Technology deep dives

Why did Kodak fail? It missed the rate, and so did the CCD makers

GetFocus

·

November 19, 2026

·

8

min read·Updated

November 19, 2026

Why did Kodak fail? It missed the rate, and so did the CCD makers

Stay ahead of the curve

Get exclusive insights delivered directly to your inbox. Join thousands of tech leaders.

Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.

TL;DR

  • Kodak built the first digital camera in 1975, was warned by its own study in 1981, and filed for Chapter 11 on 19 January 2012.
  • In the GetFocus backtest, CCD imagers improved at 89.8% a year in 1980 against 17.3% for film, while film still out-filed every digital technology into the mid-2000s.
  • The technology that beat film lost next: CMOS passed CCD on improvement rate in 2000, five years before the filings agreed, and in 2015 Sony announced it would end CCD production.
  • The rate is directional. It shows which technology is closing and how fast; the decision and its timing stay with management.

Why did Kodak fail? Kodak failed because it kept funding silver-halide film while digital imaging improved faster every year, and the boardroom had no number that measured the gap. Kodak invented the digital camera in 1975, received an internal warning in 1981, and still filed for Chapter 11 on 19 January 2012.

The patent record holds a second story that the usual telling leaves out. CCD, the sensor technology that beat film, was itself overtaken by CMOS, and in 2015 Sony announced it would end CCD production. We build GetFocus, which measures how fast technologies improve from the way patents cite each other. We replayed the imaging record year by year, using only the patents that were public at each date, and the full series is on the Kodak backtest page.

Why did Kodak fail when it invented the digital camera?

The first Kodak digital camera was built 37 years before the bankruptcy. In December 1975, Kodak engineer Steven Sasson took the first photo with a device built around a 100 by 100 pixel CCD from Fairchild. It weighed 3.6 kg, was about the size of a toaster and needed 23 seconds to record one black-and-white image to cassette tape, according to IEEE Spectrum's history of the camera. After the demonstrations, executives said they saw no market for it, and Sasson was barred from showing the prototype outside the company.

The decision made sense on the numbers Kodak watched. Film was a consumable with extraordinary margins, and digital hardware promised thin ones. Kodak's market value reached $31 billion in February 1997, 22 years after Sasson's prototype, and the company employed more than 140,000 people at its peak.

Did Kodak see digital photography coming?

Yes, and it wrote the finding down. In 1981 Sony announced the Mavica, an electronic camera, and Kodak commissioned an internal assessment. The study concluded that digital would replace film in roughly a decade, according to the 2012 accounts cited on the Kodak backtest page.

Kodak also had the capability. In 1994 it designed the QuickTake 100 for Apple, one of the early consumer digital cameras, which its manufacturing partner Chinon built. Its own answers to digital were photo kiosks, Photo CD, the Advantix film system and an early digital SLR line. They were adjacent bets that protected film's role.

The 1981 study gave management a horizon of about ten years. With no yearly reading to check it, the warning aged while film's results stayed excellent.

What did the improvement rate say while film still looked like a fortress?

An improvement rate works like an interest rate on performance: a technology that improves 30% a year compounds past one that improves 20%, whatever their starting positions. The rate tells you who is closing the gap. Film was ahead on image quality for decades, while the rate tracked the speed of the technology coming up behind it.

The consensus chart is patent filing volume. In 1997, the year of peak Kodak, silver-halide film ran 2,611 new patent families, more than every digital imaging technology combined, and film kept out-filing digital into the mid-2000s. Filing volume shows where the industry's money and attention sit, and in the 1990s it described film as a fortress.

The rate chart read the opposite, from the first reading onwards. These are the readings from the GetFocus backtest, in improvement per year:

  • 1980: CCD 89.8%, film 17.3%
  • 1994: CCD 36.5%, CMOS 30.9%, film 25.8%
  • 1997: CCD 33.0%, CMOS 31.5%, film 24.0%
  • 2000: CMOS 32.3%, CCD 30.6%, film 23.1%
  • 2005: CMOS 58.1%, CCD 30.2%, film 23.0%
  • 2012: CMOS 54.4%, CCD 29.0%, film 22.9%

Take the 1997 line. CCD at 33.0% against film at 24.0% looks like a modest gap. Held for ten years, that gap would roughly double CCD's performance relative to film (our arithmetic on those two readings, as an illustration; the rates moved from year to year). Film read between 17% and 26% in each of these six readings, and both digital technologies read above it each time.

On the backtest, the separation was clear from the first reading in 1980. The Kodak case study puts the steep fall in film sales at 2000, so the signal led the market by two decades. By 2012, the year of the Chapter 11 filing, film was down to 56 new patent families.

How is the improvement rate measured?

