“Du bist Direktor eines Gefängnisses im deutschen Kaiserreich. Kaiser Wilhelm hat angekündigt, dass er zur Hinrichtung eines Hochverräters persönlich anreisen werde.
Zum besagten Termin herrscht übles Winterwetter. Zur vereinbarten Zeit ist vom Sonderzug des Kaisers keine Spur zu sehen. Der Zeitpunkt für die Hinrichtung rückt näher.
Da erhältst Du vom Kaiser ein Telegramm, auf dem steht:
Wartet nicht schiessen
Was tust Du?
Dieser Artikel könnte jetzt von der Missverständlichkeit von Führungskommunikation handeln und davon, dass Untergebene immer am kürzeren Hebel sind, selbst wenn es sich wie bei den meisten Managern um Gefängnisdirektoren handelt. Aber darum geht es uns hier nicht. Die Chancen, dass Du das fehlende Komma an der falschen Stelle setzt, stehen 50 zu 50. Und je komplexer und undurchsichtiger die Welt, desto öfter sehen wir uns in die Lage des Gefängnisdirektors versetzt: Wir müssen entscheiden, obwohl es keine Wahrscheinlichkeit gibt, richtig zu liegen, geschweige denn eine Sicherheit.
Eine erste Heuristik, in solchen Situationen zu entscheiden, wäre der kategorische Imperative von Heinz von Förster: “Handle stets so, dass die Anzahl der Wahlmöglichkeiten zunimmt.” In diesem Fall ist es für unsern Gefängnisdirektor besser, mit der Hinrichtung zuzuwarten. Sollte er falsch liegen, hat er zwar den Befehl missverstanden, aber er kann die Hinrichtung nachträglich vollstrecken. Falls er sofort schiesst und damit falsch liegt, kann er seine Handlung nicht mehr korrigieren. Eine Ableitung aus diesem Grundsatz wäre, mit schwer korrigierbaren Entscheiden so lange wie möglich zuzuwarten.
Aber wie so oft bei Heuristiken gibt es auch gute Gründe für das Gegenteil: einen Teil der Optionen bewusst und frühzeitig auszuschliessen. Ich war einmal an einer Veranstaltung, wo Führungskräfte einer Regionalbank über Massnahmen der Strategieumsetzung diskutierten. Ganz zu Beginn stellte der Direktor klar, dass alle denkbaren Optionen diskutiert werden sollen, mit einer Ausnahme: Die Schliessung von Filialen in abgelegenen Gemeinden stehe hier nicht zur Debatte. Was hat er damit erreicht? Auf einen Schlag sind geschätzte 30-40% der Diskussionszeit für andere Themen freigesetzt worden, da dieses hochemotionale Thema soviel Aufmerksamkeit gebunden hätte. Und selbst wenn manch eine Filialschliessung vielleicht sogar eine ganz gute Idee gewesen wäre, so wäre die wirkliche Diskussion dieses Thema ganz sicher mit den politischen Stakeholdern zu führen, und nicht mit dem Management.
Spätes Entscheiden bringt uns also Handlungsmöglichkeiten, frühes Entscheiden Fokus und Speed. Was ist besser?
Und hier sind wir wieder einmal bei der passendsten Antwort für komplexe Verhältnisse angelangt, die nur leider auch so enttäuschend ist: “Es kommt drauf an.” Mary und Tom Poppendieck nennen dies “im letzten verantwortbaren Moment” zu entscheiden.
The most overlooked digital transformation challenge for managers is not so much digitalisation itself, but its secondary effects.
In part one of this article, we have explored the secondary effects of the fact that digital technology rides on the back of Moore’s Law.
In this part, we want to look at effects of the internet, or more precisely, net neutrality.
What makes the internet different from most other means of contact such as shops, post, or phone calls? It is that the cost of physical infrastructure is disconnected from the degree of usage. Network providers charge users in various forms, but most often they pay for access, not for traffic. And even in the case of traffic (Gigabyte of data downloaded), the tariff does not take into account whether the data package in question has crossed half the globe’s net infrastructure, or come from a server just around the corner. While this fact has some effect for the average consumer, it has massive effects for content providers or over-the-top services such as Skype, who in some cases use more internet traffic than many countries.
What does this imply?
The cost of scaling
Imagine you run a local bakery. You are very successful, therefore you like to scale, and open up subsidiaries in other towns. You need to get permissions, rent a place, buy machinery, hire and train staff – before you have sold a single additional brioche. It is the amount of upfront investment that limits your speed of growth. New subsidiaries have to break even, and contribute to your capital reserve (or justify your next loan), before you can take the next step.
