How A Beer Game Explains Nokia's Supply Chain

May 30, 2013 8:55 PM ETNokia Oyj (NOK)AAPL, MSFT, TMUS84 Comments
Abu Bakr Hussain profile picture
Abu Bakr Hussain

Nokia (NYSE:NOK) and beer don't appear to have a lot of parallels at first glance. Beer is the most widely consumed alcoholic beverage in the world with huge quantities consumed during major holidays and other festive occasions. Nokia on the other hand is a beleaguered mobile phone manufacturer which has only just begun to get back on its feet after a disastrous few years. Yet the peaks and troughs of beer consumption have a great deal to tell us about Nokia's supply chain problems. Why they might have happened and why they, thankfully, may well be over.

The Beer Game

The beer game sadly is not a drinking game. Though emanating from MIT in the 1960s, it may possibly be the result of one. What it actually does is to take the variability of beer consumption and to translate that into a supply chain model such as the one shown below which I'll be using throughout this article.


This game is actually simple to play and I would invite all interested readers to have a go by clicking on the source link after reading the article.

This four step supply chain models what happens from the production of beer, through to its distribution, onto the wholesaler and finally the retailer where it is bought by the consumer. The model is deliberately simple to understand, but the results from running the model and trying to solve the problem of supply chain coordination have vexed supply chain analysts for decades and led to several PhDs on the subject. It is also a very insightful way of looking at Nokia's supply chain and goes a long way towards showing just why Nokia has had well documented problems in coordinating its supply with consumer demand.

The problems the beer game models are many. Firstly as the preceding picture shows, there is a time lag from an order being made, to it being processed, to it being delivered. This seems reasonable but the problem arises in that no link knows when its next order will arrive. So, a retailer does not know exactly how much beer it will sell that week. In turn, the wholesaler does not know how much beer the retailer will order that week. In the third link of the chain, the distributor doesn't know the amount of beer the wholesaler requires that week and finally at the end of the chain, the factory does not know how much beer the distributor will require that week. This model thus is designed to help us see how low visibility of what is going on in other links affects inventory decisions.

That is not the only problem however. Each link must also consider that holding surplus stock results in storage costs. Therefore no link wants to carry excess inventory. Yet on the other side, backlog costs are also problematic and cause knock on effects for other links in the chain.

To best illustrate the model, I have run a simple variation.

A Beer Game Example

In this model, all links are controlled by the computer with customer orders to retailers generated at random. This also means I have no control over the chain and no foreknowledge of what will transpire. I will periodically halt the running of the model and give a takeaway as to how it relates to Nokia's supply chain.

Day 0:

On Day 0, each link of the supply chain is perfectly stocked with an inventory level of 50. However, this perfection is about to be damaged by the retailer receiving its first order.

In reality this ideal can never be achieved. For example, in November 2011, the month of the Lumia 800 launch, the inventories were stocked full of Symbian phones. Not only that, but most of Nokia's supplier and carrier relationships, manufacturing capability and marketing were still geared towards moving Symbian phones and not Microsoft (MSFT) Windows Phone Lumias. This overhang took a long time to dissipate with the last Symbian stock being sold off only in 1Q2013.

Day 10:

Just 10 days in, the picture is completely different. The retailer, wholesaler and distributor have all run out of stock. The small red letters represent orders upstream with several in process at present. The yellow boxes represent orders en route. As can be seen, the distributor in particular has a severe order backlog which it has compensated for by ordering several times from the factory, none of which have arrived yet as it takes 7 days for each parcel to arrive. This has caused it to have a large backlog cost of $574, a problem the retailer is not facing.

With this we can already see one of the problems with supply chains. Not all links experience the same levels of profit or of loss. In this case, the retailer with total inventory costs of $100 is considerably better off than the distributor with total inventory costs of $649.

This asymmetry in supply chain costs explains a lot about Nokia already. Nokia receives a lot of criticism that it does not produce enough units, but one must consider that it hasn't received enough orders from down the chain for it to process. It also means that as carriers do not have an incentive to minimize Nokia's supply chain costs, they would rather order piece-meal stocks and let Nokia carry the burden of inventory costs. This would explain why for instance the L920 was sold out at several places in its opening month with limited stock available.

Day 25:

25 days in, it appears that the retailer, wholesaler and distributor have all experienced supply backlogs causing their costs to escalate rapidly. The factory on the other hand has not experienced a single backlog in the last 15 days which has led it to having spent the least on inventory costs at this stage in the game.

