POST / 03 · ENGLISH · SCROLL STORY

Is early mover advantage a lie?

Scroll, and the squares will tell you who wins: the first, the early, or the late.

DATE: 10 SEP 26 READING TIME: 8 MIN LANGUAGE: ENGLISH

DOWNLOAD THE DATASET (XLSX) ↓

CATEGORIES54first products from 1925 to 2018

PIONEER STILL #120 / 5437 percent, if you are generous about "first"

MEDIAN WINNEREntrant #3fifteen months after the pioneer

01 / THE PREMISE

Two clichés.

They say "get there first and the market is yours". The other, well, actually someone from Twitter, says that first mover advantage is the biggest myth.

Both are told with anecdotes, which is convenient, because it is the best and, most of the time, not so scientific way of trying to prove something. As a result, I wanted to check this. I curated lots of categories, found who shipped first, who is number one now (or at the category's peak, if it is dead), and looked at the gap between the two.

Hardware: a physical product is easy to check in terms of who came first and who is selling the most. So we are going with those. There are 54 categories, from a 1925 Leica film camera to a 2018 folding phone. Each one is a square below. Scroll and enjoy.

FIG. 01 / THE BOARD

STEP 01 / 54 SQUARES

Fifty-four categories, sorted by the year the first product shipped. Top left is the Leica I, 1925. Bottom right is the foldable phone, 2018. Cameras, televisions, phones, chips, printers, consoles, a robot vacuum, two cars.

STEP 02 / THE PIONEERS

Every square has a pioneer: the first brand to put a commercial product in the category. Hover a square to see it. For 17 of the 54 the "first" is disputed, which is already a warning about how this story usually gets told.

STEP 03 / STILL ON TOP

Now the question. In how many of these is the pioneer still number one? Twenty. GoPro, Intel, HP in both inkjet and laser printers, Toyota in hybrids, Amazon in smart speakers, Fitbit, Raspberry Pi, the iPad. Thirty-seven percent.

STEP 04 / PASSED, OR GONE

In ten more the pioneer is still in the top five but got passed: Polaroid behind Instax, Motorola behind Nokia, Toshiba behind Sony in DVD players. And in 24, nearly half, the pioneer is not in the top five at all. Casio, Kodak, 3dfx, MakerBot, Osborne, Magnavox, Pebble.

STEP 05 / CHANGE THE QUESTION

So 37 percent hold on and 44 percent fall off the leaderboard. On that count the cliché looks about half true. But "first" is doing a lot of work in that sentence. Ask something else: not who was first, but how long after the first did the eventual winner show up? Each square is now a bar. Its length is that delay, in months.

STEP 06 / THE FIRST YEAR

In 26 of 54 categories the brand that ended up on top entered within twelve months of the pioneer. Sixteen were the pioneer itself. Ten arrived in the first year: Apple, one month after LG's Prada phone; Nvidia, ten months after 3dfx; Samsung, nine months after Sharp's LCD television.

STEP 07 / THE LONG TAIL

Stretch the window to two years and it is 30 of 54, more than half. The tail is the other story: 14 winners arrived more than five years late. Canon in 35mm, 129 months after Leica. Samsung in DRAM, 161. Sony's PlayStation, 267. Seagate, 285. Fujifilm's Instax, 395 months, thirty-three years after Polaroid.

STEP 08 / ENTRY ORDER

Regroup by entry order instead of time. The winner was the first entrant 16 times, the second 10 times, the third 7 times, and then it thins out. The median winner is entrant number three, fifteen months after the pioneer. Not first. Not tenth. Early.

STEP 09 / HOW THE LATE ONES WON

And the late winners have a pattern. Nearly all of them won on manufacturing scale (Samsung, Lenovo, Seagate, Lite-On, TP-Link, Creality, Roborock) or on a platform reset: the PlayStation with CDs, the Switch, Instax as a revival. Nobody won late by shipping a slightly better version of the pioneer's product.

02 / THE FULL LIST

If you want to check, here are all 54.

Number one is measured at the current market (2023–24) for living categories and at the peak year for dead ones. Delay is months between the pioneer's release and the winner's first qualifying product.

#	Category	Pioneer (first release)	#1 brand	Delay (months)	Entry order	Pioneer now

03 / THE VERDICT

So, is it a lie?

Being literally first is a weak bet. About a third of pioneers keep the crown, closer to 30 percent if you refuse to count shared or disputed firsts, and 44 percent are not on the leaderboard by the time the category matures. If that is what someone means by first-mover advantage, they are right to be suspicious of it.

Being early is a strong bet. In half of the categories the winner entered within a year of the pioneer, in more than half within two, and in 37 of 54 at least one first-year entrant is still in the top five. The advantage is real. It just belongs to the first wave, not the first name.

THE SHORT VERSION The pioneer's job is to prove the category exists. The winner's job is to arrive before anyone else believes it.

The median winner entered third, fifteen months late, after the pioneer had spent the money to find out what the product actually was.

This is not a new finding, which is reassuring. Golder and Tellis counted consumer goods in 1993 and found that pioneers failed 47 percent of the time and that "early leaders," firms that entered later but before the category took off, ended up with roughly three times the pioneer's share. Markides and Geroski wrote a whole book about it and called it Fast Second. Consumer hardware from Leica to the Quest does not look any different.

04 / WHAT IT DOESN'T SHOW

Correlation, definitions, survivors.

First, there is survivorship. Every category here matured into something with a market share to measure. However, there are lots of categories that... never became categories. As a result, they are not in this set. Which means that though they failed, the pioneers of those categories are not counted in this research, making the pioneer result look better than it actually is.

Secondly, there is "definition". Logitech counts as the webcam pioneer because it bought Connectix; Apple shares the touchscreen smartphone with LG because the Prada shipped a month earlier; 17 of 54 pioneers are flagged as disputed in the data. Read the entry-order numbers as roughly right, not exact.

Third, measurement. "Number one" mixes current shipment shares (2023–24) for living categories with peak-year shares for dead ones, and the shares come from mixed public sources: company reports, CIPA, IDC, Counterpoint, Statista. The dataset grades 47 of the 54 categories A for evidence quality and 7 as B. Every row carries its own source links in the spreadsheet above.

Lastly, the scope. I only checked hardware. Software, platforms, services: there are many other things to check, and we cannot be sure whether the result looks similar there. However, I was lazy and did not check those. So I am leaving this as an exercise to the reader.

IF YOU ONLY REMEMBER ONE ROW: Median winner: entrant number three, fifteen months after the pioneer.

5 REFERENCES

Two clichés.

If you want to check, here are all 54.

So, is it a lie?

Correlation, definitions, survivors.

Sources.