Now imagine that instead of bread, you sell an app that allows people to keep their shopping list on the smartphone. If you sell two hundred copies, and want to scale, your only upfront investment consists in advertising. If your app works, you are capable to go from two hundred to two hundred million with extremely little impediments. This is the reason why companies that can reach out over the App Store, Amazon marketplace or the like, touch a much larger potential than those who are bound to other means of contact.
The death of the experience curve
The experience curve, as developed by PIMS in the Sixties, states that in industrial production, every time you double the volume of pieces produced, you have the potential to reduce the cost per piece by about 20%. This is the economy of scales, the advantage of industrialisation and standardisation, the law of the discounter, the reason why big players beat small players in price wars.
For our little bakery, this may be an advantage of scaling up. If you produce your bread in one big bakery and distribute it to many subsidiaries, you can produce it cheaper. But how does this play out for the shopping list app? Your cost of creating the app, and keeping it bug-free, remains roughly the same, whether you sell twenty or twenty million. The marginal cost of selling another million copies is close to zero.
This explains why digital companies, that can grow very quickly, have such an exponential profitability, and are more often valued based on their growth potential, instead of their actual profit. In 2014, WhatsApp famously was five years old, had 55 employees, 450 million users, and was sold for $ 19 000 000 000.-.
For a digital business model, it therefore becomes essential to optimise for this kind of scalability. Facebook, Twitter, Instagram, even Meetup all have an extremely small part of their value chain outside of digital scalability. By comparison, Amazon, Zappos and Zalando have their real-world logistics. Uber and Lift have to manage real-world drivers, complaints, regulators, and they need to quickly reach critical mass in every new city they take on. Most of Bookings.com’s 13 000 employees are employed in customer service. The more the bakeries in your business model decrease, and the apps increase, the higher your potential to profit from digital scalability. This is the direction of the race.
It’s not about size, but dominance
The dominant predator theory states that an ecosystem organises around the species at the top of its food chain – its dominant predator. Check out trophic cascades, e.g. how the re-introduction of a few wolves in Yellowstone National Park created ripple effects down the food chain, until it changed the very appearance of the geographical landscape. If you take this as an analogy for business, you find that in many cases, economies organise around a product that acts as the dominant predator in their field. Often this product is a platform. Many people use Microsoft office not necessarily because it is the best product of its kind, but because everybody with whom they interact and share documents uses Microsoft office. The same is even more visible with platforms such as the App Store: It makes sense to create apps which are marketed on the App Store, and as a consequence are guaranteed to run on the ecosystem of Apple devices. Therefore, many solutions for contactless payment in taxies and small shops, or for augmented reality in industrial contexts, bind their operators to owning Apple hardware.
A consequence is that investors often seek to finance for market dominance, not size or profitability. Amazon continue to get investment to ensure the dominant predator status, although they hardly ever have been profitable yet.
Connected to the dominant predator situation is the rule of first starter advantage. If you are at the beginning of an innovation S-curve, e.g. a new technology, or a new business model, and are ahead of the competition, you have the best chances of becoming the dominant predator. This is so because once an ecosystem has emerged around you, it is almost impossible to remove you from its centre. The race may be fierce for this reason. You may remember the triumph of VHS as the system for video cassettes (another platform). Sony’s BETA system produced better quality video, but VHS won the race to become dominant predator, not the least because they, contrary to Sony, allowed for porn videos to be sold on their technology.
I once had the following discussion: Imagine your business consists in collecting and selling a particular kind of specialist data, and algorithmic evaluations of it. You are the largest player in the market, but under fierce competition from companies that are backed by large investors. Also, you suffer from data theft: many competitors use copies of the data you have tediously collected, because somewhere, dubious clients illegally forward the data to the competition. At this point, here’s a suggestion: make the raw data publicly available for free. Are you shocked by the idea? It does make perfect sense from a digital ecosystem perspective: First, your own profit is made mostly with the algorithms, not with the data itself. Second, the competition can no longer sell the stolen data at a profit, and as a consequence goes out of business. And third, you immediately become the industry standard – the dominant predator.
Now, the VHS example dates from before the times of the internet, and the Microsoft Office phenomenon may depend on email more than the web itself. What the scalability of internet-based services adds to this is the speed and easiness of growth, together with the potential connectivity of user networks. It is in the nature of internet-based ecosystems to form around dominant predators, and it is easier for internet-based solutions to transform first starter advantages into dominant predator positions. Due to the nature of the internet, many new markets tend to be the-winner-takes-it-all-games. In such conditions, it is often more intelligent to become the platform of an exponentially scaling market, than to do everything yourself. Let the users do the growing for you – and profit in their wake.