GetFocus reads the rate from the patent citation network, through two signals described on the method page. Cycle time is the number of years between one generation of inventions and the next. Knowledge flow is how much later inventions build on earlier ones. Short cycles and high knowledge flow mean a technology is advancing fast. The method is validated on more than 150 technology domains at R² ≈ 0.7, meaning the forecast explains about 70% of the variation in measured improvement rates.

Patents make a fair replay possible because each one carries a date. For any past year, the evidence can be rebuilt with only what was public then, so hindsight stays out of the series. Our article on how to backtest a forecasting method explains the test, and how to measure technology improvement covers the research behind it.

What happened to the CCD makers that beat Kodak?

They met the same measurement twenty years later. CCD imagers started at 89.8% a year in 1980, and their rate decayed to about 30% by 2000. That year the CMOS rate passed it, at 32.3% against 30.6%. The first camera phones shipped the same year.

At the time, the CCD camp had a true argument: CCD delivered better image quality, with lower noise and better uniformity, into the mid-2000s. The filings agreed with them until 2005, when CMOS out-filed CCD for the first time. By then the CMOS rate read 58.1% against CCD's 30.2%. The iPhone shipped with a CMOS sensor in 2007, and in 2015 Sony announced it would end CCD production and put its investment into CMOS. Sony was the company whose Mavica had alarmed Kodak in 1981.

The two underlying signals show why. CCD's cycle time stretched from under four years to almost seven, so each generation took longer to arrive, and its knowledge flow halved. CMOS held its knowledge flow steady for two decades. Film stayed flat on both signals for the whole period.

Why did smart people back the slower technology?

Their arguments were true. Film did out-perform digital for decades, and CCD did beat CMOS on image quality into the mid-2000s. Each argument described the present accurately, and each one held at board level.

What the arguments left out was speed. A former research fellow at Eastman Kodak put it this way in the Kodak case study:

"It would have been great if we'd had something like GetFocus to show management how serious the situation was."

What are the limits of this reading?

The improvement rate is directional. It shows which technology is closing and roughly how fast. The year a market tips also depends on cost, manufacturing scale and distribution, which the rate leaves to other analysis.

Close readings deserve caution. In 1997, CCD and CMOS sat 1.5 points apart, at 33.0% against 31.5%, and a call between them on the rate alone would have been weak before the 2000 crossover. The film reading was clearer: CCD read above film in every reading listed above, by at least six points.

The backtest reads patents only. Progress that companies keep as trade secrets stays outside it. The rate also leaves the decision with people. Kodak held a written warning from 1981 and the engineering to act on it, and keeping film at the centre was still management's choice.

What does the Kodak digital camera story mean for an R&D leader today?

Kodak had the invention, the warning and the capability, and it lacked a yearly reading that could stand next to the P&L. Three checks carry over to any portfolio:

  1. List the competing technologies for the job your customers pay for, including the ones your own labs have shelved.
  2. Compare their improvement rates, and treat filing volume and today's performance lead as the consensus view, which trails the rate by years in this record.
  3. Set the trigger in advance: the reading at which funding moves. Our article on when to move away from a core technology shows how to stage that move.

This is the comparison technology forecasting software is for: every option for one technical problem, ranked by how fast it improves, re-read as new patents publish.

FAQ

Did Kodak invent the digital camera?

Yes. Kodak engineer Steven Sasson took the first photo with his prototype in December 1975. It used a 100 by 100 pixel CCD, weighed 3.6 kg and needed 23 seconds to record one black-and-white image to cassette tape. Kodak kept it inside the company.

When did Kodak go bankrupt?

Kodak filed for Chapter 11 bankruptcy protection on 19 January 2012, in the U.S. Bankruptcy Court for the Southern District of New York. That was 37 years after the first digital camera prototype and 15 years after its market value peaked near $31 billion in February 1997.

Could Kodak have known earlier that film would lose?

The GetFocus backtest says yes. Replayed with only the patents public at each date, CCD imagers read 89.8% a year in 1980 against 17.3% for silver-halide film, and film stayed near 23% a year from 2000 onwards. The Kodak case study dates the steep fall in film sales to 2000, two decades after that first reading.

Why did CCD sensors lose to CMOS?

CMOS kept improving faster. Its improvement rate passed CCD's in 2000, at 32.3% a year against 30.6%, and by 2005 it read 58.1% against 30.2%. CCD's cycle time stretched from under four years to almost seven while its knowledge flow halved. In 2015 Sony announced it would end CCD production.

Is there a film business in your portfolio?

Film still won the filing chart and the P&L in 1997, and the rate had been pointing at digital since 1980. If you are weighing a technology that looks safe on those same two charts, send us the decision. Get a free 48-hour Decision Brief: a first-pass radar of the competing options, their improvement rates and suggested triggers.

Ready to see it in action?

Bring us one technology decision you're weighing. We'll send back the competing routes mapped by improvement rate and maturity.

Talk to us