Landa Digital Printing – a textbook example of ‘sunk cost fallacy’?

If you are a regular reader of my blog, you will have noticed that I’m a fan of ‘behavioral economics’. It’s a fascinating field in economics, that studies the irrational behavior of human beings. One of the concepts in that field is the ‘sunk cost fallacy’, a bias that prevents rational decisions, such as investments. And, in my opinion, Landa Digital Printing is a textbook example. It’s the story of a passionate man who deserved his place in the history books, chasing his dream, even when evolutions of ‘old’ and new technologies filled the gap he was targeting.

Maybe I should first provide some background on behavioral economics. For a very long time, people studying economics considered human beings to be rational; it’s akin to math: a € 1 product is cheaper than a € 2 product. However, over the years, it became apparent that people aren’t that rational… One of the pioneers was the late Prof. Daniel Kahneman, who received a Nobel Prize for his work. His book ‘Thinking, Fast and Slow’ is highly recommended, it’s probably the most influential book I’ve read thus far. It reads like a thriller, and I can assure you: people are not rational at all, there are a lot of biases.

The Sunk Cost Fallacy
And one of those biases, is the ‘Sunk Cost Fallacy’. This website describes it as follows: “Sunk costs represent past expenditures that cannot be recovered and should not factor into current decisions, yet they often irrationally influence individuals’ economic choices. This concept is pivotal in understanding how individuals and businesses evaluate their options and make decisions. Despite the irrecoverability of sunk costs, they frequently cause a cognitive bias where people continue investing in a project, decision, or transaction based on the time, money, or resources they have already committed, rather than evaluating the prospect’s current and future value. This bias can lead to suboptimal decision-making and is known as the sunk cost fallacy.”

The recent details uncovered about Landa Digital Printing, particularly the total investments shareholders have made in the company, support my assessment that this is a textbook example of the sunk cost fallacy. In total, some 1,3 billion US$ was invested in Landa Digital Printing. That’s a lot. Plus: the assessement of the current shareholders is that Landa Digital Printing needs another 5 years and 300 million US$ to get break-even…

The development of the ‘nanography’ idea started already two decades ago, the official launch was in 2012, and – according to Benny Landa – immediately generated ‘over 400 orders’. Which didn’t materialize: from the 6 models (S5, S7, S10, W5, W10 and W50), only one survived: the B1 S10 and it’s perfecting sibling S10P. And even those were delayed time after time. The claims that the machines would go to beta customers in 2013, and then would be ready in 2015 (“One year from today”), weren’t fulfilled. The promise that dozens of machines were to be installed in 2019? Didn’t come true either. But investors kept raising their investment… Because they believed the dream that nanography would be a game changer, and because they had already put that amount of money into it. Just a few years more, just a few hundred million extra… The sunk cost fallacy…

Fail fast, fail often
An essential element of the ‘lean startup’ idea is ‘fail fast, fail often’. It’s the mantra of Silicon Valley. And when you fail: learn from it. That’s why failure as such isn’t shameful. But that second word of the mantra is essential: fast. If you prolong the struggle, you can end up with the sunk cost fallacy. Especially if you don’t learn in between, such as, for example, how the market and the market fit you were looking for have changed…
Since the start of the development of nanography in 2002, the gap that Landa wanted to fill, has disappeared. Due to significant declines in makeready time, sheetfed offset printing is viable for a few hundred copies. There is even a Belgian printing company that puts jobs on its webfed press starting at 1000 copies, one thousand copies, it’s not a typo. Other digital technologies moved upward and became profitable at higher run lengths… Both were filling the gap that Landa Digital Printing was targeting.

Developing digital presses: other examples
To put things in perspective, let’s examine several other printing technologies, digital printing presses, that have been developed over the past few decades. Both successful and unsuccessful ones. I’ll put them in chronological order.