Platforms – the inequality of mass participation
The scalability of internet connections also means that, like a large flock of small birds can chase the eagle away, it becomes possible to organise things with many small players, which sometimes work better than a few big players. Let us look at one of them: crowdfunding.
If you are a startup in need of investment, you can go to venture capitalists or similar investors. Or you can go to platforms such as Kickstarter, and collect a very large number of very small investments. What is the difference? The few investors have joined your endeavour with large sums each. Even if you are only a piece in their portfolio, they consider their investment substantial, and interact with you accordingly. The crowdfunders may have invested a few dozen, or a few hundred Euro in most cases, and this may have a different impact as to how they think about risk.
However, the most fundamental difference appears at the end of the engagement. If your startup fails, both the investors and the crowdfunders lose their money. There is no difference in how they participate in your risk. But what if your startup succeeds and becomes exponentially more valuable? The investors’ share participates in the exponential growth: they still own a large part of your company. The crowdfunders, on the other hand, get their perks and disappear. The exponential part of your success belongs to you alone.
It is called individualised gains at distributed risks.
A similar situation is created by platforms such as the App Store. For all profitable apps, Apple takes a substantial chunk of their sales. All successful non-profit apps help grow the ecosystem at a very low additional cost to Apple. All failed apps are a loss to their developers, but not to Apple. The same goes for Etsy, Amazon marketplace, Alibaba and the likes.
Service over product
The ease of contact over the internet shatters many of the more brittle forces that create customer loyalty. Think about where you regularly spend your money. Your supermarket and other shops, your car dealer, cinemas and restaurants. Imagine you found a better one, but further away. How much further are you ready to drive to substitute your old choice once and for all? Maybe it’s different for a new car than it is for grocery shopping, but still we tend to let our preferences be heavily influenced by convenience of access: distance, means of traffic needed, time, risk of ending up in a traffic jam etc.
On the web, a less convenient access means a few more clicks.
A side effect of this phenomenon is the rise in importance of user experience design – starting with eliminating unnecessary clicks and keystrokes, for example to re-enter your shipping address, and ending with sophisticated search engine optimisation. But in times when a small shop and a multinational retailer can have the same visibility on Amazon or Ebay, the more fundamental question is how to create solid ties with a customer, that last beyond the single visit or sale. Sometimes this can be achieved with modular products – think how Lego become more interesting as long as a kid continues to buy more toys to integrate with what they’ve got. But the ease of the web is best beaten by the ease of the web: from re-ordering supplies to help with design and application, in the face of an evasive customer’s ease of choice, the importance of additional services to bind them to the supplier has increased fundamentally.
Why did Dollar Shave Club, the no-fuss subscription service for shaving blades, got from 0 to 18% US market share in 5 years, and was sold for $ 1bn? It was not only for its viral advertising on youtube, but mostly because their customer base consists in a growing, multi-million number of – subscribers.
The fragmented supply chain eats the customer
The ease of data exchange over the web is an open invitation to outsourcing. From software development to graphic design to recruiting, many steps in a company’s activities may be executed by third parties. One of the more interesting outsourcing partners is the customer. On the one hand, as Gunter Dueck put it nicely, the digital transformation brings about the end of the consultant-to-the-backside-of-their-screen, i.e. the person who reads out the content of their computer screen to the customer who sits on the other side of it.
On the other hand, with self service come responsibilities. Imagine you want to print a poster in an online print-shop. You may enjoy to preview the format, or a frame. But interesting enough, many providers ask the customer to deliver specific formats (transfer png to pdf), and leave the responsibility for the correct application of printers marks and bleeds, or choice of paper thickness, with the often inexperienced customer (do you know what printers marks and bleeds are? I didn’t.). Or think about the best choice of ticket fare in a complicated ride – Deutsche Bahn are famous for their maze of options. Or who of us knows exactly what we are choosing when asked to make a credit card payment abroad in local vs. home currency? In the name of self-service, and in a careful balance between less tedious access and more value-adding steps in the workflow, today’s online customers are ready to take over an ever growing part of the remaining work.
The inherent libertarianism of platforms
Some situations have an inherent bias. In the world of work, this bias plays out in favour of internet-based business, simply because connections between demand and supply are easily established, and can easily circumvent impediments which appear outside the web: physical distance, average wage level, regulatory environments. The ease of global reach through the internet has created occasions for business models to bypass existing organisations by connecting demand and supply on a global basis, thus creating micro-entrepreneurs in a libertarian marketplace. Think Uber instead of taxi drivers. AirBnb instead of hotels. TaskRabbit instead of movers. Upwork instead of admin employees. Ebay vendors instead of shop assistants. Etsy instead of gift shops.
