Skip navigation

I've been busy this week...setting up my own supply chain. What...have I suddenly started buying and selling goods, you ask, but no, I'm talking about the supply chain of my main blog or website, While it has been more of a technical than a logistical challenge, the paralells to real-world logistics is strikinggly similar, as you may discover when you read on.


A years ago or so I wrote about the supply and demand side of a blog, and even set up a model framework for the supply chain of a blog. However, focusing primarily on the chain itself, I did not pay much attention to how the outgoing flows were delivered. In logistics lingo, I did not pay much attention to the carrier. I should have.


The delivery network is as important to a blog as it is to any business. It is important to have the right infrastructure, and what matters most as in much of real-life logistics, is delivery speed. Of course, initially, it may seem that the delivery speed for blog content depends primarily on the infrastructure on the reader side, i.e. the individual Internet connection, the current Internet traffic in the vicinity of the reader, the reader's device (PC, Mac, iPhone, Blackberry or whatever) and even which browser the reader is using, but not quite. In fact, it also depends on how I serve my readers.


Now, very few bloggers will use 1PL, that is serving the blog from their own computer in their own home or office. Most bloggers will use 2PL, that is hosting their blog with some webhost, where the actual blog resides somewhere on some server, most often in the US, as it is in my case. More correctly perhaps, this is 3PL, as you have little or no control or say in type of server or Internet connection from that server to the reader. In logistics terms this is a centralized warehouse.


As a rule, your server will also host many other blogs, so you will have to share resources with those, and if they have more traffic than you do, your 3PL, is likely to give them more priority, or you will at least have to queue until it's your turn to send out what the reader wants. This can be abated by securing a dedicated server that only hosts you, making sure that you are the preferred customer. That way you can make sure that you will not have to compete with others for the server's scarce resources. Ok, done that, my warehouse is now my own and not shared.


The reason why this is important is that a website served from a server is built on a Make-To-Order basis. The reader asks for a certain post or page, and the server assembles it on the fly. There's nothing on the shelf. However, things can be improved by using Make-To-Stock. In blogging lingo that is called caching. A cache will essentially preload or preassemble your posts, and when a reader requests them they are sent off in one go, much faster than making to order. That is very useful since the server doesn't have to assemble posts again and again, but can simply deliver what is in the inventory. Ok, done that, but still not happy.


Coming back to speed and the reader side, while caching delivers content en bloc, and thus improves reader experience, if the reader sits far from the server there will be some lead or lag time, even more so if his or her Internet connection is not up to par. While my post travels on a ten-lane freeway while still in the US, in some remote village in India, it may have to follow a narrow and winding dirt road. That said, considering Internet traffic, an empty dirt road can be much faster than a clogged freeway.


This is where decentralized warehousing can help, or in blogging lingo, Content Delivery Network, or cloud computing to use another word. Not only is it possible to have my posts made to stock, I can also predistribute them as close to the intended reader as possible, which means they may not have to travel the whole globe before reaching my reader, but can take the shortest and fastest path. This ensures timely and reliable delivery of my blog posts to any reader, wherever he or she may be. Ok, done that, and is it any good? Yes. In fact, it may even save me from a supply chain disruption, should my web host go down. Using a Content Delivery Network or CDN thus ensures flexibility and agility.


Is this 4PL or still 3PL? Not sure. What's your say?


So that is what I have been experimenting with this week, make-to-order versus make-to-stock and centralized warehousing versus decentralized warehousing. Judging by the performance meters, the latter won in both cases, and while my US readers on a broadband connection are likely not to see any difference in speed of my blog, I am sure that it will matter to my far-and-away readers. Because, in the end, like all businesses, I too must evaluate my customer base, i.e. reader base, and make sure that I serve them the product they ask for when they ask for it. Looking at my reader base, not all of them are in the US, but many of them are actually in the so called emerging countries, and I want to serve them just as well as I serve my American clientele.

Cocoa trading at a one-year high...a nuisance, but not worrying. Coffee trading a 13-year high...ouch. That'll hurt Starbucks and alikes.Sugar trading at a 30-year That's definitely going to have some impact. In much of today's food chains, particularly in the modern and industrialized world, these are three staple ingredients, and if these prices go up, that will feed through the entire food supply chain, let alone the economy.


