A prediction model applied to the spread of digital cinema
Marco del Mancino

Ladies and gentlemen, good afternoon!
Firstly I’m what you might call a film buff, in second place I’m a free-lance market researcher for MEDIA Salles. But to pay my bills I work for an American drug company. In the next 30 minutes I’ll try to answer a question that you may already have in mind: when will digital cinema be a tangible reality in my town, too?

I’ll try to answer this question by letting you glance at the future with a reliable forecast about the future diffusion of d-cinema on cinema screens worldwide. Let’s face the truth: I know that most of you may not trust forecasts very much. Probably because you have heard many predictions on the future of d-cinema and they have been quite different from one another and very changeable over time. I believe you’re right to be a little bit sceptical about previsions, in particular when a phenomenon is just beginning, as d-cinema is. Forecasts are not all good or all bad, forecasting is a science which uses complex tools that have to be wisely deployed. So I’d like to give you some advice for judging whether a forecast is good or not. A good prevision is a prevision with good, realistic, agreable assumptions. So if you don’t agree with the hypothesis on which a forecast is based, or the person who made the forecast doesn’t doesn’t state his hypothesis clearly before forecasting, well, trust me, don’t even listen to that!
Forecasting is not purely mechanical, you cannot start without having an idea of the every-day life of the market you’re dealing with.
Now I shall try to explain to you why in my opinion this prevision is based on good assumptions.

I’m sure that after this training course you must know everything you have to know about d-cinema, so let me just summarize with a statement that I like very much. It’s not mine, but comes from John Fithian, President of the influential American exhibitors’ association: “D-cinema is the biggest technological change in the field of the cinema since the advent of sound”. This is because d-cinema could really change the face of what we call the cinema exhibition business today. Because adopting d-cinema is not a mere technical matter, but a more complex process, in the first place a financial and commercial process.
In choosing the assumptions for my forecast, I first tried to understand what the most important pros and cons are in choosing d-cinema for the main players involved in the cinema supply chain.
In particular, I concentrated on the thinking of the potential adopters of this new technology who are, obviously, the cinema exhibitors. I have read widely and talked to exhibitors from Italy and other European contries as well as the Usa.
It’s a quite complex situation; it is not easy to decide when to adopt d-cinema (because we all know that is a matter of when, not if adopting it). In my opinion the problem is that there are important advantages for both distributors and exhibitors, but for the former the benefits are short term (because it’s mainly a matter of cost cutting), while for the latter the benefits are long term (because it’s a matter of redefining one’s business model to get more revenue through digital cinema).

So what I did first was to get inside your minds. I tried to identify myself with you – remember I’m not a cinema exhibitor – and think as you do when you wonder whether to opt for d-cinema (I mean choose to get rid of the 35 mm projector and start using a DLP cinema one with an appropriate server). This is the first brick I laid at the basis of my forecast.
When exploring your minds, what did I find out? That there is a tug of war going on: on the one hand you have the advantages in terms of cost reduction, flexibility and more revenues that could come from D-cinema and on the other hand you have the costs and risks associated with it. Therefore if the advantages win, the digital technology will quickly diffuse worldwide: in fact there is nothing to stop exhibitors going to Barco tomorrow, taking 75 thousand euros from their pockets and buying d-cinema, thus guaranteeing a promotion for Glenn Wastyn.
On the contrary, if the cons win, we shall see the more pessimistic vision circulating in the sector, which contemplates the two types of technology co-existing for a long period, something that would happen only if the problems due to the definition of standards or to an “excess of inertia” (the “Shall-I-go-first-no-after-you” syndrome) failed to be solved. How can we forecast which “team” will win?
Well, I know that we are talking of a b-2-b market that is less driven by irrational behavior than a consumer market, (for example the I-pod market), but I don’t forget that exhibitors are human beings and all human beings are social animals and thus look around before taking a decision, following a sort of imitation behaviour pattern.
So I have assumed that, for the adoption of d-cinema, what is relevant is mainly the information that circulates among potential users through direct contact with other operators already using the new technology or, to a lesser extent, through micro-social events, such as conferences and specialised events like today’s.
In particular it is clear that the more exhibitors opt for d-cinema, the less psychological resistance there will be amongst the others, thus reducing legitimate fears accompanying the change from a well-known and reliable technology (that of 35mm) to a new one to be tried out. If the faction in favour of d-cinema prevails and changes the view of exhibitors worldwide, then the process of diffusion will be faster.

