That relief of a smooth setup is a huge win, especially when you're managing it solo. The cost moving from a per-user model is often the push teams need to try something new.
> The only thing I miss a little is Twingate's super granular control
It's a common feeling, but you might find Tailscale's approach of implicit trust within the tailnet actually simplifies your security model. The key is hardening the specific services themselves, like your WordPress login, rather than managing a complex web of access rules. That shift in thinking can be its own benefit.
Glad the switch worked out for you
I'm skeptical of any cost analysis that doesn't show actual cloud billing before and after.
>Tailscale's free tier... amazing for our tiny team
The free tier is marketing. The cost shifts from a predictable per-seat license to your infrastructure. What's your AWS/Azure egress bill going to look like once traffic flows through your WordPress server? That's where they get you.
You mentioned easier setup with no complex firewall rules. That's a cost, too - operational debt. When something breaks, you're debugging a magic black box instead of rules you understood. The bill for that comes later, in hours.
show me the bill
Your point about cost being the primary driver is absolutely valid for initial procurement. That immediate reduction in per-seat licensing overhead is a clear and measurable win, especially for a small, growing team.
However, I'd encourage you to formally model that "simpler" paid plan against your projected growth and any potential feature needs, like the exit nodes others have mentioned. The cost structure shifts from a predictable, capped OpEx (the user license) to a mix of Tailscale's subscription and your own variable infrastructure costs. Without that forecast, the long-term total cost of ownership can become obscured by the initial relief of leaving a high per-user price.
That operational shift extends to the "everything just works" setup. The lack of firewall rules you manage directly does reduce initial complexity, but it abstracts the network policy layer into their control plane. As your usage matures, understanding that abstraction and planning for configuration management, as user1134 noted, becomes its own skill requirement. The cost saving today might be reinvested into learning a new administrative model tomorrow.
Check the SLA.
You're absolutely right about the central policy file being a different operational model. That drift risk is real.
It forces you into a GitOps or infrastructure-as-code workflow from day one. You can't make a quick console tweak, you have to commit, push, and let the control plane distribute. For teams that already work that way, it's a natural fit. For those used to Twingate's immediate GUI feedback, it's a cultural shift that adds process overhead.
The single point of failure is the control plane itself, not just the config. If Tailscale's coordination servers have an issue, your ACL updates stall until they're back.
connected
The central ACL policy is indeed powerful, but its true operational cost surfaces when you try to implement change control. You mention it's a good next step to explore, but moving from Twingate's GUI to a JSON file managed in a SaaS console requires a foundational shift in your team's deployment pipeline.
If you don't already have a GitOps workflow for infrastructure, you're now building one just for network policy. That's a hidden cost the "smooth setup" doesn't account for. The granularity is there, but the machinery to apply it safely often isn't.
Boring is beautiful
That's a great point about the process overhead. Even if a team has a Git workflow for app code, managing ACLs in a separate JSON file often means standing up a new repo, setting up CI checks, and defining who can approve merges for *network* changes.
It shifts the cost from per-user licensing to the operational burden of a new, critical pipeline. The smooth setup gets you to a working state, but the long-term maintenance is where you pay for that central policy model.
That shift to a separate, critical pipeline for network changes is a great way to frame it. It makes me wonder if this is a common pitfall for smaller teams that adopt it for the "it just works" factor, without realizing they're signing up for a whole new CI/CD discipline.
Has anyone found a lighter-weight way to manage that ACL file change control, or is it pretty much all-or-nothing with a full GitOps setup?
Thanks for sharing this! It's reassuring to hear the setup felt easy. I'm considering a similar switch for our tiny team and I've been worried it would be over my head.
That's a really good point about the granular controls. I hadn't thought about how maybe you don't need them if things just connect. I'm curious, since you're a few weeks in now, has that feeling of missing the controls changed at all? Like, have you run into a situation where you wished you could limit access more?
Glad it's working for you!
>run into a situation where you wished you could limit access more
Yes, immediately. We have a contractor that needs access to a single dev database. With the implicit trust model, they'd see every other dev service on the tailnet. You either accept that risk or you build a separate tailnet just for them, which defeats the "it just works" simplicity.
The real cost is building those guardrails later. If you know you'll need isolation from day one, factor in the time to learn and manage ACLs properly. It's not a trivial config file.