Xeikon
Let’s first look at Xeikon, which was launched at the same time as Indigo and should also share the credit for being the first digital color press, jointly with Indigo. Not my words, but those of prof. em. Frank Romano, who witnessed the entire evolution of digital color printing from a front row seat.

Xeikon was founded by Lucien De Schamphelaere, an engineer who worked at AGFA and developed the first monochrome digital printing press: the P400. And this got him thinking about a full-color digital printing press. In 1988, he founded the company ‘Ellith’ and in early 1989, they started the development of the Xeikon press. Its DCP-1 was launched on June 23, 1993, during a press conference in Antwerp. The period between the start of development and launch is approximately 4 years.

You can find out more about the history of Xeikon in this article by Dutch print journalist Ed Boogaard, which was published to commemorate Xeikon’s 25th anniversary. The first paragraphs are in Dutch, but a little later, everything is in English. From that article: “On the morning of the unveiling and live demonstration of Xeikon’s DCP-1 digital printing press, De Tijd, Belgium’s leading financial newspaper, ran an article stating that Benny Landa had spoiled the Flemish world premiere by announcing his digital press earlier in various US, British, and French newspapers. Lucien De Schamphelaere commented in the same article: “When we read those articles. Benny Landa knew about our invitation and has tried to bring the news first.” De Schamphelaere didn’t think his premiere had been overshadowed by it: “We will be first to go to market and to demonstrate it.””

Frank Romano once told me how Benny Landa got word of the Xeikon event: an American journalist who couldn’t come to Belgium started calling around to find out more (Xeikon hadn’t disclosed anything before the event), and by doing so, Benny Landa learned about the Xeikon event. Since he absolutely wanted to be able to claim he was the first, he issued a press release the day (or a few days?) before the Xeikon event…

Xeikon had a difficult moment in 2002, including a bankruptcy, but the company eventually survived. The toner technology currently used in their dry toner presses is an evolution of the technology launched in 1993.

KBA Karat
An interesting concept from two decades ago was the KBA Karat, a cooperation between KBA (now Koenig & Bauer) and Scitex. They put their expertise together to create an offset press that was always ‘spot on’: you could not control the color on the press. The vision was that color needed to be controlled in prepress, the press was just an output device that printed exactly what prepress handed them over. An excellent idea.
Three elements of the press were crucial: the inking system was an anilox roller, not a conventional inking system, the printing plates were imaged on the press using DI (direct imaging),and it used waterless offset printing.

The cooperation likely began in 1997, with the joint venture agreement being signed in January 1998. The press went into serial production in 2000. In 2001, the joint venture was terminated, and KBA assumed all the activities. But eventually, it was discontinued. I do not have an exact date for that. In 2005, the KBA Karat 46 DI was still listed on the website.

Although it was definitely feasible from a technological point of view, market acceptance ultimately failed. Especially print buyers couldn’t accept a life without press checks, without being able to tweak the colors up to the very last moment, when it was on the press… Eventually, even the press design needed to be changed: a heating element could slightly influence the amount of ink, to accommodate the wishes of the print buyers doing press checks…

However, it also required changes for press operators: it was a waterless offset process and used an anilox roller for ink feed. This anilox technology (which is coming in flexo printing) is still being used in the Heidelberg Anicolor offset press. And its DI (direct imaging) technology, is also still in use it seems, in the Mark Andy Presstek presses. The complete project, from start to termination of the product line, likely lasted approximately 8 years.

Barco Graphics – The.Factory
Barco Graphics showed its webfed inkjet press, The.Factory (pronounced: The Dot Factory), at drupa 2000. If I recall correctly, a few months before that drupa, during a visit to Barco Graphics, I was invited to go down to the basement, where an engineer was working on that press… I found an article from 2000 in a Flemish newspaper that provides information about the development. It began in 1997, in cooperation with Xaar (printheads) and Metronic, a KBA spinoff, for the frame. According to that article, the investment between 1997 and 2000 was between 7 and 8 million euros. The article also mentioned Chiyoda (a producer of ‘decorative printing’) as the test customer. So, between the start of the project and showing the first model at drupa was only three years.

