By Hiran de Silva
Something has changed in the Excel world over the past 30 years.
I don’t mean Excel itself. Obviously, Excel has changed enormously. It has acquired new functions, new interfaces, new data tools and a continuous stream of improvements.
I mean something rather more fundamental.
The way we think about Excel has changed.
And I want to approach this subject carefully, because it would be very easy to interpret what follows as criticism of today’s Excel users. It isn’t.
Quite the opposite.
Excel is now available to vastly more people than it was 30 years ago. That democratisation of technology is an extraordinarily good thing.
But when the demographic changes, the conversation changes too.
And I believe something important has been lost along the way.
A small feature raises a much bigger question
This article was prompted by a LinkedIn discussion about a new or returning Excel feature: automatically refreshing PivotTables when their source data changes.
That’s useful.
If somebody changes the underlying data, it is reassuring to know that the PivotTable automatically reflects that change. There is no uncertainty about whether what you’re looking at is current.
So I have no objection whatsoever to the feature.
But it made me ask a different question.
Where does this rank in the wider scheme of Excel capability?
For an experienced Excel developer, automatically refreshing a PivotTable has never been particularly difficult. You could refresh it when a workbook opens, when a worksheet is activated or in response to another event. You could simply provide a Refresh button.
The problem was already solvable.
That doesn’t make Microsoft’s new implementation pointless. Making useful capabilities easier and more accessible is a good thing.
But it illustrates something about how the Excel conversation has evolved.
Thirty years ago, the expectation was often:
Excel doesn’t do exactly what I need, so I’ll make it do it.
Increasingly, the expectation today is:
Does Excel have a feature that does what I need?
That is a profound change in thinking.
Something remarkable happened in the 1990s
By the mid-1990s, Excel had become an extraordinarily programmable environment.
Excel 5 brought VBA into the product. PivotTables were available and programmable. Excel’s object model allowed developers to manipulate the application itself.
Connectivity to external data was also emerging as an important part of the architecture.
The significance of this is easily underestimated today.
Excel wasn’t merely a collection of functions and buttons.
It was becoming a platform on which professionals could build things.
That distinction matters enormously.
If a particular function didn’t exist, you could create one.
If a repetitive operation wasn’t automated, you could automate it.
If information lived somewhere else, you could retrieve it.
If users needed an interface, you could build one.
If a business process needed to send information back to a database, you could program Excel to do that too.
Microsoft’s engineers didn’t have to anticipate every possible business requirement and provide a button for it.
They had given us the tools with which we could help ourselves.
That was a step change.
The £11,000 laptop
There is another part of this history that we tend to forget.
Computers were expensive.
The most expensive laptop I bought for my work, in January 1998, would cost something in the region of £11,000 in today’s money by my calculation.
Think about that.
Would the average Excel user today spend £11,000 on a laptop?
Of course not.
But that is precisely the point.
The people buying their own computers and seriously developing with Excel in the 1990s represented a much narrower demographic than the enormous population using Excel today.
Computing has become dramatically cheaper.
Excel has become ubiquitous.
Technology has been democratised.
The population changed.
And when the population changed, naturally the educational market changed with it.
That isn’t an insult to today’s users. It is economics.
From professional community to mass audience
The Excel community I encountered decades ago was populated by people such as Stephen Bullen, Bill Manville, Bob Phillips, Myrna Larson, Tom Ogilvy, Rob Bovey, John Green, Dave Hawley and Chip Pearson.
These people weren’t primarily trying to accumulate millions of casual viewers.
They were helping people solve problems.
The audience itself was different.
Today we have something that barely existed then: the professional social-media Excel educator.
And social media has its own economics.
If you are building a YouTube channel, LinkedIn following or training business, there is an obvious incentive to create material that is useful to the largest possible audience.
Again, there is nothing inherently wrong with that.
Millions of people have learned useful Excel skills as a result.
But inevitably, the centre of gravity moves.
The question becomes less:
What is ultimately possible with Excel?
and more:
What will be immediately useful and understandable to the largest number of Excel users?
Those are not the same question.
Imagine calling an electrician to switch on the light
Here’s an intentionally exaggerated analogy.
Imagine a community that has never had electricity.
Thirty years from now, electricity arrives. Everybody loves it. Electric lights transform people’s lives.
But because electricity is unfamiliar, people regard anything associated with it as specialist territory.
When somebody wants the light turned on, they call an electrician.
A whole industry could even emerge around helping people turn on their lights.
From inside that world, this would seem perfectly normal.
But you and I would find it extraordinary.
We would simply walk across the room and flick the switch.
The difference isn’t intelligence.
It is knowledge, familiarity and expectation.
And that, I believe, is close to what has happened in parts of the Excel world.
People are sometimes being taught elaborate ways of solving problems that an earlier generation of professional Excel developers would have approached from an entirely different architectural direction.
Not because today’s users are incapable of understanding the alternative.
They often haven’t been shown that the alternative exists.
When old technology looks like magic
This produces a fascinating consequence.
I can demonstrate an Excel solution today that looks futuristic even though the fundamental architecture could have been implemented decades ago.
My Gordon Ramsay Experiment is one example.