These are the core assumptions that let us make a forecast, but they are only thoughts and words.
Now what we have to do is to translate them into another language that is more useful for us because unfortunately with only words and thoughts (even when they are wise) you cannot make a forecast; this is why I told you that forecasting is a kind of science what we have to use now is the language of mathematics to build a forecasting model that is a well-established tool, used in scientific management and demographic research, which has for a long time been applied successfully in various sectors, with the aim of foreseeing the spread of various phenomena: disease, information, products or, as in this case, a new type of technology.
So please bear in mind that a forecasting model is not a (magic) crystal ball that can anticipate what will happen in the future, but it makes it possible to indicate, on the basis of figures and strict mathematical logic, what will happen if the market continues to grow according to the patterns seen up to the present.
This is why I have called my work a reliable forecast: you can agree with it or not, but at least it provides a solid basis you can use for discussion.
Amongst the various models offered by the literature in this sector, the epidemiological pure imitation type is the one I have chosen for our needs.

So to sum up: first we have tried to gain an insight into how exhibitors choose to adopt d-cinema, then we translate this into mathematical language but in order to make a forecast one element is missing: let’s call it the x factor. Thankfully, scientists have demonstrated that we can quite easily make an estimation of it just by using the data of the first year’s sales after the lunch of the new technology in 1999 with the famous screening of George Lucas’ Star Wars-the Phantom Menace.
So as often happens, we look at the past not because the future will be exactly the same, but to make use of it in predicting the future.
To make our forecast we just need to find the sales history, or the yearly number of “digital screens” throughout the world from the introduction of the technology up to now. Luckily we can rely upon the precious free services from MEDIA Salles, that publish these figures every year in their European Cinema Yearbook.
I’d like to point out that we take into consideration only cinema screens, so excluding post-production houses or private screening rooms. Today there are probably a little under 3 thousand d-cinema screens worldwide.
As you can see adoptions have actually followed an almost perfect exponential growth curve, supporting the hypothesis of an “epidemiological” spread characterised, in this introductory phase, by a very small number of innovators, growing more and more quickly year by year and thus thankfully confirming the validity of our model.
This may be an explanation of why the forecasts you may have heard for d-cinema are very different from one another. Especially if a solid mathematical model is lacking for reference, it’s difficult to anticipate developments in processes of an exponential nature. Typically, these phenomena tend to be under-estimated in their very early stages and then over-estimated when the exponential growth becomes more evident (just think of the internet economy). Please note that we have to be fair: d-cinema has grown exponentially, but is still a real niche market. In fact only about 3% of cinema screens worldwide use d-cinema.

Well, finally here is the forecast we have obtained not by using a magic wand, or because this is just my guess, but using the logical process I’ve showed you.
For some of you this may be merely a coulored chart, and you’re not wrong. This is not a maths lesson - but we want to give advice and information for your own cinema business so the question is: what does this forecast say to an exhibitor who’s wondering whether to opt for digital cinema?
One: if you’re a sceptical person who thinks that d-cinema is science fiction, this forecast says that our horizon is perhaps not short-term but time marches on: in 5 years d-cinema will shift from being marginal to a market that is worth billions of euro (I have estimated about 8 billion euros). The peak of adoptions will already have taken place at 2013 (with over 20 thousand new installations), a year in which technology based on 35mm film should be overtaken. Two: it is interesting to note that the model foresees a particularly slow phase in the completion of market saturation: over 9 years for the final 1% of screens to adopt d-cinema. This forecast is coherent with a sentiment that is widespread in Europe, i.e. that a small number of cinemas will be more resistant to the changeover to digital, either because of being economic-commercial outsiders or because of an extra-rational choice which will lead them to continue using technology by then considered outdated. For example, I am reminded of Davide, a friend of mine, who owns a little country cinema in a beautiful little town in Tuscany and who’d like to screen only Fellini and Antonioni!
Third practical implication: our model, although perfectly compatible with an initial phase of exponential increase as has actually happened, does not assume constant growth: as you can see the yearly growth rate is going to stay high until 2010, after which it will begin to fall rapidly.
This is not a secondary discovery, in fact if you take my advice, you won’t trust anybody who says that a market, any market, will grow exponentially forever (I mean that the yearly growth rate will never fall). People used to say this about the internet economy before they learnt their lesson from the explosion of the dotcom bubble in 2000.