You could blame recent disasters like the Australian flooding for the hike in sugar prices, as was debated in a lengthy post titled Queensland could transform our view on commodities that I found on the Tranformational Logistics blog, but disasters decimating crops are only part of the picture. More to blame (although "blame" is  obviously the wrong word here) are the emerging economies and their  rising demand for food and energy. Imagine, in 2010, for the first time, more cars were sold in China than in the US. That hasn't happened since Henry Ford introduced his Ford Model T some hundred years ago.  And China has now overtaken Japan as the world's second largest economy  and is poised to surpass the US in 15 years or so or maybe even less.  India is the runner up and may overtake  Japan in 25 or so years. Now,  what will that do the world's food and energy reserves?


I guess it's no coincidence that the soft commodity crunch is making rounds on CNN's Quest Means Business these days. In fact, CNN has just launched a year long series titled "The Price We Pay", where they will examine in-depth the factors and direct impact of global price instability around the world and how the growing imbalance of supply and demand will affect each and everyone of us. Here are some interesting videos:


- Richard Quest explains the price we pay

- Why food prices are rising

- Food prices - rising or peaked


Looking at the World Food Situation index provided by the UN or FAO to be exact, the picture is grim: Food prices have more than doubled since 2002. If I remember correctly, here in Norway people spend around 10-15% of their monthly income on food. Double the price, and the impact is still manageable. In emerging or developing countries, it is estimated that people spend more than 50% of their income on food, in remorter areas perhaps even more. Double the prices here and you have a potential crisis at your hands, with Tunisia and Egypt being the first countries to illustrate what could happen.


Fortunately for the emerging Asian market is that the price of rice is only half of what it was during the peak of 2008. Interestingly, if you look at it in a longer historical perspective, the rise in food prices is even more frigthening, looking perhaps a bit like Al Gore's hockey stick of global warming:



Is there an end in sight? In my personal opinion, no. I think that the food supply chain and the security of supply will be one of the hottest supply chain trends in years to come. France is already putting rising food prices at the top of the G-20 agenda, the Wall Street Journal reports.

Jim Fulcher had an interesting blog post on the Supply Chain Expert Community the other day, when he wrote about Volatility, uncertainty and their effect on the supply chain. In his post, Jim referred to the Tompkins Supply Chain Consortium survey, which said that companies in the future must become highly flexible while remaining efficient so they are able to react to changes quickly and efficiently.


This reminded me of a journal article that came across my own desk recently: "Supply Chain 2.0": managing supply chains in the era of turbulence, by Martin Christopher and Matthias Holweg.  The paper questions the established approach of supply chain management and argues that in the  light of increasing turbulence a different approach to is needed. What is needed is an approach that builds structural flexibility into the supply chain. Many companies only appear to have flexibility in their supply chains, but this dynamic flexibility is very different from the structural flexibility that Christopher and Holweg advocate in their article.


The traditional approach to variability and uncertainty in the supply chain has been to add measures of control and stability. This has given rise to operational practices such as Lean, Six Sigma, push-based production, vendor-managed inventory and many more strategies. Firstly, although seemingly effective, these are very specific strategies geared towards certain industries and/or certain products, but what is needed, so the authors say, is a generic strategy that anticipates, rather than reacts to, turbulence. Secondly, although many companies have managed to build dynamic flexibility into their supply chains by being able to quickly reacting to changes, these are more like contingency plans than  actual structural changes to the supply chain.


Structural flexibility implies the use of the following strategies:


- Dual sourcing

- Asset sharing with other companies, including competitors

- Separating base demand from surge demand

- Postponement of final product assembly

- Flexible labor arrangements

- Rapid manufacturing of small batches

- Outsourcing and contract manufacturing


So what exactly is structural flexibility? Well, according to Christopher and Holweg it refers to the ability of the supply chain to adapt to fundamental changes in the business environment. This, so they say, is done by first and foremost designing the supply chain for flexibility, not for efficiency, which is the case with most supply chains today, where flexibility is a sort of added option. In a structurally flexible supply chain, flexibility takes center stage, and efficiency becomes the second violin. I find that an interesting thought.



In the efficient supply chain, the focus is on control and the aim is to reduce too much variability. In the adaptable supply chain the focus is on volatility and the aim is to develop a superior ability to adapt. The former sees turbulence as bad for business. The latter sees it as unevitable and thus creates adaptable structures to accomdate it. While the efficient supply chain tries to eradicate turbulence, the adaptable supply chain embraces it and works with it.