But it is also a great entry for the secondary effects we have observed about platforms. Platforms have the tendency to individualise gains and distribute risks. Globally mobile digital services are flexible to choose preferred regulatory environments. What are inherent biases of such a system? The five-day-work-week micro-entrepreneur competes against the seven-day-work-week micro-entrepreneur. The one with health insurance and pension plan against the one without. The one who obeys safety regulations against the one who doesn’t. The inherent bias of the digital transformation points away from what many of us, at least in Europe, consider achievements of civilisation in labor protection, and towards libertarianism. At the other end of the stick you will find a growing income gap between those who offer their services on platforms, and those who own them, and as a consequence, a growing precariat.
Where does this put corporate social responsibility in your digital business model?
The most overlooked digital transformation challenge for managers is not so much digitalisation itself, but its secondary effects.
Of course, digitalisation proposes a host of opportunities, and managing them is not easy. How to use cloud, big data analysis, AI, predictive maintenance, industry 4.0 etc. are big challenges for themselves. Even things like creating a unique platform where the customer can access all your offerings online is sometimes quite difficult.
But once you look at what secondary effects emerge because of these changes in your organisation and its environment, you realise that the foundations of almost everything you know are profoundly shaken.
It has mainly to do with two aspects of digitalisation: Moore’s law and net neutrality. In this article, I am going to explore implication of Moore’s law. Click here for the second part.
The first source of change: Moore’s law
Moore’s law is not so much a law, but the statement that every 18 months, the capacity of a microprocessor doubles, and its price is cut in half. This has been true from the 1960 to this day. It simply implies an enormous speed of exponential innovation. If the same speed of innovation were applied to air travel from 1970 to today, a flight from Paris to New York would cost less than a penny and last less than a second.
So just by coincidence, everything connected to digital technology sits on the back of an exponential accelerator of innovation.
What are the implications of Moore’s law?
Exponential growth – outside in
Curt Carlson, at the time CEO of SRI international, the most productive developer of the foundational technologies of Silicon Valley, has famously stated:
In exponential times, if you improve your performance incrementally, you fall behind exponentially.
If you focus on the history of your own organisation, and compare its present state to where you’ve come from, a continuous past of incremental improvement will make you proud. Every year you’re better than before! Only – the speed of improvement around you is faster, and if you miss the early signals, or miss the capacity to accelerate with it, you’re gone before you know what hit you. It also means that the legacy (old contracts, old technology) of established players is not as big an advantage as it may seem. Sometimes, this not only applies to companies, but to whole industries.
The speed of commodity
The speed of innovation also means that novelties become common very fast. This relates to hardware and software. Take for example space technology. In classical space technology, there is the paradoxical saying “Only use in space what has been proven in space.” This makes sense, since it is tremendously expensive, or impossible to fix things in space. So you won’t risk a billion-dollar satellite by using technology of which you don’t know yet whether it works in space. The consequence is that satellites run on very old technology. Meteosat (2nd generation), has taken 20 years of development, there are currently 4 in orbit, and they cost 1 Bn € each. The capacity of their in-board camera is to cover 1 square km with 1 pixel – that was what the technology was capable of 30 years ago.
By comparison, the start-up Planet Labs has built a satellite for a similar task: take 1 picture of all the Earth’s surface per day. Their “Doves” (you can see one in the picture above) were built on a completely different combination of technology, price, and quantity. They use gyroscope and cameras from current smartphones – cheap, modern, but not tested in space. Since they cost less than 200 000 € (estimate), it was possible to send 149 into orbit – one batch at the time. Price and quantity were the condition to include one or two experiments with each batch – if some of them break down, it’s not the end of the mission. Relying on existing, commoditized technology, development of the first satellite took 3 years. The quality? One pixel covers 3 square meters on Earth – no surprise with today’s Smartphone camera technology.
This means that the option to piggy-back on a commodity instead of an expensive customised or niche product, has become much more widespread and powerful because of Moore’s law.
The substitution divide
Many products include digital as well as non-digital parts. Since the digital part substitutes faster, and is usually also more connected to a digital ecosystem which substitutes faster, it can affect the whole concept of product life-cycle.
After two years my car is old because it cannot stream Spotify.
Shift left on the S-Curve
The S-Curve is the form of development of most technologies and markets over time. Depending on the phase of the S-curve, different rules apply to competition. At the beginning, it is about attracting interest with innovation. In the steep part, it’s about expanding the attention with marketing, and ramping up production. In saturated markets at the top of the curve, it is about optimising cost, and the big players are in a better position to survive a price war because of their potential to produce at lower cost. In the old world, that was usually when a new technology came in, and the question was who had jumped to the new S-curve early enough.