In 2004, when AGFA acquired the then-independent company ‘Dotrix’, a first press was installed at a customer site, Gardner in the UK, which acquired a second machine in 2008.In 2012, however, AGFA decided to discontinue Dotrix and focus on wide-format inkjet instead.

Unfortunately, the market for inkjet was still in its early development at that moment. The Dotrix was a decent machine, but being a public company, AGFA had shareholders nagging about investments that didn’t provide a return in a short timeframe… leading to the discontinuation. And this is something Benny Landa understood very well, that’s why he didn’t transform Landa Digital Printing into a public company. If you are a public company, the market sentiment can significantly impact your future. If you are still in an early stage, without or with only a limited amount of sales, that can be a liability for the future.

Aprion
That same drupa, there was also a new, Israeli player: Aprion. Which did live up to the fail fast mantra. It was founded in 1999 as a spin-off from Scitex, and at drupa 2000, it showcased some interesting concepts based on its ‘MAGIC’ technology (Multiple Array Graphic Inkjet Color). One of them was the ‘Booknet’ (which I covered in this article), another concept was geared toward corrugated packaging. However, I don’t recall that any of the concepts ever reached the market. In 2003, Aprion merged with Scitex Vision, which was later acquired by HP in 2005. As a result, some of the technology that was developed by Aprion may still be in use within HP. But as a separate company, life was short.

Xeikon has another example: it’s Trillium technology.
In case this doesn’t ring a bell, in 2012, Xeikon also unveiled a liquid toner technology, which would eventually be called Trillium. If my memory serves me right, they acquired the IP (patents) from another company, but I couldn’t find any references to who owned the IP before. What I did find is this article, showing “the promise of lower running costs and significant speed improvements, while maintaining the high quality level of its dry toner presses.” Four years later, Xeikon again showed Trillium at drupa, but eventually in October 2017, Xeikon pulled the plug. This is from the press release: “Changing market dynamics but mainly ongoing technical issues in developing the liquid toner technology, including press uptime issues, encouraged us to take this difficult decision, and to communicate it to the marketplace in the straightforward, no-nonsense approach we have always taken.”

A sample printed in the Xeikon Trillium, from drupa 2016.

When I talked about this with a few of the Xeikon VPs some time after that decision, they stated that they noticed that inkjet was evolving much faster, and that in a few years, inkjet would be cheaper than liquid toner. So, they made the rational decision to stop the Trillium program and focus on inkjet, next to their dry toner. And I applaud this bold decision. BTW, if you want to know more details about this Trillium, here’s an excellent article.

Océ / Canon VarioPRINT
A last one I want to mention is the Océ / Canon VarioPRINT, their color sheetfed inkjet press. As a full disclosure: when the project to develop this sheetfed inkjet press started, project Niagara, Océ knew their knowledge of the printing industry was limited. With that in mind, they hired me to do train both R&D engineers and people of the ‘business unit’ To help them understand both the technology used in printing and the market. I think we had half a dozen, or maybe close to a dozen, trainings.

The formal decision to start the development of that machine – after some preliminary tests with the technology – was made in fall 2010. During PRINT 13 in Chicago, there was a teaser about this upcoming product, which generated a lot of interest. The launch itself was early 2015. I checked with one of my friends at Canon in Venlo and at this moment some 725 devices of this series are installed. If I’m not mistaken, the current iX version is the third iteration. (Please correct me if I’m wrong!)

From the examples above, it becomes clear that the two decades of investment in research and development in a new type of printing press is long, very long. However, it’s similar to the time Benny Landa required for the development and launch of the Indigo presses, as I will demonstrate below. And in case you argue that nanography is a completely new technology, well, essentially it’s inkjet with an extra variable: the belt. And as an R&D manager once told me: every inkjet is on a nano scale, otherwise the pigments are too large to jet.