I think Christopher and Holweg are on the right track here. We should not fight volatility, we should perhps not welcome it, but we should not view it as an enemy. Rather, it is an opportunity. And no, I am not thinking about the Chinese sign for crisis here, because contrary to popular misbelief, it does  not mean opportunity. The most important lesson for business concerning volatility is that it challenges the mindset of economies of scale as being the most efficient manufacturing strategy. It is most certainly not the most efficient strategy for dealing with volatility. Responding to volatilty requires small numbers and the ability to fill niches. It may appear more costly, but in the long run, the return on investment will pay off.


Reference: Christopher, M. and Holweg, M. (2011). "Supply Chain 2.0": managing supply chains in the era of turbulence. International Journal of Physical Distribution and Logistics Management, 41 (1), 63-82.

How much product variety is too much product variety? That is the question in a recent article in the International Journal of Operations & Production Management titled "Theoretical versus actual product variety: how much customisation do customers really demand?" Here the authors challenge the common claim of "infinite variety" being demanded in the  marketplace, by measuring not just how much variety theoretically could  be produced, but also measuring how much variety that is actually demanded by the customer.


Take mobile phones. At the  time of my first Nokia in the mid-90s, the range of products was still  manageable and it was relatively easy to choose the phone you wanted.  Today, there seems to be a new model, slighty different from the  previous every couple of weeks or so...maybe not, but it certainly feels  that way. And it's almost impossible to choose the phone you really  really want, particularly if models are linked to subscripton plans with  special prices or offers. Not only is it hard to choose the phone you  want, chossing the right subscription is even worse. Do we really need  this plethora of models and plans?


Perhaps  mobile phones should come like cars or like Dell computers, with a few  basic models or shells, which you may then configure as you wish. At  least then you would be able to assemble a phone with the functions that  you really need. While I do love my Nokia E52 which I bought a year ago and which by now is hopelessly obsolete, I  have never used most of its functions, but I wanted a phone with integrated e-mail,  that easily syncs up with Outlook, that has Internet and WLAN, and GPS for navigation. What I didn't need  were all all the options for personalization and hundreds of ringtones or  whatever. But such is product variety these days that if I want to have some functions I need to buy a phone that has a c..pload (pardon my French) of other and to me uneccessary  functions as well. Why is that so? And why does Nokia have some 100+ models to choose from in Europe, but only 20+ model in the US?


I remember when I bought my first car in the early 80s. Back then cars came with a limited range of choices, both enginewise and interiorwise. Usually there would be an austerely fitted basic version, a  somewhat acceptable mid-range version and a  all-you-want luxury version, perhaps even a sports version. There were very few, if any, optional extras, and more often than not, only available if you had already bought the luxury version. Today, my new Skoda Fabia, while having merely three possible lines and only seven possible engines, has a list of no less than 50 factory-fitted or dealer-fitted extras. Is that too much? A friend of mine, also a Skoda driver has been waiting for more than six months for his new car because there are less then 20 people worldwide desiring his particular configuration and the Skoda factory is not producing a configuration until there are at least 20 orders for it. Perhaps a sign of too much product variety?


And indeed, it is the automtive industry that is at the heart of the article I mentioned at the beginning of this post. The article analyses production and sales data of 226,106 passenger cars,  comprising of three models of one vehicle manufacturer sold across four  global market regions. The authors propose and validate product variety measures based on  actual customer orders, and empirically demonstrate how these measures  can be used to assess the impact of late configuration and option  bundling strategies. Moreover, the paper highlights how actual variety differs from theoretical variety  in practice, which in turn co-determines the effectiveness of  mitigation strategies applied by firms.Obviously, having too much variety is not a good thing, if this means a wide range of flexible solutuons to accomodate this variety and a whole battery of contingency plans in case these varieties cannot be supplied. Product design does have an impact on supply chain risk.


Perhaps I should consider the lessons learned in this article the next time I decide to buy a new car? My friend, the one I mentioned above, is already regretting his choice. Not the choice as such, but the consequences it had for the lead time for this new car that he is stil waiting for. Anyone with similar experiences?


Reference: Thomas Stäblein, Matthias Holweg, Joe Miemczyk, (2011) "Theoretical versus actual product variety: how much customisation do customers really demand?", International Journal of Operations & Production Management, Vol. 31 Iss: 3, pp.350 - 370