In the digital world, technologies substitute each other quicker. This often implies that the new S-curve starts long before the old one has reached saturation. The relation between the investment into fast expansion (the steep part of the curve) and its pay-off in the saturated market has fundamentally changed.
One example of this is the music industry. The change from vynil to CD was still “old world”. Now, CDs still are still sold at 50% of their peak in 2000, while download (such as iTunes) is already being overtaken by streaming (such as Spotify).
Attention to customer value
The customer needs, their changes, and the discovery or creation of new needs prove to be more than ever the most valuable compass for the direction of product or service strategy. With digital innovations impacting the customer’s ecosystem with such speed and complexity, the relation of your offering to the customer’s needs is a constant source of surprise.
When the video games company Valve published Team Fortress 2, they had introduced the option for players to create their own hats for the characters they played. Within weeks, the players not only did that, but started trading the hats with each other, exchanging thousands of dollars in the process. For Valve, who had not foreseen that, the option was to immediately design a marketplace for this trade, or loose the connection to that part of their customers’ activity. As their CEO Gabe Newell put it:
Your fiercest competitor is not the other games producers. It is your players.
The amount and lifespan of options
Buy one now? Wait for the next?
Every decision for an asset is a decision against other options. This fact alone is vexing. Whenever you set on one technology, you exclude some features which are only available in other technologies. In a slow moving world, this decision was between the alternatives on the table. You could analyse your options, and then decide. The faster the technological substitution, the more your decision is against alternatives that are just around the corner. Think about buying a computer. If you are looking for an experience which guarantees to ruin your day, just visit a shop a couple of months after you’ve bought a new laptop, and see what you could have gotten now. In retrospect, it would always have been better to wait – but you cannot wait forever. In terms of complexity theory, the gap between premature conversion (closing your search too early) and eternal boiling (closing your search too late), has become narrower.
The same intensified dilemma applies between alternatives which are not fully developed yet. Think about the car industry’s decade-long indecision between the fuel cell and the battery as a substitution for the combustion engine. The digital world is full of such phases of indecision. You cannot wait until the technology, infrastructure and market demand are fully developed. You have to act, knowing that part of what you do is in vain.
“Due to the exponential innovation in tech, the rationale of digital business models is in constant change. The bottleneck is not technology, nor money. It’s the human side of the organisation.”
In cyber-social systems, it is the social which lacks behind constantly. Business models, strategies, organisations and skills are in constant change. This means a shift of management attention between running the operational business and developing it. It means an increased weight of everything bureaucratic, everything difficult to change. Like so many others, automotive is a whole industry concentrated on optimising for low cost of production. Compare this to Wikispeed, or the Saab Gripen plane, and you see a new paradigm coming around the corner: optimise for low cost of change, so you can incorporate innovations in every small iteration.
It also means people’s capability to learn, and time dedicated to it, is put under pressure – think about the divide between young and old employees. It also means that if you hire a new person, their capability to fill the vacancy at hand is still important, but their capability to be useful at different, unknown positions within the company gains in relevance, because chances are, the vacancy at hand will not exist in a few years from now, and you cannot afford to replace the majority of your workforce every few years.
The changing relevance of relationships
Last but not least, the speed of digital innovation implies that a growing number of things are either automated, delegated to highly specialised departments, or outsourced to suppliers.
In many U.S. courts, judges use algorithms to indicate a defendant’s “risk”. Based on this risk factor, they may increase bail, sentence, or parole terms. These algorithms are not transparent, but have proven to be very accurate when checked against past cases – reason enough to start using them, isn’t it? The judges were very surprised to learn that on thorough inspection, one of the things these algorithms did was pure racial profiling. Just for growing up in a African-American neighbourhood, the risk index attributed to you went up considerably.
What does this mean for other contexts? You find an increasing number of people who yield the results of something digital they don’t really understand – starting with you as the chief generalist of your business model. The pressure on trust and reliability of your relationships has just gone up – not because you lack transparency on what is going on on the other side of the interface, but because albeit the transparency, you don’t understand what it means, and have to rely on the other party’s word.
The second part of this article takes a look at the effects of the internet, more precicely, of net neutrality: click here.
Imagine you’re sitting in a restaurant with an important client. She has given you the wine list, but courtesy demand that you will make a suggestion, and ask if she agrees. Now you are a big fan of Californian wines. And on the list, you see a great wine from Nappa Valley. A blend of Cabernet Sauvignon and Merlot – you really want to try it. But you know from previous conversations: your client is not a big fan of wines from the “New World”. She likes European wines, and especially Bordeaux – in her words the apogee of wines. What do you do?
