Excel is ending the forty-year rule that a cell holds one value. I read all 50 comments. Exactly one named Microsoft — 0.2% of the vote — while 50.3% went to people asking whether the files they already have will still open

At 7:20pm on 28 September 2026, ITmedia NEWS published on Yahoo! News.

Microsoft had announced, on 24 September local time, a new Excel feature called Lists. It lets you put several pieces of data into a single cell and then use each of them separately for filtering and calculation.

You could always type two names into one cell before, but the whole thing was treated as one string of text. Now, if a cell holds both Tanaka and Sato, searching for either name will find that cell.

Microsoft itself describes this as changing the forty-year-old rule that a cell holds one value.

The article says one more thing. This is a preview, rolling out to some Windows and Mac users. There are limits — data inside an array cannot be used in a pivot table — and the specification may change before general availability, so Microsoft does not recommend using it on important Excel files.

Fifty people wrote in the Yahoo! News comment section underneath.

What I thought

When I saw this article, here is what I thought.

Microsoft, bolting on features nobody needs.

The impression is of a company that has already slipped to third place or below.

A company that will shove any unnecessary, unwanted feature into an update just to stay visible.

It is finished.

All it has left is clinging to the operating system.

That was my reading.

You will notice that all of it is about the company. A company that adds needless things. A company that has slipped. A company that wants attention. A company that is finished. A company clinging to its operating system.

So this time I turned that into something countable. Is this thread talking about the company?

What came back was not what I expected.

85 to 15 — 50 comments, about eleven hours old

I read all 50 parent comments. The article carries 62 comments in total, 12 of them replies.

The buttons: 1,474 agree, 204 insightful, 307 disagree. Supportive presses against pushback comes out at 85 to 15.

Of the 57 threads I have counted, this is the fourth smallest. Fifty comments is a number you can read one by one in under an hour.

The vote sits high up. The single top comment holds 311 agree presses, 21.1% of everything pressed, with 52.0% on the top three, 66.9% on the top five, 82.3% on the top ten and 93.8% on the top twenty.

Lower down, 6 of the 50 comments (12.0%) drew no agree press at all. Two (4.0%) drew no button of any kind, and nine (18.0%) drew more disagree presses than agree.

The average comment runs 49.2 characters; the longest is 145. This is a short thread.

The most-pressed comment says: a cell is the smallest unit of data

Here is the comment this thread read most.

311 agree presses — 21.1% of everything this thread pressed.

Look at what it does. It does not call Microsoft anything. There is no company name in it, no market position, no operating system.

There are two things in it: every single function, and the smallest unit. SUM sits on top of the one-value-per-cell rule. So does VLOOKUP. So do the tens of thousands of spreadsheets running in offices right now. Move the rule and everything stacked on it has to be checked again.

That is the one point this thread pressed hardest.

Exactly one comment wrote the name Microsoft

Here is the scorecard against my reading.

Of 50 comments, one wrote the name Microsoft. It drew 3 agree presses, 0.2% of the vote.

That is the only one.

The rest went like this.

• Comments mentioning third place, market share or rank — none
• Comments naming Google Sheets or any other rival spreadsheet — none
• Comments saying the company does this for attention, publicity or show — none
• Comments mentioning money, profit, subscriptions or price rises — none
• Comments writing the words Windows or OS — none

All five, zero. My reading was built entirely out of the company. The company is almost absent from this thread.

The words unnecessary and unwanted appear in 4 comments — 1.8% of the vote

I counted the literal words too. Comments using unnecessary, unwanted, pointless or a waste: 4, or 8.0%, holding 27 agree presses between them — 1.8% of the vote.

In words, they say what I said. In votes, they were not what this thread came for.

I also counted comments insulting the company's standing — finished, incompetent, laughable, insane, stupid. 4, or 8.0%, holding 41 agree presses, 2.8% of the vote.

One of them comes closest of all to my own reading.

Four agree presses. 0.3% of the vote.

That comment writes down almost everything I wrote. Needless. Finished. Something better to be doing. And four people pressed it.

Half the vote went to: will what I already have still work?

So where did those 1,474 agree presses go?

Function specifications. Save formats. File extensions. VBA. Macros. csv. Old versions. Existing work. In other words, whether the things they already own will break. Thirteen comments, 26.0% of the thread, holding 741 agree presses — 50.3% of the vote.

The second-placed comment is the model of it.

244 agree presses, 16.6% of the vote.

Note that it leaves room for the other side: a fair few people will find it handy. Then it says that where it sits, caution comes first. It is not against the new thing. It is protecting what is already on the desk.

Fix something else first — 8 comments, 31.3% of the vote, including the one shared on X

There is a second block. Before this, please fix something else.

Eight comments, 16.0%, holding 462 agree presses — 31.3% of the vote.

