Offline Software Is Boring. That's Exactly Why Your Competitors Still Use It.

February 21, 2026

I sent my 6am motivational text to my client list from the parking garage on Tuesday. The wifi reaches the second level from the Panera next door. I've got it mapped. A man adapts. That's the kind of resourcefulness I want to talk about today, because the tech industry has spent the last decade telling businesses that if you're not on cloud-native, AI-integrated, always-on SaaS platforms, you're already dead. And I'm here to tell you that narrative is wrong, it's serving somebody else's financial interests, and the companies quietly ignoring it are the ones your salespeople keep losing to.

Here's the news that's been building into a real story: on-premise, offline-first software is not dying. It is, by almost every serious measure, holding its ground. Cloud computing has grown to the point that it is now the largest environment where workloads are being executed - but in 2025, it represented just 36% of all workloads, an equal share to on-premises workloads, according to Omdia chief analyst Roy Illsley. That's not the stat the cloud vendors are leading with in their pitch decks. But it's real.

And it gets more interesting. According to a 2024 IDC survey, 80% of IT decision-makers expect to repatriate some workloads from the cloud within the next 12 months. Repatriation. That's the word we're using now. Moving back. Not because these companies are technologically backward - because they ran the math and found out what they were actually paying for.

The Cloud's Dirty Secret Is the Bill

I'll be real with you. My personal finances are currently in a state that a therapist would call "a growth opportunity." Car's gone. Savings are a concept I respect but cannot access. So I am perhaps more attuned than most people to what it feels like when a recurring cost turns out to be higher than you budgeted for. The cloud does this to businesses constantly.

Companies like Basecamp reported spending $3.2 million on cloud services in a single year, only to find they could save significantly by moving back to on-premises solutions. That's not a niche horror story. The most common motivator for cloud repatriation is cost - more than 43% of IT leaders found that moving applications and data from on-premises to the cloud was more expensive than expected.

And here's what really gets me: about 32% of cloud budgets go to idle resources, and only 30% of companies can track costs accurately. A third of what you're spending is on nothing. At least with on-premise software, you know what you paid. It's sitting on a server in a room. You own it. That capital expenditure conversation is cleaner than watching an AWS bill balloon every month while your finance team tries to figure out which team provisioned what.

Cloud costs have a tendency to balloon over time: storage costs per GB might seem low, but when dealing with terabytes of data, the costs add up - and retrieval and egress fees only compound it. The cloud is elegant at the start. It's the long game that bites you.

62% of Organizations Still Run Legacy Software. That's Not a Bug.

The framing I keep seeing in the tech press is that legacy and on-premise software usage is a problem to be solved. A liability. Evidence of organizational dysfunction. And that framing is wrong in a very specific, financially motivated way.

62% of respondents in a 2025 survey of 500 U.S. IT professionals said their organization continues to rely on legacy software. The consultants who write about this number treat it like a slow-moving emergency. I think it's closer to proof of something.

The number one reason IT pros say they haven't upgraded? "The current system still works." Half of all respondents cited this as their primary blocker. The consulting class says that "working" doesn't mean optimal. And sure, technically. But a system that works, that your team knows deeply, that doesn't require six months of onboarding, that doesn't have a pricing tier that can change on a vendor's quarterly earnings call - that system has real economic value that doesn't show up in the ROI deck from the SaaS company trying to replace it.

Derek spent two weeks last month trying to get everyone onto a new project tool. Nobody used it. We went back to the spreadsheet. The spreadsheet is offline. The spreadsheet works. I'm not saying spreadsheets are the answer to enterprise software needs - I'm saying Derek's experience is not unusual, and the productivity cost of failed migrations is something nobody's lining item in their cloud transition budget.

The Industries That Never Left Were Right the Whole Time

Defense. Healthcare. Government. Finance. These sectors kept their critical workloads on-premise while the rest of the business world chased SaaS dashboards, and they were not being backwards. They were being correct.

On-premise installations persist in finance, defense, and healthcare where sovereignty or latency concerns prevail. That's not resistance to change. That's a rational response to the actual risk profile of their data. Fewer than 20% of organizations plan to move sensitive data to the cloud, preferring to keep it on-premises because of regulatory concerns.

And look at what SAP, one of the largest enterprise software companies on the planet, is actually doing with its install base. A large portion of SAP's install base is still running legacy SAP Business Suite / SAP ECC on-premise, and SAP pushed the end of mainstream maintenance for ECC 6.0 from 2025 all the way to 2027, with extended support available until 2030. SAP didn't do that because they felt nostalgic. They did it because their largest customers - the ones with real leverage - weren't moving. SAP announced a "private edition" transition option to extend support through 2031-2033 for select customers with very complex landscapes, effectively giving large enterprise customers on ECC or S/4HANA on-premise additional breathing room until 2033.

2033. Think about that. While the tech press writes breathless pieces about cloud-native transformation, SAP is quietly guaranteeing on-premise support for another eight years because that's what the actual market demanded. The gap between what gets written about and what enterprises actually run is enormous, and I think small and mid-size businesses lose money every year by reading tech media instead of watching what the big players actually do.

