Responsible IT - Part 1

Drivers for rethinking IT

Uwe Friedrichsen

12 minute read

San Francisco-Oakland Bay Bridge

Responsible IT - Part 1

Lately, I attended an IT decision maker conference. A few hundred CIOs and other IT decision makers under a single roof. When looking at the conference schedule, you found the usual suspects:

  • Lots of talks about AI, especially GenAI
  • Quite some talks about modernization of IT and digital transformation ((mis-)using digital transformation as a synonym for IT modernization)
  • Some talks about cloud migration and platforms, complementing the previous topic
  • A few talks about data, data quality, green IT and security
  • A tiny amount of talks about low code and digital sovereignty
  • All that garnished with two or three talks about diversity and ethics, especially regarding AI

Up to this point nothing surprising. But when talking to some of those decision makers and trying to understand what really moves them, a much more interesting picture started to unfold:

  • Almost everyone suffers from extremely complex, technologically highly diverse and outdated IT system landscapes and struggles to satisfy the expectation placed on them and their IT departments while being trapped with that IT system landscape mess.
  • Basically everyone suffers from the intensifying demographic change and the skills shortage based on that.
  • Business-IT-Alignment is still a huge topic and poorly understood. Company decision makers often still consider IT being a cost center and solely measure them by their expenses, pressing them to reduce costs. This adds additional pressure on the IT decision makers, making it almost impossible to meet the expectations placed on their IT departments by the rest of the company.
  • Interestingly for me, nobody really mentioned growing market pressure and dynamics on their own. If you asked the people explicitly, they agreed that their business departments requires more and more features in ever-shrinking time and they are almost always accused of being “too slow”. But overall, it did not seem to be a major concern for them.

It is not the problem that shapes our response to it. It is our perception of the problem that shapes our response. And based on the perceptions I was able to observe, the response patterns discussed at this conference were not surprising for me:

  • Everyone wanted to increase “developer productivity”, meaning efficiency, creating more software in less time – with GenAI of course (even if quite some decision makers still seemed to be a bit cautious about that), low code platforms, COTS software and SaaS offerings, the modern and more fashionable version of COTS.
  • There was the (faint) hope that GenAI will be the silver bullet to solve the skills shortage – and of course to save costs by laying off people. Nobody seemed to find it at least questionable to eagerly replace people with AI solutions if they should get the chance to do it. Quite the opposite, it was openly discussed with a fake tear in the eye – because costs rule and efficiency is king!
  • Modernization of IT landscapes was reduced to replacing existing systems with COTS software and cloud SaaS offerings. The COTS software and especially SaaS vendors who were major sponsors of the conference took painstakingly care that nobody thought beyond that (or at least did not get a chance to say it out loud).
  • Cloud transformation was still lift & shift for most companies. Only a few started to ponder how to actually leverage the extended options a cloud platform offers.
  • Resilience was a synonym for cyber-security (I mention it here because I will come back to resilience later).

Overall, the whole event felt a bit like sitting in a time machine back to the early 2000s and before. People only considered increasing efficiency. It did not come to anyone’s mind at all that improving their effectiveness could offer a much bigger lever. At least I did not meet anyone who thought about potential solutions for their problems from first principles. Everyone seemed to be anxious not to leave the well-trodden paths of the last 40 years.

So much for the IT decision maker conference. Some insights, some disappointment, a lot of disenchantment. I mean, is there not a better way to respond to the challenges, IT faces these days? Apparently, not under that roof.

Taking a few steps back

However, I think, there are much better and (not only in the ecological sense) sustainable ways to respond to the challenges IT organizations face these days. Therefore, let us try to take a few steps back and rethink everything from first principles.

To start with: What are the problems most IT organizations face these days?

From what I observe, there are 4 core problems. We already heard about demographic change and the resulting skills shortage (which is a big issue at least for the Western hemisphere) as well as about highly complex and outdated system landscapes. Those two problems are the obvious one. They are basically understood, even if the typical response patterns are not particularly helpful. But more about that later.

The other two are a bit better disguised (but not a lot). Understanding them makes it so painfully obvious why the usual response patterns are doomed to fail in the (not so) long run.

Digital transformation is poorly understood

The first poorly understood problem are the consequences of the ongoing digital transformation:

  • Digital transformation started in the 1960s, digitizing selected business functions.
  • With increasing compute power and the rise of LAN, in the late 1970s, the digitization of whole business processes began.
  • With the rise of the Internet (specifically the WWW), digitization left the company boundaries. The digitization of the customer and business partner interactions began.
  • Starting in the late 2000s, smartphones reinforced the digitization of the customer interface and interactions massively because figuratively speaking the users did not need to go to the Internet anymore (a stationary computer) but the Internet went with them (as part of their smartphones’ capabilities).
  • In the recent years, whole business models were digitized with the rise of public APIs – think, e.g., Stripe. The whole revenue stream of the company basically goes through an API. Additionally, previously unthinkable business models became possible, leveraging APIs from multiple business domains, leaving behind the limitations of traditional business domains.

The most profound effect of this ongoing development is:

Business and IT have become inseparable.

They are the same side of the same coin.

The other side are the market and customers.

This in turn has several consequences, most importantly 1:

  • You cannot change your business anymore without touching IT. Every new feature, every slight process change requires you touching IT. This means by implication, you treat your business as you treat your IT. E.g., if you treat your IT as a cost center … well, you also treat your business as a cost center because your “cost center” IT limits how fast and good you can respond to changing market needs and demands – like it or not.
  • IT has become indispensable. Switching off IT is not an option anymore. Highly dependable IT systems have become a must. If systems fail, people suffer in their business and private lives. IT organizations carry a responsibility for the well-being of all the people they affect with their IT solutions. They did not ask for this big responsibility, but they have it – like it or not.