Maybe some of you will remember that MEDIA Salles published last year a prediction of mine in their newsletter “DGT Online Informer” no. 12.
Well in order to test the reliability of this forecast, I have tried out my new forecast by matching it with the previous one.
As I said, the forecast is based on the sales history, so every year we have new data and we can make new forecasts, but if every new year the forecast changed substantially from the previous year’s this would mean it was totally unreliable. Well I’m happy and proud to say that, despite the extraordinary growth recorded last year, the effects on our forecast are modest: an acceleration in the growth process of only one year.

I’m proud because in 2006 a totally unexpected little revolution happened.
We can say that the watershed has defined the mood of those keeping a close watch on the spread of d-cinema internationally. After the increase in the number of screens adopting digital projection technology in 2005, the idea circulating in the sector was that 2006 would be slower, “a transition year, with a more moderate growth rate compared to the previous year”.
For example Michael Karagosian, who is an American veteran of digital technology applied to the cinema, foresaw a slow-down in d-cinema adoption since the 2006 data was published.
As usual, his thoughts are analytical and interesting.
To sum up, we can say that he considered new d-cinema installations would be slowed down by specification and interoperability problems, since he says that specifications do exist in some form today, but are not complete and not sufficiently detailed to ensure interoperability at the level exhibitors expect. Apart from this, Michael thinks that specification and interoperability problems regard not only the playback products, but the entire supply chain. So, in his opinion, the problem is that exbibitors making an investment now may find out that a better and cheaper system will be available shortly.

Michael was not the only one who, until a few months ago, thought that 2006 would be a transition year, since the market was supposed to be at the point of having to “cross the chasm” of the drop in sales that is sometimes seen after the launch of a new product, i.e. when purchases by innovators and early adopters are no longer present and a way has not yet been found of getting through to the late majority, as suggested by the famous book “Crossing the Chasm” by Goeffrey. A. Moore. Even though I don’t appreciate Moore’s theory very much, because it is a very simple, typical, American way of thinking.

On the contrary, what actually happened in 2006?
Look at this chart: at the end of December 2006 the number of digital screens had risen to 2,866, with a 383% increase compared to the previous year! An exceptional growth rate - in fact it’s the highest yearly rate growth ever recorded since the introduction of d-cinema in 1999, even higher than the growth rate in the first couple of years when there were only a few dozen digital screens.

We have howerever to consider that this enormous growth is actually an average of very different rates. Asia and Europe continued to grow as in the past, whilst it was North America that took a stride forward with a growth of 1,031%.