You have two options. Either you try to evangelise your client on Californian wines, to convince her that her view is too narrow, and that she should give it a try. Or, you focus on another aspect of your suggestion. Bordeaux wines, too, are made of a blend of Cabernet Sauvignon and Merlot. So you say: “I remember you like Cabernet Sauvignon and Merlot blends?” “I love them.” will be her reply. “I have here a wonderful blend, from 2010 which was a great year – shall we take it?”
What happened? Here are five observations.
First: Whenever we argue something – propose, defend, legitimise – it is less the facts that influence people’s opinion, but the aspects, or perspectives on the facts. Already Epictetus knew that. To each fact, there are many perspectives, so we have a relative freedom of choice. You may remember the Greek government-dept crisis in the aftermath of 2008. The essential facts on the table were not the point of dispute, but the perspective from which people looked at it. Was it a crisis of the greek population, their poverty, their loss of employment? Was it a crisis of the Greek government, their decisions and their struggle to assert themselves between democratic legitimacy and the pressure from abroad? Was it the crisis of the Euro? Was it the crisis of the European Union and its future? Or was it the crisis of foreign investments in the country? Depending on our assumption about which perspective is relevant, the same facts lead to fundamentally different conclusions. Whatever our choice of perspective, it is not the facts that suggest one perspective over the other, but our values.
So, coming back to our choice of wine, you can choose various aspects with which to “sell” your preferred wine. The wine itself does not tell you which of them you should choose.
Second: Many interactions with other people happen in a context of pragmatic interdependence. Pragmatic interdependence means that we have a goal, but in order to reach it, we depend on somebody else. Our interaction includes a claim, e.g. for agreement, contribution or tolerance. And our success depends on whether our claim is accepted or refused. We want a certain wine, but we need our client’s agreement. We want to start a project, but we need our boss to approve the budget. Even something as simple as a welcome includes claims. We extend our hand, and we implicitly claim “please shake my hand”. This looks very self-evident, but imagine you just had a row with that person, now you meet them again in front of important people, and you extend your hand. The claim now is “please join me in a gesture of reconciliation”, and you charge your claim with situational power, because now, the alternative for the other person would be to very publicly be rude to you – which may be not as far as they want to take it.
So the shaping of acceptable claims is part of our tactics in contexts of pragmatic interdependence.
Third: You cannot evangelise and sell at the same time. In situations of pragmatic interdependence, we want to make our claim acceptable, to “sell” it. If at that moment we choose to try to change the other person’s values, they tend to see all our “missionary” arguments just as an instrument to our salesmanship, and therefore won’t take them seriously. A missionary is only good as long as he doesn’t need to sell anything. As Castiglione observed in his Libro del Cortegiano in 1528: In social interaction, as soon as a technique is seen as a technique, it doesn’t work anymore.
Coming back to our context of pragmatic interdependence, where we have to “sell” our claims by choosing certain perspectives or attributes, this means that we will choose those perspectives which appeal to widely accepted values. In short, when we are pragmatic, we are conformist in our choice of values.
Fourth: Therefore, in a context of pragmatic interdependence, the more a value is being used, the more it is seen as being widespread. The more widespread a value, the more it is being used. It is like a deer path in the woods. The first animal to cross the wood has ample choice whether to pass this tree to the left or to the right. The next animal that comes, sees a slight trace where the first animal has passed – it has a 51% preference to follow in those tracks. With each animal that passes, more dead leaves get trampled on the ground, twigs are broken, young plants don’t grow in the path. After a thousand crossings, the path has become very evidently the easiest and fastest way across the wood, and no-one would think of passing the tree on the other side. The same with values. Think about evergreens such as customer orientation. Or cross-functional collaboration. Or attention to cost. Those values have the best chance to be strong in the future, which are strong now. And this is true with no regard as to the reason why those values have become strong in the first place. In complexity science, this is called an attractor basin. When a value is strong in a group, people use it in order to sell their claims to each other. This means that whenever we see strong cultural habits, and values which seem irrational, it is worth looking into the past, or to the local perspective, where we often find very rational reasons why they became strong – only the reasons have disappeared, while the value remains strong.
An example? BMW are a very successful company with some strong cultural habits. One of them is that every decision, before it is made, needs to be circulated and discussed with many people. If some have been excluded from the consultation, it is well possible that they disagree afterwards, and this can in fact lead to a re-elaboration of the decision. As one head of the board once said: he couldn’t implement a decision to introduce Six Sigma and have everyone schooled within six months – like Jack Welch famously claimed for General Electric – because their culture would not allow it. Now this habit, and some others with it, originated during a great trauma, when the company was in a huge crisis and nearly got sold to a competitor, and was saved by, amongst others, a great union of management and workers. The latter allowed for heavy cuts in their wages, but got a bigger place at the table in turn. When was that? In the 1950ies. Nobody who underwent that trauma is still in the company. But the culture is there, solid as rock.