Perseverance vs sunk cost fallacy
Some comments over the past few weeks have admired Benny Landa for his perseverance in chasing his ‘nanography’ dream. Which made me wonder: when does perseverance, essentially a positive trait, become a liability, and should the plug be pulled? Because I didn’t have an answer to that one, I asked Google Gemini.

A few highlights:

Both perseverance and knowing when to pull the plug (to avoid the sunk cost fallacy) are critically important in innovation, and neither is inherently “more” important than the other. They represent two sides of the same coin, and the true art of successful innovation lies in finding the right balance between them.
Innovation is rarely linear: Most truly innovative ideas face numerous setbacks, rejections, and failures before they succeed. Perseverance is the ability to push through these difficulties, learn from mistakes, and keep trying.
Emotional Detachment: Innovators can become emotionally attached to their ideas.9 Knowing when to quit requires objective evaluation and separating emotions from business decisions.
Strategic Pivoting: Sometimes, an idea isn’t entirely wrong, but needs a fundamental change in strategy. This “pivot” is a form of knowing when to pull the plug on the current approach to pursue a new, more viable one.
Set Clear Criteria: Before starting a project, define clear success metrics and “fail conditions.” This helps in objectively evaluating progress and deciding when to pivot or stop, rather than making emotional decisions based on sunk costs.
Regular Review: Schedule regular “pivot or persevere” meetings to assess the project’s viability based on current data and future potential, not past investment.
Seek External Perspectives: Get input from trusted advisors or peers who are not emotionally invested in the project.

And Gemini’s conclusion: “(…) both perseverance and the ability to strategically quit are vital. Perseverance allows good ideas to mature and overcome inevitable hurdles, while the wisdom to pull the plug ensures that resources are not endlessly poured into dead ends, freeing them up for up-and-coming innovations. The most successful innovators are those who master the delicate dance between these two critical capabilities.”

The development of Indigo
It’s not just Landa Digital Printing and nanography that took a long time to develop; the Indigo press also took a very long time. According to this history of Indigo, it started in 1977. In Benny Landa his own words: “With Indigo, we started our liquid toner research where Ken Metcalfe left off – and we got great patents.” Yes, liquid toner wasn’t something new; it was already invented in 1953, and over the years, a lot of companies had printers with liquid toner. This paper gives a nice overview.

Eventually, the Indigo press was launched in 1993, 16 years after the initial announcement. But the press wasn’t ready, not my words, Benny Landa confirmed it in this interview with Globes from 2003: “The product wasn’t ready, it wasn’t sufficiently reliable, and it therefore needed a lot of support.” But, as I already stated above, he always wanted to be the first to launch a digital color press, so when he heard about a big press event happening in Belgium – the launch of Xeikon – he put out a press release to be able to claim he was the first…

And it showed that the Indigo E-print 1000 wasn’t ready; the early buyers weren’t that happy. It required a lot of maintenance, and when your business proposition is that you can deliver print in hours instead of days or weeks, you can’t afford any kind of issues with your press. Additionally, one could argue that the market was not ready for fully digital presses: in those days, memory was still extremely expensive, and the digitization of prepress had just begun.

From an article in 1997: “But something went wrong along the way. The Indigo printers suffered from reliability problems. The company had to bring in one engineer for every two machines sold, clients seldom used their new machines, and potential clients shied away”

The Indigo IPO that didn’t go as intended, late 1995 there was even a class action from investors when shares dropped significantly. And then there was the dot-com bubble… HP invested $ 100 million in 2000 and eventually acquired the remaining shares in 2001. And after a thorough redesign of the print engine by HP engineers, it became the successful product that’s still on the market. (early 2000s, I was involved in the European chapter of The Society for Imaging Science & Technology, IS&T, and I had regular contact with R&D people, managers off a lot of digital printing vendors; someone from the HP R&D team in Barcelona one evening shared insights on the aftermath of the Indigo acquisition, a thorough redesign of the print engine was needed)