Nokia hasn't had this situation arise in the last two years. It is however very similar to Apple's (AAPL) situation where it deals with backlogs downstream on a regular basis whenever a new iPhone launches.

Day 50:

The next 25 days, it appears there has been a seasonal slowdown. However, the retailer did not react appropriately to this slowdown. Instead of drip-feeding small orders to the wholesaler throughout the period, it sent a bulk order which as can be seen completely emptied out the wholesaler. The small red letter that can be seen leaving the wholesaler, I predict is a very large order which is about to cause a severe backlog for the distributor. Note that all of this is occurring despite there being less total sales than in the preceding 25 day period. This goes to show that low volumes present just as much an issue for supply chains as do high volumes. Again this links in with Nokia's relatively low volume of Lumia sales. Averaging between 4-6 million a quarter, these are a small fraction of the 60 odd million total mobile phone sales it makes. Nevertheless, as the above model shows, these low volumes can still cause trouble.

Day 75:

It appears a large order indeed was sent through the system which caused the factory to briefly have its first backlog since the beginning of the simulation. Nevertheless all links in the chain are beginning to return to ideal inventory levels though all are running slightly low. This is probably due to customer demand remaining relatively stable. In reality, in the smartphone market, customer demand is highly variable dependent on seasonality as well as phone release dates. Nokia could thus never hope for such a pleasant scenario.

Day 100:

The final 25 days of the model were another quiet period. Interestingly enough, the factory processed a solitary order, went to 120 units of inventory and sat idle for the rest of the period. If we consider Nokia's factories in this situation, it is inconceivable they would sit idle. It would be considered a waste of resources. They would thus continue to churn out units, regardless of actual demand levels. This would then pose a problem for it as it would have to reduce ASPs to clear this inventory. Thus, with lower than expected feature phone sales in 1Q2013, we can expect to see greatly reduce ASPs as Nokia clears this surplus stock.

With the model having concluded its run, let's look at the broader story of its lessons.

Beer Game Takeaways And An Insight Into Nokia

With the retailer being fortunate enough to be closer to the end-user, it appears it has been best able to manage its total costs in this instance. However as the end-user order volume is random, the above result is not a given in all circumstances. Depending on the behavior of each link, each link should be able to successfully reduce its total costs. For example, if the distributor had run higher inventories, it would both have benefited itself (by reducing backlog costs) as well as the factory which would have run lower storage costs. This would have resulted in a reduction in total supply chain costs and is one of the key reasons that all links in a supply chain should work together.

The reason this game is particularly popular is because it shows the difficulty in managing the randomness of demand by consumers. The resultant timings in new orders being placed upstream leads to the "bullwhip effect". This is where at each stage, imbalances in supply exacerbate leading to large discrete orders being sent through the supply chain, causing backlogs at each step. This was especially noticeable at Day 50.

This bullwhip effect is actually something Nokia in all likelihood has been going through which has resulted in severe delays in getting enough phones to consumers. So if one thinks back to June 2012, we had a large surfeit of L800s and L900s which had been caused by Nokia over-ordering. We can infer this from that fact that the WP8 launch had effectively osborned these products causing a decline in Nokia Lumia sales of 30% from 2Q2012 to 3Q2012. Nokia probably also did not help itself by having overoptimistic projection of L800/L900 sales which the market shared as the following estimate by Morgan Stanley shows.

In fact, in 2012 Nokia sold just 21 million Symbian handsets and 13 million Lumias (approximate figures based on company reports). This would undoubtedly have generated an oversupply of product. As supply creates its own demand, Nokia would have had to severely reduce ASPs to clear this inventory. We can see evidence of this in low smartphone ASPs in 1Q2012 of $190.

Fast forwarding to the present day, with reports of low stocks in newly launched Nokias such as the L520, we can now surmise this inventory has now cleared which has resulted in considerably higher ASPs of $249 in 1Q2013.

The beer game thus is a very useful beginning insight into the complexities of supply chain management for Nokia. However, to properly consider Nokia's difficulties we now need to turn to its real supply chain.

Nokia's Real Supply Chain

Nokia's actual supply chain is much more complex than the beer game's which makes it considerably more difficult to manage. It involves many more links for a start. Furthermore, rather than a "chain" going in a linear fashion from upstream to downstream, it is better to consider it a web. In this web, links at the same stage can impact throughout the network. To illustrate this, I have created a model of the Nokia supply chain below.