So, strong values remain strong, because we tend to use them more often for pragmatic reasons.
Fifth: Culture therefore tends to be more conformist, or unified, if there is more pragmatic interdependence. If we have a group of salespeople who all need to achieve their individual targets, but don’t need to collaborate, and don’t need to convince anybody in the company to accept their claims, there are few occasions for strong conformisms to grow. Taking them out to a teambuilding exercise every year doesn’t change that. If, on the other hand, people depend a lot on each others agreement, judgment, and reputation, then the culture is strong. This information is key if, from a management perspective, we want to influence culture through structural measures. Make pragmatic interdependence strong where it depends on certain values, and these values become strong.
So, structural reasons for pragmatic interdependence reinforce culture.
In 1999, Frederic Vester published a report to the Club of Rome named “The Art of Interconnected Thinking”. The main focus of the book is about understanding complex systems, and how a number of interconnected models, what he called the Sensitivity Model, can help us do so. The Sensitivity Model is an IT-based approach, today in the ownership of Malik Management. While other IT-based approaches try to connect some 200+ variables into a database, Vester is frugal in comparison, with 10-20 variables. The advantage of his approach over the more mathematical siblings is the acceptance and use of fuzziness. We simply cannot expect to be able to get a total picture of our system with sharply differentiated concepts and mathematical variables, so stop trying to do it anyway. The consequence is: we better accept that whatever model we use, it will be incomplete and partially wrong. It would be foolish to attempt something that is 100% correct. Therefore, a more realistic ambition is to create a model which is relevant to the pragmatical perspective of the beholder, and is sufficiently apt to produce this relevance.
“The map is not the territory.”
— attr. Alfred Korbynszky
I have facilitated the creation of sensitivity models, and they still take a group at least a day to construct. With this in mind, I would like to propose something more pragmatical: something that you can do alone or in a small team, without external assistance, in less than half an hour if need be.
If we take a blank sheet of paper and try to visualise a complex system, most often we start writing down some elements of the system and connect them with lines. Often, these lines become arrows, and we end up with an influence map, or, as many call it, a spaghetti diagram. Unfortunately, just like spaghetti on a plate, the spaghetti diagram has a practical fault: in most cases, if we think about our system long enough, we end up with something like this:
Like Alice’s rabbit hole, it draws us in, and we can spend our time following the lines, but our most significant learning from this reverie is something we already know from the start: it’s complex.
In comparison to the spaghetti diagram, another model from Vester’s treasure chest, the role distribution map is less intuitive, but vastly more helpful from a pragmatical point of view. What is our interest in modelling our complex system? We want to learn more about how we can relate to the system and its various elements, what we can do with them. Can we use them as a leaver? What are the chances and risks of doing so? What are elements that we should watch with a careful eye? What are elements that we can happily ignore? In order to answer these questions, it makes sense to think about the role these elements play within the system.
A role distribution map is the visualisation of two questions: To what extent does an element influence the system, and to what extent is the element itself influenced by the system? Notice the fuzziness of the word “influence”: is it a specific cause-effect influence, or a diffused, unspecific force out of which something emerges? For the sake of this model, the answer to this question is of no importance. What is more important is the collection of essential elements, and the holistic assessment of their influence. With the answers to these two questions, we can place our essential elements on a role distribution map.
As a result of these reflections we have now identified elements which are a) critical factors of the auto-dynamic nature of the system, where external manipulation may easily trigger off surprising side-effects, b) reactive elements which may indicate the state of the system, but do not wield any power, c) buffering elements which contribute to the system’s stability, and d) active, i.e. relatively independent elements which actively influence the system. It is this last group which interests us most if we want to influence the system: If we can place a modulator on some of these elements, we have found the easiest way to make a difference.
This method, along with others, is part of the Palladio training program Beyond Bureaucracy: Navigating complexity with the Amber Compass. Click here for more information.
The German Sociologist Gerhard Schulze uses a wonderful analogy when he describes societies’ capabilities to deal with crisis, and I want to adapt the picture to organisations facing complexity. Most of us may not have heard of the Polynesian Migration. Within a timespan of about 800 years preceding 1300 AD, the Southern Pacific area was populated, starting in South-East Asia, and ending up in places as far away as Hawaii, the Easter Islands and New Zealand. The spread from island to island happened in individual steps – sometimes more frequent, sometimes less – across the centuries. For our purpose it is useful to note that during this period, society had something like two states of existence: State A consisted in the expoitation of existing resources on a given island: Societies lived together in buildings, they took food from land and sea, the basic events of life happened, such as birth, death and several rites of passage in between. For all this to work, there was a division of labour with appropriate tasks, roles and rules.
