Hybrid work is a consistency problem, not a remote problem

Most organizations declared victory on remote work in 2020. Buy laptops, install VPN, hand out MFA tokens, done. Five years on, the work-from-anywhere reality is messier than the laptop-plus-VPN story admitted. The people on the receiving end of it know — every IT survey we run with customers, the top complaint isn't security or cost or even speed. It's inconsistency. Desktop A at the office. Desktop B at home. Desktop C on a hotel Wi-Fi. Three slightly different versions of the same job.

The metric that matters is consistency. Same files. Same apps. Same speed. Same shortcuts. Same browser sessions. Same identity. Three locations on the same Wednesday should feel like one workday, not three. When that's true, productivity is no longer location-dependent. When it isn't, the helpdesk eats the difference one ticket at a time.

Azure Virtual Desktop is what hybrid actually wants. AVD's whole architecture is built around exactly this problem: one desktop image, delivered identically to any client device, anywhere it connects from. The corporate environment lives in the cloud session host; the device on the user's lap is a thin window into it. Lose the laptop? The desktop is unaffected. Switch to a tablet in an airport lounge? Same desktop. Open a browser on a partner's machine? Same desktop. AVD is the right tech because hybrid is the question it was designed to answer.

FSLogix is where AVD goes from working to feeling fast. Profile containers, when designed right, give you sub-2-second logons and roaming Office data, OneDrive cache, search index, Outlook OST. When designed wrong, they give you 90-second logons, profile corruption, and a stream of 'why is my Outlook missing emails again' tickets. The difference is whether someone has thought hard about redirection rules, container sizing, exclusion patterns, and where the SMB storage actually lives. Most organizations don't have that someone in-house — and that's the gap most Ultiblob AVD engagements close in the first week.

Now for the part no AVD marketing page wants to talk about: the cloud has bad mornings. Entra ID has had multi-hour authentication outages in production. Specific Azure regions have lost the AVD control plane for windows long enough to make CFOs notice. The honest assessment is that any pure-cloud VDI estate is one outage from a stopped business, and the bigger the estate the more public the pain when that outage happens.

This is why we don't ship AVD-only. Every Ultiblob workforce platform pairs AVD with a traditional RDS deployment in our own private datacenter, running session hosts with a parallel identity provider that can authenticate users when Entra is unavailable. When the cloud is healthy — most days — everything runs through AVD because that's the better-cost-per-seat path. When the cloud has a bad morning, sessions fail over to RDS in under 90 seconds and your team keeps working. The CFO calls about why the helpdesk is calm during a Microsoft outage; the helpdesk calls because the failover ran cleanly and there's nothing to triage.

The cost question. Yes, RDS-as-resilience costs something. No, it doesn't double your spend. Session hosts sized for failover-only traffic are 20-30% of the AVD footprint by design. Most weeks they're parked. The ROI math is straightforward: if your team can't work for a full day during an Entra outage, what does that cost? Multiply by the frequency of major outages over a three-year horizon. The answer is almost always larger than the RDS reserve.

What we tell customers in scoping. Three questions. One: how many of your people change locations during a typical week? If the answer is more than a third, AVD is the right primary path because consistency-at-scale is the problem AVD is best at. Two: what's the cost of a full work day lost to a cloud authentication outage? If it's above six figures, traditional RDS resilience pays for itself within a year. Three: who owns FSLogix profile design today? If the answer is 'nobody' or 'a vendor we contracted for the initial deployment,' that's why your AVD feels slow.

Hybrid isn't a 2020 problem you can declare solved with VPN. It's the operating model for the next decade — and consistency, not connectivity, is the metric to optimise for. Get that right, and remote work stops being a productivity tax and starts being what it was always supposed to be: a tool that lets your team do their best work, wherever the work happens to be that day.

If you want to see the AVD + RDS hybrid stack running in 48 hours for your team, or want a second opinion on a deployment that already feels off, the scoping call is 15 minutes. We'll tell you what we'd change, and what we'd leave alone.