Source: Author's analysis based on publicly available information

As the above diagram shows, Nokia is not dealing with a simple chain at all. In my model, there are six stages and 20 separate links which is still a considerable simplification of Nokia's real supply chain. Each one of these links has the potential to generate a knock-on effect to all other parts of the chain. So for example, if China were to limit rare earth exports as it did in 2010, this would quickly begin to impact on Nokia's subcomponent manufacturers feeding through to component manufacturers causing a reduction in supplies reaching Nokia. This would reduce total output at Nokia's production facilities despite all other inputs being freely available.

This is just the beginning of the complexity; I have documented other ways in which Nokia's supply chain is more complicated than the Beer Game's below.

The Complexities Of Nokia's Supply Chain

1. Nokia does not produce a single product - It produces several different phones, all which must share valuable machinery time at Nokia's production facilities. It currently produces 9 separate WP8 phones (Lumia 920/925/928/822/820/720/620/520/521) compared with Apple's 3 (iPhone 5/4S/4). Thus Nokia needs to manage its expectations of consumer demand much more carefully, especially as different phones will have different production times. For example, a L920 may require 25% more time to manufacture than a L520 (hypothetically), but will have a higher gross margin. This is not even including the different color variations, the Asha line and general feature phones. When we factor this in, Nokia will be juggling manufacturing capacity between literally over a hundred different phones. This makes it understandable why cyan and gray L920s have had very limited availability globally.

2. All Nokia phones have variable consumer demand over time - They will be relatively popular at launch, however sales will diminish over time. Nokia needs to be able to calculate the rate of decline and decide when to reduce production and by how much. At the moment, L920 demand will be tailing off with it being a 6 month old phone. How quickly to reduce the production of the phone (and to make space for new flagship phones such as the L925/L928 and the rumored Eos) will be vexing Nokia's supply chain management as of right now.

3. As Nokia has very few physical stores and does not typically sell directly, it is one link removed from its consumers. This places its orders in the hands of carriers and independent retailers. As these carriers have coped with a severe decline in Symbian phone sales, it is understandable that they would order few new Lumia devices in the expectation they would not sell many. Nokia thus has to market to its carriers also and convince them that the Lumia will sell more than carrier estimates. This is probably one of the key reasons Nokia develops carrier specific variants of its phones such as the L822, L928, L920T and so forth.

4. Bullish or bearish inventories? - Linked with three, if carriers are generally pessimistic of total Lumia sales and Nokia is bullish, Nokia supply chain managers are caught in a very problematic dilemma. If they produce more than their carriers may expect, they may be burdened with excessive inventory. This is what happened with the Lumia 800 which was produced in vastly more numbers than carriers wanted. On the other hand, if carriers face unexpected consumer demand, they will ask for more orders from Nokia. If Nokia is unable to provide these orders, the carriers will divert their customers onto other available (non-Nokia) models in stock.

5. Nokia's marketing efforts need to be coordinated with its supply chain to shift stock at the right time. This is much more difficult than in the case of beer. With beer, one can run weekly promotions to shift excessive inventory or raise prices to manage tight stock levels. With phones, this is much more difficult. Price reductions typically cannot be reversed. Additionally with Nokia not generally selling directly to consumers, Nokia marketing is typically more aimed at raising awareness. This makes it more difficult to carefully calibrate sales to manage inventories better. Apple with its physical stores can run the right promotions to manage its inventories to the desired level.

So What Is The Future Of Nokia's Supply Chain?

All I have stated so far indicates that Nokia's supply chain is systemically difficult to manage. Unlike Apple, it is one step removed from its consumers, thus finding it more difficult to manage demand. It also makes several different variations of its phones. Apple with its policy of producing only three variants at a time (for evidence of this, see the iPhone 4 disappear from shops in the next 6 months as the new iPhone makes its appearance), has a much simpler supply chain to manage.

However, a key difference between Apple and Nokia finally appears to have been overcome by the latter.

Take a look at this graph of iPhone unit sales.

Source: CNN

Notice the yearly increase in sales. For sure there are peaks and troughs, but the general trend is upwards. This certainly does not match the "beer game" where sales average out over time. Apple thus doesn't really have a "bullwhip effect" to deal with where it can find itself with surplus stock.

Now let's take a look at Lumia sales over the last 2 years.