From all I see, most people and companies have not yet understood what digital transformation actually means and especially not its consequences. Otherwise, we would not see overly complex outdated IT landscapes all over the place while company decision makers at the same time press to reduce IT costs further and further.

Post-industrial markets are handled poorly

The second poorly understood problem are the consequences of post-industrial markets. I wrote about post-industrial markets time and again in this blog (see, e.g., this post, this post and this post as well as in many other posts). Hence, I will not bore you with discussing it from scratch again.

The core consequence of post-industrial markets is uncertainty. Sales are no longer certain because a surplus of demand as in industrial markets does not exist anymore. Customers only buy where their individual needs and demands are met best because they have the choice. The individual demands are not necessarily the lowest price. It is rather the ratio of perceived value for money where “perceived” is highly individual.

This means, producing as much as possible at the lowest price possible – which was the key to success in an industrial market setting – does not help anymore. The new key to success is responding to the ever-changing needs and demands of the customers as fast as possible. Those companies that trace and respond to the changing needs and demands faster than other companies have a significant competitive advantage.

Figuring out the needs and demands of the customers once in a while is not sufficient because as soon as a new idea enthuses the customers and makes them buy, it almost immediately becomes the new baseline. The novel purchase incentive very quickly turns into an expected feature. Hence, it is always a Red Queen’s race for the next customer excitement feature.

What makes all this even harder is the fact that the customers cannot tell you what they want, what will excite them. They can only tell you if they see it. Also, traditional market research is only of limited use because liking something and buying something are two very different things (remember: customers have the choice and very individual demands). Therefore, companies continuously need to offer their (potential) customers new ideas, knowing that most of the ideas are in vain and only a few will be successful.

This creates a huge uncertainty regarding what to do next, what to offer the customers next, what to bet on next, what to implement next – because in the end all features, IT needs to implement are nothing but (parts of) bets against the market in a post-industrial setting.

Global political and economic developments add more uncertainty and surprises to the mix (definitely for the Western hemisphere). There is climate change which adds more uncertainty how to act in the future and leads to unexpected adverse events.

And then, we have the extreme over-optimization of all production processes to maximize the potential adverse effects of all this uncertainty. In the last 50+ years, we have optimized all production processes further and further to maximize efficiency in order to produce more with less (the industrial market’s key to success).

Highly optimized processes are okay in highly predictable and controllable environments which the industrial production paradigm implicitly assumes. They work nicely within very narrow boundaries of an expected, pre-defined environment behavior.

However, in environments that are characterized by a high degree of uncertainty, this approach takes its toll. Highly optimized systems respond very grouchily, often savagely to deviations from the expected norm and it usually takes them a very long time to resume normal operations.

Let us briefly take the global supply chain as an example of a highly optimized system. Enter the Ever Given: One ship blocks the Suez Canal for a few days and it takes the global supply chains more than six months (some people say more than a year) to recover from this event. It also cost the global economy several billion euros 2. A single unexpected event. This is how an over-optimized system responds to an unexpected event.

Now take into account that unexpected events are the norm in environments with a high degree of uncertainty.

It is well known and understood how to deal best with uncertainty. Unfortunately, most companies respond with “more of the proven”, i.e., they press to further optimize towards more efficiency which only reinforces the problems.

Especially when it comes to IT, it feels a lot like companies respond with a shotgun to uncertainty. They only think in terms of increasing efficiency, of “boosting software production”, trying to speed up the software assembly line, getting more software implemented with less resources (humans also considered “resources”) – a brute force response to the growing uncertainty, true to the motto of industrial production “more is better”. 3

Or to rephrase that overused smart aleck phrase: “Work harder, not smarter!”

Summing up the problems

For now, let us sum up the problems most IT organizations face these days:

  1. Post-industrial markets and the resulting decision uncertainty (which equally affects IT due to the consequences of digital transformation).
  2. Digital transformation and its consequences, the fact that business and IT have become inseparable and especially the indispensability of IT.
  3. Overly complex IT system landscapes, very often highly diverse and outdated, held together with lots of duct tape and kept running with applying WD-40 all over the place.
  4. Demographic change and the resulting skills shortage (which is a big issue at least for the Western hemisphere).

As this post is long enough, I will leave it here for now. In the next post (link will follow), we will look at the typical response patterns, why they only reinforce the problems and try to rethink the solution space from first principles. Stay tuned …


  1. I deliberately left out the consequence that business and IT influence each other mutually, i.e., that it is not sufficient anymore to have a one-way communication from business to IT only dictating IT what to do. Novel technology options in IT change what is possible at the business side. Unless the business side understands what the novel IT options mean for them, they are not able to leverage them. Therefore, the creation of new digital products needs to be a continuous co-creation process of business and IT together (or more precisely: all relevant parties) to create successful digital products. As this need for co-creation is basically already covered by the other depicted consequences of digital transformation, I left it out to not unnecessarily complicate the line of reasoning. ↩︎

  2. I tried to find a commonly agreed-upon estimation of the overall economic losses the grounding of the Ever Given in the Suez Canal caused but failed to find one. However, several articles reported an estimated accumulated loss of more than 10 billion dollars. ↩︎

  3. Maybe you want to remark that I seem to contradict myself because I wrote a few paragraphs before that companies continuously need to offer their customers new ideas, i.e., to ship new features. Well observed! Please bear with me until the second post where I will resolve this seeming contradiction. A little teaser upfront: It does not help to increase the efficiency of the traditional IT feature delivery processes. Instead, you need to rethink those processes and improve how to ask the questions, i.e., how to slice the ideas into a series of feature fragments and how to measure the customers’ feedback and immediately learn from it (see, e.g., this post for more details). ↩︎