The Repatriation Story Nobody Is Telling Loudly Enough

There's a real movement happening and it's not getting the coverage it deserves. Companies that migrated to the cloud - fully, enthusiastically, the way everyone said you were supposed to - are moving back. Partially. Selectively. Strategically.

25% of organizations surveyed in the United Kingdom have already moved half or more of their cloud-based workloads back to on-premises infrastructures, according to a study by Citrix. A quarter of companies. That's not a rounding error. The survey also showed that 93% of respondents had been involved with a cloud repatriation project in the past three years.

And here's my favorite detail in this whole story: AWS itself testified that analyst reports suggest cloud services only make up around 15% of the IT services market as a whole, and AWS acknowledged that the perception that once customers move to the cloud, they never return to on-premises is not correct. Amazon. The company whose entire business model depends on you staying in the cloud, testified to a regulator that people leave. That's not a concession you make unless the data is undeniable.

I had a client - I'm still coaching a few people through the transition, life continues - who called me last Thursday to check in. Chris answered instead because my phone was on his desk charging. I got the phone back and called her back and she said she'd just moved three of her five business tools back to desktop-installed software because her internet kept cutting out during client calls and she was losing data. She'd spent eight months migrating to cloud tools. She spent three days moving back. The cloud's uptime promises don't survive contact with a rural ISP.

The Business Argument for Boring

I want to be honest about what I'm not saying. I'm not saying cloud software is bad. I use cloud tools. Cloud-based CRMs have made small business sales operations genuinely more functional than they were a decade ago. The collaboration tools are real. The flexibility is real. If you're scaling fast and your team is distributed, the cloud is probably the right call for a lot of your stack.

But I am saying that the contempt for on-premise and offline software - the assumption that it's only used by laggards who haven't gotten the memo - is costing businesses money and competitive clarity. Despite the hype around digital transformation and the seeming ubiquity of the cloud, many companies prefer legacy systems because of their reliability, cost-effectiveness, and familiarity. Those aren't sentimental reasons. Those are operational ones.

The boring software that's been running your competitor's inventory system for eleven years has something your shiny new SaaS stack doesn't: a team that knows exactly what it does, zero migration risk, and a monthly cost of zero beyond maintenance. Companies have often invested millions in an on-premise platform - investments for the long term, not lightly discarded if the platform still functions fine. Why pay monthly or annually for functionality in the cloud if you already have a solid platform? That's not a failure of vision. That's capital allocation discipline.

There's a version of this conversation that applies directly to sales intelligence and outreach tools too. A lot of teams are running sales intelligence platforms with cloud-synced everything, real-time enrichment, constant integrations - and they're also spending 30% of their budget on idle capacity and fighting sync errors between four different tools. The team running a simpler stack, maybe even with some offline components, with higher adoption rates because it actually works consistently - that team is probably outperforming on pipeline per dollar spent.

I ran a cold email sequence for one of my remaining coaching clients from my laptop, offline-composed, queued for send when I hit wifi. It sent clean. No dashboard to check, no integration to break. Sometimes the simplest pipeline is the one that actually runs. (I have thoughts on email automation more broadly but the short version is: tool complexity and send quality are often inversely correlated.)

Expressive pencil and ink wash sketch of an old server tower glowing with a single green light in a dim corner room, with a neon cityscape visible through a foggy window behind it — Showed this to Derek and he said it looks like our old office. That was not a compliment. But I think that light still being on is kind of the whole point.

What Your Competitors Actually Know

Here's my actual take: the companies still running on-premise and offline software are not ignorant. They made a different bet, and in a surprising number of cases, the bet is paying off. They're not paying cloud egress fees. They're not dealing with vendor lock-in. They're not scrambling every time a SaaS company gets acquired or raises prices. Public cloud platforms often use proprietary tools and APIs that make it difficult - and expensive - to migrate. When you're locked in, you're negotiating from weakness every renewal cycle.

Meanwhile, by deployment, the on-premise segment contributed the highest market share in the business software market in 2024. The highest market share. Not cloud. Not hybrid. On-premise. This is not ancient history. This is last year's data.

Linda showed me an article a couple weeks ago about how Gerald's company - she works with a few of his old colleagues - had just finished a three-year cloud migration and immediately started a review of what to pull back on-premise. She read me the headline. I said that tracks. She said Gerald had said the same thing. Gerald is usually right about this stuff, apparently.

The tech industry has a marketing problem dressed up as an innovation narrative. Cloud software is genuinely better for some use cases. Offline and on-premise software is genuinely better for others. What's not genuinely true is the story that one is smart and one is stupid. That framing serves vendors. It doesn't serve your P&L.

The businesses that will win the next decade aren't the ones that chased every SaaS trend. They're the ones that were honest about what they actually needed, ran boring systems well, and spent the money they saved on something that actually moved the needle. That's not a failure of ambition. That's what competence looks like from the outside.

I'm aware I'm writing this from a parking garage. I am also aware I'm right.