Then, with the population growing on a given island, at some point the density came to be felt as pressure, and society switched into a state B: Young men took their ships on exploration to ever further areas of the Sea, some of them never came back. Those who made it back home, evaluated what islands they found as potential destinations, maybe comparing alternatives. There might have been wars with other settlers. Then the population on the island split up in those who stayed, and those who left, a process which might have started as early as the beginning of state B – we may imagine an ongoing debate between “we must go and seek” vs. “it is useless to go and seek”. For those who eventually left, a dangerous journey began, in open boats across the long distances of the Pacific, towards a tiny target difficult to spot. Once arrived, they had to build and organise in temporary arrangements, until at last they could switch back into a new state A.
Now it is obvious that for these two states of exploitation and exploration, rules, priorities, criteria for decision were fundamentally different, as were the kind of people who excelled at the task.
If we take these two states as analogy for an organisation, the exploitation state A consists in administering operational business, taking the daily decisions, dealing with incidents, and incrementally improving the status quo. State B, the exploration, would be the reorganisation, the development of a new strategy or business model, and innovation or skunkwork projects. People spend all their time in workshops…
Here’s why I believe that it is important to remember this analogy today:
The more complex and dynamic your environment, the more it is the capability of an organisation to master state B, and to switch between the two states, that defines its competitive advantage.
First and foremost, Agile is one of the most radical hijackings of human language. “Oh you mean agile in the sense of nimble? That’s not what Agile means. Agile means instant, frictionless and intimate delivery of value.” Steve Denning told me in a recent conversation. Problem is: First, Steve’s right. Second, most stakeholders of Agile organisations haven’t understood that yet. Third, you can’t blame them, because confusion is what you get when you hijack an existing word for something different. Fourth, the confusion is cunningly hidden by the fact that to some extent there is an overlap between Agile and agile: both include in their meaning the capability of an organisation to deal better with complex, unpredictable, shifting environments. But soon after, the overlap comes to an end. Both for Agilists and their stakeholders, that area would be a healthy focus of learning.
Yesterday morning, my daughter asked me to help her with some last moment review for a maths exam. It was about solving word problems. The calculations themselves were too difficult for the kids’ level, but you were allowed to work by approximation.
We both found that one of the most difficult things about this was to be able – without knowing the result of the precise calculation, nor the way the calculation needed to be done – to guess how much rounding would be adequate, and when it would throw us off course so much that the result would be marked as a fail. So here’s the complexity challenge for you: Sometimes, instead of getting lost in all the details of the trees, it is worth taking a step back and having a look at the wood, knowing that you will miss out most of the information on the tree level.
On a physical level you can try this with the famous rasterized image of Abraham Lincoln: If you squint, you see the face much better. How could that work on a practical level? Here’s an example: We all know the meaning of active listening, or empathic listening, in what kind of contexts it would be a good thing to do, and most of us would think of themselves that they should be doing more of it. But how do you learn it?
There is the approach on the tree level. Make a list of the activities that this includes: Lean forward. Look at the speaker. Mirror body language. Confirm with nods, facial expressions or short statements. Paraphrase what the speaker has said. Ask open questions. Etc. etc. And in each context, some of these things may be inappropriate, others would be helpful that are not on the list, and you can probably do any of these items in a way which is awkward and produces unwanted consequences.
What if instead of making this list ever longer and more precise, we take a step back and give ourselves another kind of instruction: Instead of listening in order to reply, listen in order to learn. Or: Try to understand and feel what the world looks like from the speaker’s perspective. You don’t describe any precise activity. But you can be sure that each person who follows this general imperative, will practice active listening in a way which is both authentic, and adequate to the situation. The “imprecise” instruction works better.
Remember your last New Year’s Resolution? To quit smoking? Spend more time with your family? Finally do something about (insert imbarassing vice here)? The truth is, we often didn’t act on them, because when the moment came to do something, we just happened not to think of our resolution.
It’s the same with many good leadership behaviours. We understand they would improve things. We really want to do something about them. But we cannot plan them, because the next occasion to try them out is unpredictable.
So here’s a trick: Put a small blackboard or something similar somewhere either at work or on your way to work. Write down your Motto of the Moment, so you get reminded of it at least once a day. After some time, choose another motto.
In the comment section there is a collection of mottoes from Palladio customers and friends. Feel free to contribute!