Welcome to this week’s Saturday Strategy. Grab a coffee, buckle in, and let’s talk about the most powerful, persistent, and occasionally perilous force in the corporate world: the gravity of Microsoft Excel.
For decades, we’ve predicted the "death of the spreadsheet." We’ve heralded the era of the "Single Source of Truth" and promised that enterprise-grade Business Intelligence (BI) would finally liberate us from the clutches of .xlsx files (or, worse, .xls but let's deal with that another day, my nerves can only take so much….. I'm tired, people.) Yet, here we are in 2026, and Excel's gravity is deepening. If anything, Excel is as addictive as a drug for many businesses, and it is the "secret database" that is an open secret in an organisation.
Microsoft recently introduced two new functions that might seem like minor productivity wins at first glance: IMPORTTEXT and IMPORTCSV. However, if you sit where I sit, at the intersection of AI strategy and data governance, these functions represent a significant shift in how data moves within an organisation. They are the latest boosters in Excel’s orbit, pulling business users further away from governed environments and deeper into what I call the "secret database."
The Gravity of Convenience
Why does Excel have such immense gravity? It’s simple and I see it time and time again. We humans love convenience.
When a business user needs an answer, they don’t want to wait for a ticket to be cleared by the data engineering team. They don’t want to navigate a complex Power BI workspace if they only need to look at three columns from a CSV file. They want the "grid."
The new IMPORTTEXT and IMPORTCSV functions provide immediate productivity wins. They allow users to bring external data into the spreadsheet environment with unprecedented ease. This is self-sufficiency in action. But as I’ve often said, counting is easy, but measuring is hard. When we make it easier to pull data in, we also make it easier to pull in errors, duplicates, and unverified "facts" that soon become the basis for executive decisions.
The "Daisychain" Effect and the Secret Database
The real strategic challenge isn't just one person with one spreadsheet. It’s the "Daisychain" effect.
With these new import functions, it becomes incredibly easy to link spreadsheets together in a fragile, invisible web. Spreadsheet A imports a CSV from a local drive; Spreadsheet B links to a range in Spreadsheet A; Spreadsheet C pulls it all together for a board report. This is the birth of the "Secret Database", a sprawling, undocumented architecture that exists entirely outside the purview of IT and security teams.
For business users, this is a dream of agility. For the organization, it is a governance nightmare. These daisychains are brittle. If one file is moved, or one column header in a source CSV changes, the entire house of cards collapses. We are seeing a return to Excel Hell (if we ever climbed out of it, of course….) but this time, it’s powered by high-speed automation and a false sense of security.
Shadow AI: When Copilot Joins the Chain
The gravity of Excel is intensified by the integration of Microsoft Copilot. This is where we need to buckle up: we are talking about automated data ingestion combined with manual data entry.
When Copilot is asked to "Analyze this data," it can suggest and execute imports, transforming "messy" data into "clean-looking" tables in seconds. This creates a significant "Shadow AI" risk, with everyone directing blame at everyone else. For example, IT teams may often feel that the "shared responsibility" model of cloud means that they will not be responsible or accountable, and the refrain becomes "It's Microsoft's fault" because people are comfortable with blaming external factors, particularly ones that are Jupiter-sized and just as remote and inaccessible.
If your AI is training on or analysing data pulled in via unmonitored IMPORTCSV functions, how do you verify the lineage of that data? Or do you simply trust the machine, and don't worry too much about it? We’ve discussed the risks of the AI Pilot Trap before, and this is a prime example. The ease of remaining within the spreadsheet grid may further entrench data silos when it is bolstered by AI. Why move to a centralised data warehouse when Copilot can give you a "good enough" answer right here in your "Secret Database"?
The real challenge is data fluency, not tools. We are giving people faster cars without checking if they know how to navigate the terrain.
The Strategic Limit: Accessibility vs. Fluency
As an analyst, I see a recurring pattern: organisations mistake tool accessibility for data fluency.
The fact that an employee can import 50,000 rows of data from a text file into Excel doesn't mean they understand the limitations of that data. Does the data contain nulls that will skew the average? Are the date formats consistent? Is there a sampling bias?
When tools become easier to use, the "barrier to entry" for data analysis drops, but the "barrier to accuracy" remains high. We are reaching a point where the strategic limit of the spreadsheet is being obscured by its own convenience.
Is Your Power BI a Spotlight or an Instagram Filter?
For many organisations, the push toward Power BI and its governance was intended to solve the Excel problem. However, the deepening gravity of Excel means that many Power BI reports are simply "beautified" versions of those same secret spreadsheets.
I recently wrote about whether your Power BI is a spotlight or an Instagram filter. When Excel becomes the primary ingestion point for BI, we aren't creating a transparent data culture; we are just putting a professional veneer over a chaotic foundation. The new import functions make it even more likely that the data feeding your high-level dashboards originated in a "daisychained" spreadsheet that nobody in IT knows exists.
Securing the Grid: A Middle Ground?
We cannot, and should not, try to ban Excel. It is the world’s most successful data tool for a reason, with over 3 billion users (which is a small planet of broken spreadsheets, I'm sure). However, we must address the security implications.
We need to treat these "Secret Databases" with the same diligence we apply to enterprise systems. This means focusing on securing Schrodinger’s Spreadsheets, those files that are simultaneously critical and invisible until they break.
If your strategy is to ignore the "Secret Database," you are essentially accepting a massive amount of hidden technical debt. Every new function like IMPORTCSV makes Excel more powerful. However, it also means your organisation needs to have a robust data literacy program in place.
The Saturday Strategy Takeaway
As we look at the source of these updates from the Microsoft Tech Community, it’s clear that Microsoft is leaning into Excel's role as the primary data hub for the individual.
The question for you, as a leader, is this: Is Excel’s gravity pulling your data strategy off course?
- Productivity Gains: Yes, these functions save time.
- Strategic Risks: Yes, they encourage silos, daisychaining, and Shadow AI with no accountability.
With Excel, is it ever possible to fight the gravity now? The way forward is to lean into the business need for speed with the human need for convenience. We need to bridge the gap between the convenience of the spreadsheet and the safety of the enterprise warehouse. This starts with data fluency: teaching users not just how to import data, but when a spreadsheet has reached its strategic limit and needs to be migrated to a governed environment.
Let’s Discuss
How is your organisation navigating this change? Are you encouraging these functional gains to empower your users, or are you doubling down on centralised governance to prevent the "Secret Database" from taking over?
Are you worried about the fragility of the "daisychain" effect? Or are you excited by the potential for business users to become more self-sufficient through Copilot-led imports?
Convenience is a powerful drug, and Excel is as addictive as crack for many businesses. However, in the world of AI and data strategy, clarity and governance are the only things that keep us grounded.
Join the conversation. Reach out via and let's discuss how to balance the gravity of Excel with the needs of a modern, data-driven enterprise. 🤝🚀
References:
- Microsoft Tech Community: Bring data into Excel with the new import functions
- Jen Stirrup Consulting: Excel Hell: 7 Signs Your Business Needs Data Automation
- Jen Stirrup Consulting: Escaping the AI Pilot Trap