Two files with the same name in different folders cannot be opened at once. What you see on screen is not what prints. Printing goes out of alignment. Startup is slow. It freezes. Pivot tables are unreliable. The Outlook VBA link. These are ten-year-old complaints, lined up.

The comment shared on X belongs to this block.

156 agree presses, 4th in this thread, 10.6% of the vote, with one reply.

It too says nothing about the company. There are only two things in it: who asked for this, and there are things to do first.

The two blocks — thirteen comments about breakage and eight about fixing something else — overlap in one comment. Take that out and you have 20 comments, 40.0% of the thread, holding 1,199 agree presses — 81.3% of the vote.

The other 30 comments hold 18.7%.

The most pushed-back comment is the one that said it was useful

Two comments, 4.0% of the thread, come out plainly in favour of the feature. Between them they hold 18 agree presses, 1.2% of the vote.

They also hold 104 disagree presses — 33.9% of every disagree press in the thread.

One of them holds 103 on its own: 33.6%.

17 agree presses and 11 insightful, against 103 disagree.

It is courteous about it. It stands on the other side first — it looks pointless at first glance — before saying what it thinks. It still became the most pushed-back comment in the thread.

This thread is not unanimous — at 85 to 15, the pushback is only 15% of everything pressed. But almost all of it landed on the people who took the feature's side.

Nobody wrote down the article's own caveat

One more thing surprised me in the count.

The article says clearly that this is a preview, that it is going to some users only, that the specification may change before general availability, and that Microsoft does not recommend using it on important Excel files.

Comments touching on preview status, general availability, trials, limited rollout, restrictions or that recommendation: none out of 50.

Comments writing the forty years from the headline: none either.

And yet the seventh-placed comment says this: the scary thing is that they suddenly announce actually, never mind!, so you should not use it unless you really have to. 61 agree presses.

Nobody quoted the caveat, and they are worried about exactly what the caveat says. Not because they read it. Because they have been here before.

Length, and replies

Comments under 40 characters number 27, or 54.0% — over half the thread. They hold 261 agree presses, 17.7% of the vote.

Comments over 80 characters number 9, or 18.0%. They hold 976 agree presses, 66.2%.

The same shape as last time. The vote goes to the ones that give a reason, not the ones that land a line.

There are 12 replies, three of them under the third-placed comment. 42 of the 50 comments (84.0%) drew no reply at all.

Context for readers outside Japan

Excel is used very widely in Japanese offices, and not only as a spreadsheet. Application forms, staff lists, meeting minutes, even rough drawings get made in Excel. One comment in this thread is a joke at exactly that: Japanese companies saying see, Excel really can do everything, you can even write documents in it.

So one value per cell is, in this country, more than a specification detail. Decades of paperwork are stacked on top of it.

What this thread set out to defend was not its opinion of Microsoft, and not anything about Japan. It was the files already sitting in its own drawer.

That is why the company name appears once, and rank and operating systems never. What 20 of the 50 wrote about was one question: will the thing I already have still open?

Things I have not checked

I have not checked where Microsoft actually ranks. Slipped, doing it for attention, finished — those are my words, not statements of fact. I wrote impression myself.

This note looked up no revenue figures and no spreadsheet market share. The same goes for clinging to the operating system: this is not an article about an operating system, it is an article about a feature in a spreadsheet.

I have not tried Lists. What I read was the ITmedia NEWS article and the 50 comments underneath it.

One note on the numbers. Yahoo! News listed 50 parent comments and I read 50, so every percentage here is out of that 50 — small enough that a single comment moves a share by 2%. Please read them that way.

Voices I counted but did not quote

Comments containing abuse of a person or a company: 4, or 8.0%, with 41 agree presses, 2.8% of the vote. One of them is quoted above; the rest are not.

Comments giving ethnicity, descent or race as a reason: none.

The scorecard

My reading, one line at a time.

• Features nobody needs — 4 comments, 1.8% of the vote
• Slipped to third place or below — none
• Doing it for attention — none
• Finished — 4 comments, 2.8%
• Clinging to the operating system — none

Three of the five are zero.

What took 81.3% of the vote instead was whether the things they already have will break, and whether there was something to fix first.

What this thread was actually saying

I was looking at the company. This thread was looking at its own desk.

Not one of the fifty cared where Microsoft ranks. One of them wrote the company's name. Twenty of them wrote about the files they are already carrying.

Whether Microsoft is first or third does not matter to these people. What matters is whether the spreadsheet they built last year opens the same way tomorrow morning. That is where the presses went.

The words from the most-pressed comment are worth setting down again: the fact that a cell is the smallest unit of data matters enormously.

What this thread defended was not an opinion about a company. It was the work it had stacked, for years, on top of that one line.