoon packet capture, which will be 50% video and cloud storage sync.
Those separate SKUs for Geo-IP and posture mean a three-condition policy costs 3x. Always get the line-item quote for the exact policy you build in the demo.
cost per transaction is the only metric
Spot on about the finance contact loop. That's the moment the sales script ends. You can use that moment as leverage, actually.
If they need to "check with finance" on clawback, immediately pause the whole discussion. Get that finance person on a separate call before you do another demo minute. That's the only way to get the real terms, because the sales engineer's incentives are completely misaligned with exposing how punitive the exit costs are.
Also, those three separate line items for a single policy mean your operational cost scales with complexity, which is the opposite of what you're buying the tool for. Ask them to show the billing page in the admin portal where those module charges are broken out. If they can't, you're looking at invoice surprises every quarter.
Been there, migrated that
That 70/30 split point is critical. When they ran our model, they didn't account for video calls and cloud backups as "web" traffic, so it fell into the expensive optimized category. We had to argue it was standard web traffic to avoid the multiplier.
Is the support tier difference documented anywhere, or is it purely a sales negotiation thing? It feels like a hidden variable they use to balance the quotes.