Source: Nokia Company Reports and author estimates

Again, one can see a general upward trend. This figure has been disguised to many carriers and outside observers by the general downtrend in Symbian sales. With Symbian phones no longer in Nokia's supply chain, this upward trend will become a lot more noticeable to everyone.

It also means that Nokia's supply chain will fare a lot better over the coming years. The "bullwhip effect" where we see shortages of Lumias will persist for a while. However I do believe that the news that Lumias do indeed sell well is finally getting through to carriers. I believe that over the coming months, they will continue to order more from Nokia. Even better this should create a virtuous cycle. For example, as T-Mobile (TMUS) notices the L521 continually sell out, it will ramp up its order requests to Nokia. In the future, to cope with this increased demand, it may also increase the amount of marketing it conducts on the Lumia to ensure it clears this stock. This will mean when we open the T-Mobile webpage, in the future we should see Lumias prominently advertised. This is not yet the case as its current front page doesn't even note that the L521 has been created especially for it.

Source: (notice the absence of any Lumias advertised)

It will be an interesting exercise for readers to go back and check carriers over time. As Lumia sales increase, I predict more and more concerted advertising by carriers, both to take advantage of the Lumia popularity and to shift their stock causing yet more sales. I do not think it unreasonable to expect a Lumia to be advertised on the front page of the T-Mobile website by this time next year.

Does this mean that an increase in Lumia sales magically eliminates Nokia's supply chain problems? Not at all. In the early 2000s Nokia's supply chain was a perennial favourite in business school discussions. Now in the latest Top 25 Supply Chains in the world announced May 30th 2013, it doesn't even figure in Gartner's analysis.

Source: Gartner

Now, I do slightly question the methodology behind this ranking. It seems to me to be a retrospective analysis based on what supply chain analysts think are the best supply chains in the business. As actual supply chains are a closely guarded secret, these opinions are formed just on publicly available information and can thus be erroneous. I do think it is relatively accurate however that Apple's supply chain is the best in the smartphone sector. It isn't the best because Apple make the best phones (I do not believe they do). It is the best because they make the least variations of their phones and they manage the entirety of the chain excellently from upstream to downstream. As an aside they also know how to market obsolete phones which the iPhone 4 clearly is. Compare this with Nokia which cannot now sell the Lumia 800 unless at a steep discount. This is despite the L800 being barely 18 months old compared to the iPhone 4's 3 years on the market and with both having similar specifications.


So after all of this, why am I still bullish on Nokia? I believe at present, Nokia is priced for bankruptcy. It won't go bankrupt as its net cash which is holding steady at a healthy $5.8 billion demonstrates. Lumia sales continue to increase which can only result in an upside for Nokia. Furthermore Nokia's share of Nokia-Siemens Networks (NSN) alone is valued at $9.6 billion by fellow Seeking Alpha contributor Jonathan Verenger. This means NSN is worth approximately 80% of Nokia's total market capitalization as of today, which is an imbalance the market will surely correct resulting in significant upside once Nokia gets back into positive EPS territory. All of this is without even factoring in HERE which with its integration into Windows Phone mapping solutions will become an increasingly valuable division for Nokia.

None of my reasons for owning Nokia shares include its supply chain management skills. I do hope however that this article has offered a more sympathetic insight into why Nokia has faced problems in managing its supply chain. It is also why I am not frustrated by these backlogs and shortages which new Lumia phones seem to face. I believe that as production ramps up and as carriers work with Nokia better, these problems will dissipate. It won't make Nokia have the best supply chain in the business. But it will make it have a competitive supply chain.

Disclosure: I am long NOK. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.

This article was written by

Abu Bakr Hussain profile picture
MSc Project Management Graduate 2012, PRINCE2 Practitioner, APMP, Winner of APM Scottish Project Challenge 2012, Recipient of Technology Strategy Board start-up Grant 2013. Risk Management Product of the year 2013 with Datatecnics. Risk Management Innovation of the year 2015 with Datatecnics (beating Google!) Specialisms: Projects - especially how they relate to strategy; start-up phase; business case design; Supply Chains Direct message me here or on: LinkedIn: New Twitter account: @AbuB_Hussain

Recommended For You

Comments (84)

To ensure this doesn’t happen in the future, please enable Javascript and cookies in your browser.
Is this happening to you frequently? Please report it on our feedback forum.
If you have an ad-blocker enabled you may be blocked from proceeding. Please disable your ad-blocker and refresh.