What were the reasons? First of all it should not be forgotten that until last year the US were still a few steps behind in the process of introducing d-cinema: in Europe, for example, there were more digital screens than in the US in 2005, thanks also to the incentives offered by public financing, such as the Digital Screen Network (as you know, the UK allocated almost 18 m euro of public money to digital cinema).
Apart from that, an answer may come to mind: is the US market rolling out? We all know that this is the event that the whole sector is waiting for as the engine for worldwide spread of the new technology.
Here you have some different opinions. Glenn Wastyn disagrees: he thinks that acceleration in the d-cinema installations in North America is largely due to the effects of an aggressive promotional policy by AccessIT’s Christie/AIX, the leading player in America’s d-cinema market.
Apart from that, Michael Karagosian goes further. He thinks that: “These numbers are misleading because a lot of d-cinema projectors were sold in 2005 but in 2006”. As far as he knows, there were only a few selling and this growth in the number of digital cinema systems is due mainly to the systems that were sold the year before and were then installed in cinemas.
So, although we are probably not witnessing the roll-out of digital cinema in the US, the large growth recorded there is visible to everyone and must not be underestimated. It demonstrates the vitality and the unpredictable nature of a market that is increasingly difficult to reduce to a niche for the fanatics of new technology, and one that is capable of sudden leaps forward.
One closing remark must be made: reality is obviously (by definition) more complex than the model proposed and, like any forecast, it must be taken with a pinch of salt. I hope that it is now clear that a good prediction is a prediction with realistic even if few assumptions. For example, we can modify some hypotheses used in the forecasting model and lookat what happened to our forecast, in order to have different scenarios. Once again, not in an automatic way but taking care that the changes in assumptions are realistic.
This is also because it must be remembered that there is a sort of trade off: the accuracy of the model’s forecasts can be improved by introducing further elements, so the model could become just a little bit more precise. But it would be too complex and I assure you that when you’re forecasting, it is better to work with a simple tool than a complex one.
Let’s look at some examples.
We can consider that the the size of the potential market (represented by the over 105 thousand screens in operation throughout the world) varies over time.
It is certainly a more realistic hypothesis, but the point is that the forecast for worldwide screen growth is very small in comparison with the d-cinema growth rate (less than 5% in the next 5 years). An optimistic scenario could be that d-cinema will change the cinema exhibition business model so much (diversifying the sources of theatrical revenues thanks to alternative contents) that the cinema business will experience a new springtime and more cinemas will be opened; in this case the growth rate of screens will be higher and we’ll have then more new digital screens opening yearly but also a longer process (because the market will be larger). Let’s say that 99% of the market will adopt d-cinema not in 2019, but in 2023/4. Moreover, price of the projectors will certainly drop, thanks to scale and experience economies and the producers’ penetration strategies, (for example the price of digital projectors has fallen at an average of -16% each year since their introduction onto the market). So we can predict that, considering the effect of price reduction, the process will be a little faster compared to our basic scenario. Moreover we might imagine that the decision to adopt d-cinema could also be influenced by an external source of communication (for example, the decision by the majors to change to d-cinema) who would certainly give a precise sign to all exhibitors worldwide who want to stay in business: buy d-cinema. Besides, the future speed of d-cinema spread could be altered by technological change over time, even though the whole industry is working to prevent this from happening, it is nonetheless not impossible that, in a pessimistic scenario, one day the technology you’ve bought will need to be updated or even changed because the majors want different standards or, in an optimistic scenario, a producer like Barco might, perhaps, introduce a lower-cost (costing let’s say 20 thousand euros), lower-quality, but reliable projector, so that second-release exhibition markets that would not have not bought d-cinema can anticipate their technology changeover.
What is more, we hope that the process will not be slowed down by restrictions because Barco and other producers have all the resources to keep up with the market demand.

Lastly we could imagine that the decision to adopt d-cinema might be influenced by the varied nature of the potential adopters, i.e. the exibitors.
It’s not so difficult to forecast that the US market will act in a different way from the rest of the world since in the US the world of cinema exhibition is characterised by greater concentration (the Regal cinema chain alone possesses as many as 6,386 screens), which means that when a chain decides to change to d-cinema, this causes a sudden leap forward in the total of digital screens according to a process of stochastic diffusion, which is different to the more uniform one (and thus more easily to be foreseen), followed by markets in Europe and Asia.
The extraordinary growth in the US can be explained as follows so: certain cinema chains have decided to adopt digital cinema and therefore the count of total installations follows from the size of the chain (in fact here we have an example with Carmike).
As a confirmation of this observation, the average number of digital screens for every cinema having at least one d-cinema installation grew from 1.3 in 2005 to 2.9 in 2006, showing that it is mainly the multiplexes (generally part of large chains) that are choosing d-cinema and no longer a few pioneer independent exhibitors. In the light of these figures and considerations, what seems most important to us (and what we are already working on) is to elaborate two different forecasts: one for North America and one (at least) for the rest of the world (including the EU).

This is similar to what an important UK-based research company in the media industry, Dodona Research, has managed to do, forecasting that by 2011 half the cinema screens in the US will be digital, while, on the contrary, there will be not more than 10% coverage in Asia and 20% in Europe.
It’s interesting to note that if one takes a weighted average (considering that in the US there are almost 40% of the world’s cinema screens) he gets a forecast that is quite similar to the one I am presenting today (i.e. 30% digital screens worldwide in 2011).

In conclusion, you don’t have to believe faithfully in my forecast just as you don’t have to believe any others. I have explained to you the assumptions that underpin this forecast, so that you can, on the basis of your own opinion and experience make on, evaluating it.
In other words: I’d like you to take my work as a point of reference for your own free decision-making process. In particular, if you agree with my assumptions, please take home just this: d-cinema is not science fiction, but it will be a reality in your own backyard in a few years. Let’s face the truth: cinema exhibition is usually the weak ring in the actual cinema supply chain - therefore be prepared or be overwhelmed. In fact only those who are prepared - as soon as d-cinema becomes a widespread reality - will be able to reap all the advantages that d-cinema can offer and avoid most of the problems it might cause.
This is because d-cinema is a real revolution. It has important implications, not only in terms of business and technology, but also in socio-cultural and therefore political terms.