In that Print21 article I linked ot above, there is also a very interesting part, given the hindsight we have now:
“AM: So you’re not concerned about the length of time to get Nanographic Presses and NanoInks to market? Will you let some technologies enter the market prematurely?
BL: No, I made that mistake once. In the early days of Indigo, we did not have the money to sustain years of in-house improvement and development, so our factory began shipping presses before they were fully ready. Although the first machines weren’t reliable, we stuck with our customers and made an absolute commitment to “get it right”. They stayed with us because our service technician was with the customer day and night if necessary.”

Why is this important?
‘Kill your darlings’ is the motto of many writers when they need to select what they will publish. However, this also applies to innovation projects, especially when the investments made reach levels that make a positive business case extremely difficult. Kill your darlings closely relates to ‘fail fast, fail often’.

Having a vision is great, it’s what drives innovators and entrepreneurs. And it must be great to work for someone with a great vision, especially if that person is also a natural storyteller and genius marketeer. But you also need to execute. You also need to reevaluate your vision frequently, as market conditions can change. Just check what happened with the makeready in sheetfed offset printing. And you need to act on criticism. Please don’t walk away from it, don’t neglect it. I’ve pointed out issues with the Landa Nano prints as early as late 2020, e.g., the registration problems. It should not be that difficult to fix these, others have succeeded in doing so. But when I got hold of new samples at drupa 2024, nearly 4 years later, I still saw registration issues.

The shareholders made an extremely critical assessment, stating that the company needs another 5 years and an additional 300 million US$. Next week, we might know more about what will happen. The court granted 14 days, this Sunday is the last day of that period. With everything we know right now, the rational thing would be to cut the losses and stop. It was a nice show, the curtain should be dropped. But people aren’t rational…

However… just after I had prepared everything for publishing this article, I checked the Globes website and found out that the court has granted protection untill the end of August. With that knowledge, I also checked the CTech website… Do check this second article, especially the headline and the last two paragraphs: “Landa could become the largest printing company in the world” … [sigh]

Important PostScript: I’ve been critical of Landa Digital Printing from the start, from the moment I watched one of the demos (the small machine near the exit of the hall) for the second time and noticed that the touchscreen changed before it was touched. At that moment, it became clear to me that the machines weren’t real (yet), that it was a fundraising show.

But criticizing Benny Landa was ‘not done’. When I shared my assessment with the editor in chief of a print magazine, he replied: “What do you know about innovation??? This is Benny Landa!!!”. For the record, I worked at the innovation center for the printing industry at the time, already for 14 years. Initially, I held the position of trendwatcher, and at the time, I also combined that with the role of general manager.

When I shared my first analysis of the print quality (the “I’m not impressed” article), many years later, I received a response that I was overcritical and that I wanted to impose a super high ‘Hagen standard’ for printing. That person seemingly hadn’t read any of my posts about brand colors in real life and my quest to get rid of the 2 dE00 fetish…

That print quality came up again about a week ago, when I commented on LinkedIn that, in my opinion, the quality of the machines is the primary reason for the troubles, not the geopolitical situation. This time, I received a reply stating that Landa Digital Printing has been under much more scrutiny than any other digital press vendor. Really??? Please, show me more than 2 articles that take a critical look at the quality of Landa Digital Printing. Other than mine, obviously. The ratio of raving reviews versus critical articles is certainly above 100 to 1, probably even above 1000 to 1.

When I challenged the number of installments last year, I received an irritated reply stating that the numbers shared by Landa Digital Printing were correct, and that the author had seen the list, but – obviously – couldn’t share it. Now we know better: the information in the data room shows that there are approximately 50 installments. Last week, someone with seemingly intimate knowledge checked my list and concluded it was “fairly accurate, with a few discrepancies”.