My Annual Budgeting and Budget Review models are others.
And Tim and Reg perhaps demonstrate the point most simply.
Tim carries out a familiar sequence of spreadsheet operations. Data arrives. He manipulates it, transforms it, restructures it and produces the required result.
There are countless tutorials teaching variations of this methodology.
Reg asks a different question.
Why are we doing this work at all?
Reg changes the architecture.
And suddenly much of Tim’s work disappears.
That can look like magic to an audience accustomed to Tim’s methodology.
But technologically, it isn’t magic at all.
The fundamental capability required to build Reg’s solution has existed for decades.
What changed wasn’t necessarily Excel.
What changed was the prevailing knowledge of what Excel could be used to do.
Arthur C. Clarke explained the phenomenon
Arthur C. Clarke famously observed that sufficiently advanced technology can become indistinguishable from magic.
There is an interesting variation on that idea.
Technology doesn’t actually have to be new to look like magic.
It only has to be unknown to its audience.
That distinction explains something I encounter repeatedly.
I demonstrate Excel communicating with a central database.
People are surprised.
I demonstrate hundreds of spreadsheets sharing centrally governed data.
People are surprised.
I demonstrate Excel both GETting and PUTting information through a Digital Librarian.
People are surprised.
I demonstrate an architecture in which the spreadsheet is the working document while persistent collaborative data lives centrally.
Again, people are surprised.
Yet these ideas aren’t dependent upon some revolutionary capability Microsoft invented last year.
The foundations have existed for decades.
This also changes how we think about new Excel features
This is why I sometimes have a different reaction from others when Microsoft announces a new feature.
I may think:
That’s useful.
But I also ask:
What new business capability does this actually give me?
Those are different questions.
A feature can make something easier without making something newly possible.
Focus Cell is a good example.
Highlighting the row and column of the selected cell is useful. Microsoft providing an accessible switch for everybody to use is welcome.
But people had already created versions of that behaviour themselves.
The new feature democratises the capability.
It doesn’t necessarily invent the capability.
That distinction becomes particularly important when discussing professional Excel.
Power Query illustrates the architectural question
Power Query is enormously useful technology.
It teaches important concepts including transformation, tables, joins and ETL.
But in an enterprise environment I want people to ask an architectural question before deciding which Excel feature to use:
Why are we moving this data into this workbook?
Sometimes there is an excellent reason.
Sometimes there isn’t.
If 100 people are collaborating in a business process, repeatedly bringing copies of organisational data into 100 separate workbooks may be solving the wrong problem.
Perhaps the data belongs in a central structured location.
Perhaps the spreadsheets should GET the information they need from that location.
Perhaps they should PUT their changes back.
Perhaps the spreadsheet should remain the flexible working document while a Digital Librarian provides the shared memory.
That isn’t an argument against Power Query.
It is an argument for architecture before features.
The danger of teaching only what is popular
Social media naturally rewards content that people already understand enough to search for.
That creates a potential feedback loop.
People search for ways to perform familiar spreadsheet tasks.
Creators make tutorials showing how to perform those tasks.
Those tutorials attract large audiences.
Large audiences encourage more tutorials about those tasks.
And eventually popularity itself can begin to look like evidence that this is how the task should be done.
But popularity and architecture are two different things.
A video receiving two million views proves that there is enormous interest in the subject.
It doesn’t prove that the underlying process is necessary.
Sometimes the most valuable Excel lesson might be:
Don’t do the task more efficiently. Remove the task.
The lost professional conversation
This is what I would like to bring back.
Not the 1990s.
Not old versions of Excel.
And certainly not an exclusive world in which powerful technology was available only to people who could afford extraordinarily expensive computers.
I want to bring back a particular way of thinking.
Before asking:
How do I do this in Excel?
Ask:
Why am I doing this?
Before asking:
Which Excel feature should I learn?
Ask:
What is the business process I’m trying to create?
Before asking:
How do I transform these 50 files?
Ask:
Why do 50 files contain the data in the first place?
Before learning how to perform a repetitive operation faster, ask:
Can the operation disappear altogether?
That is professional Excel thinking.
And it is as relevant in 2026 as it was 30 years ago.
The great opportunity
There is actually something enormously positive hidden inside this argument.
We now have the best of both worlds.
Technology that once required substantial financial investment is available to almost everybody.
Excel is everywhere.
Computers are inexpensive.
Relational databases are readily accessible.
Cloud infrastructure exists.
Learning resources are abundant.
And today’s Excel users are perfectly capable of learning the architectural thinking that professional developers applied decades ago.
The missing ingredient isn’t capability.
It is awareness.
That is why some of my demonstrations can appear almost magical.
They aren’t demonstrations of the future.
In some respects, they are demonstrations of a professional Excel tradition that we forgot.
And perhaps the next evolution of Excel education shouldn’t simply be about teaching more features.
Perhaps it should be about recovering the question that professional Excel developers have always asked:
Given everything this platform can do, what is the best way to solve the business problem?
That is a very different question from:
Which button should I press?
And after 30 years of extraordinary technological progress, perhaps the most important Excel innovation we need now isn’t another button.
It is the evolution of our thinking.
Add comment