And yes, I’m not an entrepreneur, I do not have ‘hands-on’ experience as an entrepreneur. I’ve never claimed that and never aimed to be one, my goal in life is different. If I were an entrepeneur, you would not be reading these words for free, I wouldn’t share my insights and ideas for free. I also wouldn’t have worked for industry organizations for almost my entire career (one association and two innovation centers). And I have deep respect for people who have a great idea and do go that extra mile to realize that dream. However, an entrepreneur should also know when to cut their losses, as shown in the examples above.

During my years at the printing industry federation and the innovation center, I’ve been the confidant of many people, including printing companies and vendors; from press operators and R&D engineers to C-level and VPs. They’ve shared tons of information with me, even information that should have been under an NDA but wasn’t. I had the opportunity to visit numerous trade shows and conferences. I know a thing or two about innovation, R&D, and new product launches. I’ve launched several small (and free…) tools myself, projects that made a (slight) difference. For example, low TAC ICC profiles, to facilitate ink savings. Already in 2011 – well over a decade ago – we conducted tests with ICC profiles that had a significantly lower total area coverage (TAC) but without visual degradation. We launched several ICC profiles, including proiles with 260% and 220% coverage. You can still download the 260% from the ICC registry. And Project BBCG is a pretty new approach to defining and exchanging brand colors. And SAD (the ‘Smallest Actionable Difference’) takes a very different look at tolerances in brand color reproduction. These maybe not have made the headlines, but all of them are a new approach, all of them make a difference.

One last thing: I explicitly want to thank all the individuals who shared information over the past few years. I really appreciate it! And if I didn’t publish your information, there is a good reason for that: if I didn’t find any public confirmation, I decided not to publish it, avoiding that it could be linked to you… That’s also the reason why I didn’t publish any pictures I got, e.g., showing issues with the blanket… But I always kept the information in the back of my mind.

Some of the sources were anonymous, but most of them, I know by name, and I checked their profile. And it ranged from people with ‘hands-on’ experience with a Landa Nano press, to people with a much higher profile. I thank them for trusting me, for sharing the information, and for confirming what I had suspected (or feared) from day one.

PS: if you are an editor or blogger and want to use this content, please note that I put a lot of time in my writings, this one is a summary of over 30 years of experience in the printing industry. As with all writings: it’s copyrighted material. If you want to republish: please contact me so that we can make a fair deal on payments. If you don’t want to pay me, only a brief summary in one paragraph, plus a link to this article, is allowed. Don’t be a freeloader and honor this. Even if it’s ‘just post on LinkedIn’.

UPDATE 12/07/2025: Someone who read this post, send me a message. They have tested the Landa machine intensively but came to the conclusion that the quality wasn’t good enough for their kind of work.

UPDATE 13/07/2025:

Someone mentioned that the 1,3 billion US$ isn’t that uncommon, both the Xerox iGen3 and the HP PageWide web presses were over 1,5 billion US$ in investment.

Although the total amount is in both cases significant, it’s not a completely fair comparison: as Morten from Inkish also pointed out, a large part of the Landa S10 is not developed by Landa, they buy parts from others and integrate it. E.g. the frame and paper handling is (largely) from Komori, the inkjet heads are Fujifilm. The ink (obviously) and belt are Landa developments.

Also the timeline is different. For the Xerox iGen3, I found the following timeline: project initiated in 1994, first shown at drupa 2000, first customer in 2001. So 6 years between the start of the project and the first official show.

The information on the development of the HP PageWide Web presses, is limited.

Another comparison I’ve seen on LinkedIn, is with Kodak: they entered Chapter 11 in 2012. Kodak emerged from that situation and is still around. And that’s what brings Landa believers hope. However, the situation is quite different. Kodak had a long legacy and a large product range, in many industries. Landa Digital Printing only has one product (in two versions). Yes, Landa also has many patents, but the amount is probalby much less than those of Kodak. Meaning: the situation is quite different.

(Visited 1,002 times, 1 visits today)